{ "ICSI--train--1": { "original dialog id": "Bed012.json", "dialog index": 1, "original dialog info": { "dialog history": [ "B : So I guess this is more or less now just to get you up to date , Johno .", "B : This is what , uh ,", "C : This is a meeting for me .", "B : um ,", "B : Eva , Bhaskara , and I did .", "D : Did you add more stuff to it ? later ?", "B : Um . Why ?", "D : Um . I don't know .", "D : There were , like , the you know , @ @ and all that stuff .", "D : But . I thought you you said you were adding stuff", "D : but I don't know .", "B : Uh , no .", "B : This is ", "B : Um ,", "B : Ha !", "B : Very nice .", "B : Um , so we thought that , We can write up uh , an element ,", "B : and for each of the situation nodes that we observed in the Bayes - net ?", "B : So .", "B : What 's the situation like at the entity that is mentioned ?", "B : if we know anything about it ?", "B : Is it under construction ?", "B : Or is it on fire or something happening to it ?", "B : Or is it stable ?", "B : and so forth ,", "B : going all the way um , f through Parking , Location , Hotel , Car , Restroom , @ @ Riots , Fairs , Strikes , or Disasters .", "C : So is This is A situation are is all the things which can be happening right now ?", "C : Or , what is the situation type ?", "B : That 's basically just specifying the the input for the w what 's", "C : Oh , I see", "C : y Why are you specifying it in XML ?", "B : Um . Just because it forces us to be specific about the values here ?", "C : OK .", "B : And , also , I mean , this is a what the input is going to be .", "B : Right ?", "B : So , we will , uh ", "B : This is a schema . This is ", "C : Well , yeah . I just don't know if this is th l what the Does This is what Java Bayes takes ? as a Bayes - net spec ?", "B : No , because I mean if we I mean we 're sure gonna interface to We 're gonna get an XML document from somewhere .", "B : Right ?", "B : And that XML document will say \" We are able to We were able to observe that w the element , um , @ @ of the Location that the car is near . \"", "B : So that 's gonna be Um .", "C : So this is the situational context , everything in it .", "C : Is that what Situation is short for ,", "C : shi situational context ?", "B : Yep .", "C : OK .", "B : So this is just , again , a an XML schemata which defines a set of possible , uh , permissible XML structures ,", "B : which we view as input into the Bayes - net .", "B : Right ?", "C : And then we can r uh possibly run one of them uh transformations ?", "C : That put it into the format that the Bayes n or Java Bayes or whatever wants ?", "B : Yea - Are you talking are you talking about the the structure ?", "C : Well it ", "B : I mean when you observe a node .", "C : When you when you say the input to the v Java Bayes , it takes a certain format ,", "B : Um - hmm .", "C : right ?", "C : Which I don't think is this .", "C : Although I don't know .", "B : No , it 's certainly not this .", "B : Nuh .", "C : So you could just Couldn't you just run a ", "B : XSL . Yeah .", "C : Yeah .", "C : To convert it into the Java Bayes for format ?", "B : Yep .", "C : OK .", "B : That 's That 's no problem ,", "B : but I even think that , um ", "B : I mean , once Once you have this sort of as running as a module ", "B : Right ?", "B : What you want is You wanna say , \" OK , give me the posterior probabilities of the Go - there node , when this is happening . \"", "B : Right ?", "B : When the person said this , the car is there , it 's raining , and this is happening .", "B : And with this you can specify the what 's happening in the situation , and what 's happening with the user .", "B : So we get After we are done , through the Situation we get the User Vector .", "B : So , this is a ", "C : So this is just a specification of all the possible inputs ?", "B : Yep .", "B : And ,", "B : all the possible outputs , too .", "C : OK .", "B : So , we have , um , for example , the , uh , Go - there decision node", "B : which has two elements ,", "B : going - there and its posterior probability ,", "B : and not - going - there and its posterior probability ,", "B : because the output is always gonna be all the decision nodes and all the the a all the posterior probabilities for all the values .", "C : And then we would just look at the , eh , Struct that we wanna look at", "C : in terms of if if we 're only asking about one of the ", "C : So like , if I 'm just interested in the going - there node , I would just pull that information out of the Struct that gets return that would that Java Bayes would output ?", "B : Um , pretty much , yes ,", "B : but I think it 's a little bit more complex .", "B : As , if I understand it correctly , it always gives you all the posterior probabilities for all the values of all decision nodes .", "B : So , when we input something , we always get the , uh , posterior probabilities for all of these .", "B : Right ?", "C : OK .", "B : So there is no way of telling it t not to tell us about the EVA values .", "C : Yeah , wait I agree ,", "C : that 's yeah , use oh , uh Yeah , OK .", "B : So so we get this whole list of of , um , things ,", "B : and the question is what to do with it ,", "B : what to hand on ,", "B : how to interpret it ,", "B : in a sense .", "B : So y you said if you \" I 'm only interested in whether he wants to go there or not \" , then I just look at that node ,", "B : look which one ", "C : Look at that Struct in the output ,", "C : right ?", "B : Yep .", "B : Look at that Struct in the the output ,", "B : even though I wouldn't call it a \" Struct \" .", "B : But .", "C : Well i well , it 's an XML Structure that 's being res returned ,", "B : Oh . Mm - hmm .", "C : right ?", "B : So every part of a structure is a \" Struct \" .", "B : Yeah .", "C : Yeah , I just uh I just was abbreviated it to Struct in my head , and started going with that .", "B : That element or object , I would say .", "C : Not a C Struct . That 's not what I was trying to k", "B : Yeah .", "C : though yeah .", "B : OK .", "B : And , um ,", "B : the reason is why I think it 's a little bit more complex or why why we can even think about it as an interesting problem in and of itself is ", "B : Um .", "B : So . The , uh ", "B : Let 's look at an example .", "C : Well , w wouldn't we just take the structure that 's outputted and then run another transformation on it , that would just dump the one that we wanted out ?", "B : Yeah . w We 'd need to prune .", "B : Right ?", "B : Throw things away .", "C : Well , actually , you don't even need to do that with XML .", "C : D Can't you just look at one specific ", "B : No", "B : Yeah , exactly . The @ @ Xerxes allows you to say , u \" Just give me the value of that , and that , and that . \"", "B : But , we don't really know what we 're interested in before we look at the complete at at the overall result .", "B : So the person said , um , \" Where is X ? \"", "B : and so ,", "B : we want to know , um , is Does he want info ?", "B : o on this ?", "B : or know the location ?", "B : Or does he want to go there ?", "B : Let 's assume this is our our question .", "C : Sure .", "B : Nuh ?", "B : So .", "B : Um .", "B : Do this in Perl .", "B : So we get ", "B : OK .", "B : Let 's assume this is the output .", "B : So .", "B : We should con be able to conclude from that that ", "B : I mean . It 's always gonna give us a value of how likely we think i it is that he wants to go there and doesn't want to go there ,", "B : or how likely it is that he wants to get information .", "B : But , maybe w we should just reverse this to make it a little bit more delicate .", "B : So , does he wanna know where it is ? or does he wanna go there ?", "C : He wants to know where it is .", "B : Right .", "B : I I I tend to agree .", "B : And if it 's If ", "C : Well now , y I mean , you could ", "B : And i if there 's sort of a clear winner here ,", "B : and , um and this is pretty , uh indifferent ,", "B : then we then we might conclude that he actually wants to just know where , uh t uh , he does want to go there .", "C : Uh , out of curiosity , is there a reason why we wouldn't combine these three nodes ? into one smaller subnet ?", "C : that would just basically be the question for ", "C : We have \" where is X ? \" is the question ,", "C : right ?", "C : That would just be Info - on or Location ?", "C : Based upon ", "B : Or Go - there .", "B : A lot of people ask that , if they actually just wanna go there .", "B : People come up to you on campus and say , \" Where 's the library ? \"", "B : You 're gonna say y you 're gonna say , g \" Go down that way . \"", "B : You 're not gonna say \" It 's It 's five hundred yards away from you \" or \" It 's north of you \" , or \" it 's located \"", "C : Well , I mean But the there 's So you just have three decisions for the final node , that would link thes these three nodes in the net together .", "B : Um .", "B : I don't know whether I understand what you mean .", "B : But . Again , in this Given this input , we , also in some situations , may wanna postulate an opinion whether that person wants to go there now", "B : the nicest way ,", "B : use a cab ,", "B : or so s", "B : wants to know it wants to know where it is because he wants something fixed there , because he wants to visit t it or whatever .", "B : So , it n I mean a All I 'm saying is , whatever our input is , we 're always gonna get the full output .", "B : And some some things will always be sort of too not significant enough .", "C : Wha", "C : Or i or i it 'll be tight .", "C : You won't it 'll be hard to decide .", "B : Yep .", "C : But I mean , I guess I guess the thing is , uh , this is another , smaller , case of reasoning in the case of an uncertainty ,", "C : which makes me think Bayes - net should be the way to solve these things .", "C : So if you had If for every construction ,", "C : right ?", "B : Oh !", "C : you could say , \" Well , there Here 's the Where - Is construction . \"", "C : And for the Where - Is construction , we know we need to l look at this node , that merges these three things together", "B : Mm - hmm .", "C : as for th to decide the response .", "C : And since we have a finite number of constructions that we can deal with , we could have a finite number of nodes .", "B : OK .", "C : Say , if we had to y deal with arbitrary language , it wouldn't make any sense to do that ,", "B : Mm - hmm .", "C : because there 'd be no way to generate the nodes for every possible sentence .", "B : Mm - hmm .", "C : But since we can only deal with a finite amount of stuff ", "B : So , basically , the idea is to f to feed the output of that belief - net into another belief - net .", "C : Yeah , so basically take these three things and then put them into another belief - net .", "B : But , why why why only those three ?", "B : Why not the whol", "C : Well , I mean , d For the Where - Is question .", "C : So we 'd have a node for the Where - Is question .", "B : Yeah . But we believe that all the decision nodes are can be relevant for the Where - Is ,", "B : and the Where How - do - I - get - to", "B : or the Tell - me - something - about .", "C : You can come in if you want .", "B : Yes , it is allowed .", "C : As long as y you 're not wearing your h your h headphones .", "C : Well , I do I See , I don't know if this is a good idea or not .", "C : I 'm just throwing it out .", "C : But uh , it seems like we could have I mea or uh we could put all of the all of the r information that could also be relevant into the Where - Is node answer", "B : Mm - hmm .", "B : Yep .", "C : node", "C : thing", "C : stuff .", "C : And uh ", "D : OK .", "B : I mean ", "B : Let 's not forget we 're gonna get some very strong input from these sub dis from these discourse things ,", "B : right ?", "B : So .", "B : \" Tell me the location of X . \"", "B : Nuh ?", "B : Or \" Where is X located at ? \"", "C : We u", "B : Nuh ?", "C : Yeah , I know , but the Bayes - net would be able to The weights on the on the nodes in the Bayes - net would be able to do all that ,", "C : wouldn't it ?", "B : Mm - hmm .", "C : Here 's a k", "C : Oh !", "C : Oh , I 'll wait until you 're plugged in .", "C : Oh , don't sit there .", "C : Sit here .", "C : You know how you don't like that one .", "C : It 's OK .", "C : That 's the weird one .", "C : That 's the one that 's painful . That hurts . It hurts so bad .", "C : I 'm h I 'm happy that they 're recording that .", "C : That headphone . The headphone that you have to put on backwards , with the little little thing and the little little foam block on it ?", "C : It 's a painful , painful microphone .", "B : I think it 's th called \" the Crown \" .", "C : The crown ?", "D : What ?", "B : Yeah ,", "B : versus \" the Sony \" .", "A : The Crown ?", "A : Is that the actual name ?", "B : Mm - hmm .", "A : OK .", "B : The manufacturer .", "C : I don't see a manufacturer on it .", "C : Oh , wait ,", "B : You w", "C : here it is .", "C : h This thingy .", "C : Yeah , it 's \" The Crown \" .", "C : The crown of pain !", "A : Yes .", "B : You 're on - line ?", "C : Are you are your mike o Is your mike on ?", "A : Indeed .", "C : OK .", "C : So you 've been working with these guys ?", "C : You know what 's going on ?", "A : Yes ,", "A : I have .", "A : And , I do .", "A : Yeah , alright .", "C : Excellent !", "A : s So where are we ?", "B : We 're discussing this .", "A : I don't think it can handle French ,", "A : but anyway .", "B : So .", "B : Assume we have something coming in .", "B : A person says , \" Where is X ? \" ,", "B : and we get a certain We have a Situation vector and a User vector and everything is fine ?", "B : An - an and and our and our ", "C : Did you just sti Did you just stick the m the the the microphone actually in the tea ?", "A : No .", "B : And , um ,", "A : I 'm not drinking tea .", "A : What are you talking about ?", "C : Oh , yeah .", "C : Sorry .", "B : let 's just assume our Bayes - net just has three decision nodes for the time being .", "B : These three , he wants to know something about it , he wants to know where it is , he wants to go there .", "C : In terms of , these would be wha how we would answer the question Where - Is ,", "C : right ?", "C : We u", "C : This is i That 's what you s it seemed like , explained it to me earlier", "C : w We we 're we wanna know how to answer the question \" Where is X ? \"", "B : Yeah , but , mmm .", "B : Yeah .", "B : No , I can I can do the Timing node in here , too ,", "B : and say \" OK . \"", "C : Well , yeah , but in the s uh , let 's just deal with the s the simple case of we 're not worrying about timing or anything .", "C : We just want to know how we should answer \" Where is X ? \"", "B : OK .", "B : And , um ,", "B : OK , and ,", "B : Go - there has two values ,", "B : right ? ,", "B : Go - there and not - Go - there .", "B : Let 's assume those are the posterior probabilities of that .", "A : Mm - hmm .", "B : Info - on has True or False and Location .", "B : So , he wants to know something about it ,", "B : and he wants to know something he wants to know Where - it - is ,", "A : Excuse me .", "B : has these values .", "B : And , um ,", "C : Oh , I see why we can't do that .", "B : And , um , in this case we would probably all agree that he wants to go there .", "B : Our belief - net thinks he wants to go there ,", "B : right ?", "A : Yeah .", "A : Mm - hmm .", "B : In the , uh , whatever ,", "B : if we have something like this here ,", "B : and", "B : this like that", "B : and maybe here also some ", "A : You should probably make them out of ", "A : Yeah .", "C : Well , it", "B : something like that ,", "B : then we would guess , \" Aha ! He , our belief - net , has s stronger beliefs that he wants to know where it is , than actually wants to go there . \"", "B : Right ?", "C : That it Doesn't this assume , though , that they 're evenly weighted ?", "D : True .", "C : Like ", "C : I guess they are evenly weighted .", "A : The different decision nodes , you mean ?", "C : Yeah , the Go - there , the Info - on , and the Location ?", "A : Well , d yeah , this is making the assumption .", "A : Yes .", "B : What do you mean by \" differently weighted \" ?", "C : Like ", "B : They don't feed into anything really anymore .", "C : Or I jus", "A : But I mean , why do we ", "C : Le", "A : If we trusted the Go - there node more th much more than we trusted the other ones , then we would conclude , even in this situation , that he wanted to go there .", "A : So , in that sense , we weight them equally", "B : OK .", "A : right now .", "B : Makes sense .", "B : Yeah .", "C : So the But I guess the k the question that I was as er wondering or maybe Robert was proposing to me is ", "B : But ", "C : How do we d make the decision on as to which one to listen to ?", "A : Yeah , so , the final d decision is the combination of these three .", "A : So again , it 's it 's some kind of , uh ", "C : Bayes - net .", "A : Yeah , sure .", "C : OK so , then , the question i So then my question is t to you then , would be ", "C : So is the only r reason we can make all these smaller Bayes - nets , because we know we can only deal with a finite set of constructions ?", "C : Cuz oth If we 're just taking arbitrary language in , we couldn't have a node for every possible question ,", "C : you know ?", "A : A decision node for every possible question , you mean ?", "C : Well , I like , in the case of Yeah .", "C : In the ca Any piece of language , we wouldn't be able to answer it with this system , b if we just h", "C : Cuz we wouldn't have the correct node .", "C : Basically , w what you 're s proposing is a n Where - Is node ,", "C : right ?", "A : Yeah .", "C : And and if we And if someone says , you know , uh , something in Mandarin to the system , we 'd - wouldn't know which node to look at to answer that question ,", "A : So is ", "A : Yeah .", "C : right ?", "A : Yeah .", "B : Mmm ?", "C : So , but but if we have a finite ", "C : What ?", "B : I don't see your point .", "B : What what what I am thinking , or what we 're about to propose here is we 're always gonna get the whole list of values and their posterior probabilities .", "B : And now we need an expert system or belief - net or something that interprets that ,", "B : that looks at all the values and says , \" The winner is Timing .", "B : Now , go there . \"", "B : \" Uh , go there , Timing , now . \"", "B : Or , \" The winner is Info - on , Function - Off . \"", "B : So , he wants to know something about it , and what it does .", "B : Nuh ?", "B : Uh , regardless of of of the input . Wh - Regardle", "C : Yeah , but But how does the expert but how does the expert system know how who which one to declare the winner , if it doesn't know the question it is , and how that question should be answered ?", "B : Based on the k what the question was , so what the discourse , the ontology , the situation and the user model gave us , we came up with these values for these decisions .", "C : Yeah I know . But how do we weight what we get out ?", "C : As , which one i Which ones are important ?", "C : So my i So , if we were to it with a Bayes - net , we 'd have to have a node for every question that we knew how to deal with ,", "C : that would take all of the inputs and weight them appropriately for that question .", "B : Mm - hmm .", "C : Does that make sense ?", "C : Yay ,", "C : nay ?", "A : Um , I mean , are you saying that , what happens if you try to scale this up to the situation , or are we just dealing with arbitrary language ?", "C : We ", "A : Is that your point ?", "C : Well , no . I I guess my question is , Is the reason that we can make a node f", "C : or OK . So , lemme see if I 'm confused .", "C : Are we going to make a node for every question ?", "C : Does that make sense ? ", "C : Or not .", "A : For every question ?", "A : Like ", "C : Every construction .", "A : Hmm .", "A : I don't Not necessarily , I would think .", "A : I mean , it 's not based on constructions ,", "A : it 's based on things like , uh , there 's gonna be a node for Go - there or not , and there 's gonna be a node for Enter , View , Approach .", "C : Wel", "C : W OK .", "C : So , someone asked a question .", "A : Yeah .", "C : How do we decide how to answer it ?", "B : Well , look at look Face yourself with this pr question .", "B : You get this You 'll have y This is what you get .", "B : And now you have to make a decision . What do we think ?", "B : What does this tell us ?", "B : And not knowing what was asked , and what happened , and whether the person was a tourist or a local ,", "B : because all of these factors have presumably already gone into making these posterior probabilities .", "C : Yeah .", "B : What what we need is a just a mechanism that says , \" Aha ! There is \"", "C : I just don't think a \" winner - take - all \" type of thing is the ", "A : I mean , in general , like , we won't just have those three ,", "A : right ?", "A : We 'll have , uh , like , many , many nodes .", "B : Yep .", "A : So we have to , like So that it 's no longer possible to just look at the nodes themselves and figure out what the person is trying to say .", "B : Because there are interdependencies ,", "B : right ?", "B : The uh ", "B : Uh , no . So if if for example , the Go - there posterior possibility is so high , um ,", "B : uh , w if it 's if it has reached reached a certain height , then all of this becomes irrelevant .", "B : So . If even if if the function or the history or something is scoring pretty good on the true node , true value ", "C : Wel", "C : I don't know about that ,", "C : cuz that would suggest that I mean ", "B : He wants to go there and know something about it ?", "C : Do they have to be mutual", "C : Yeah . Do they have to be mutually exclusive ?", "B : I think to some extent they are .", "B : Or maybe they 're not .", "C : Cuz I , uh The way you describe what they meant , they weren't mutu uh , they didn't seem mutually exclusive to me .", "B : Well , if he doesn't want to go there , even if the Enter posterior proba", "B : So .", "C : Wel", "B : Go - there is No .", "B : Enter is High ,", "B : and Info - on is High .", "C : Well , yeah , just out of the other three , though , that you had in the ", "B : Hmm ?", "C : those three nodes .", "C : The - d They didn't seem like they were mutually exclusive .", "B : No , there 's No .", "B : But ", "B : It 's through the ", "C : So th s so , yeah , but some So , some things would drop out , and some things would still be important .", "B : Mm - hmm .", "C : But I guess what 's confusing me is , if we have a Bayes - net to deal w another Bayes - net to deal with this stuff ,", "A : Mm - hmm .", "C : you know ,", "C : uh ,", "C : is the only reason OK , so , I guess , if we have a Ba - another Bayes - net to deal with this stuff , the only r reason we can design it is cuz we know what each question is asking ?", "A : Yeah .", "A : I think that 's true .", "C : And then , so , the only reason way we would know what question he 's asking is based upon ", "C : Oh , so if Let 's say I had a construction parser , and I plug this in , I would know what each construction the communicative intent of the construction was", "A : Mm - hmm .", "C : and so then I would know how to weight the nodes appropriately , in response .", "C : So no matter what they said , if I could map it onto a Where - Is construction , I could say , \" ah !", "A : Ge", "A : Mm - hmm .", "C : well the the intent , here , was Where - Is \" ,", "A : OK , right .", "C : and I could look at those .", "A : Yeah .", "A : Yes , I mean .", "A : Sure .", "A : You do need to know I mean , to have that kind of information .", "B : Hmm .", "B : Yeah , I 'm also agreeing that a simple", "B : pru Take the ones where we have a clear winner .", "B : Forget about the ones where it 's all sort of middle ground .", "B : Prune those out", "B : and just hand over the ones where we have a winner .", "B : Yeah , because that would be the easiest way .", "B : We just compose as an output an XML mes message that says . \" Go there now . \" \" Enter historical information . \"", "B : And not care whether that 's consistent with anything .", "B : Right ?", "B : But in this case if we say , \" definitely he doesn't want to go there . He just wants to know where it is . \"", "B : or let 's call this this \" Look - At - H \"", "B : He wants to know something about the history of .", "B : So he said , \" Tell me something about the history of that . \"", "B : Now , the e But for some reason the Endpoint - Approach gets a really high score , too .", "B : We can't expect this to be sort of at O point three , three , three , O point , three , three , three , O point , three , three , three .", "B : Right ?", "B : Somebody needs to zap that .", "B : You know ?", "B : Or know There needs to be some knowledge that ", "C : We Yeah , but , the Bayes - net that would merge ", "C : I just realized that I had my hand in between my mouth and my micr er , my and my microphone .", "C : So then , the Bayes - net that would merge there , that would make the decision between Go - there , Info - on , and Location , would have a node to tell you which one of those three you wanted ,", "C : and based upon that node , then you would look at the other stuff .", "B : Yep .", "C : I mean , it i", "B : Yep .", "C : Does that make sense ?", "B : Yep . It 's sort of one of those , that 's It 's more like a decision tree , if if you want .", "B : You first look o at the lowball ones ,", "C : Yeah , i", "B : and then ", "C : Yeah , I didn't intend to say that every possible ", "C : OK .", "C : There was a confusion there ,", "C : k I didn't intend to say every possible thing should go into the Bayes - net ,", "C : because some of the things aren't relevant in the Bayes - net for a specific question .", "C : Like the Endpoint is not necessarily relevant in the Bayes - net for Where - Is until after you 've decided whether you wanna go there or not .", "B : Mm - hmm .", "A : Right .", "C : Show us the way , Bhaskara .", "A : I guess the other thing is that um , yeah .", "A : I mean , when you 're asked a specific question and you don't even ", "A : Like , if you 're asked a Where - Is question , you may not even look like , ask for the posterior probability of the , uh , EVA node ,", "A : right ?", "A : Cuz , that 's what I mean , in the Bayes - net you always ask for the posterior probability of a specific node .", "A : So , I mean ,", "A : you may not even bother to compute things you don't need .", "B : Um .", "B : Aren't we always computing all ?", "A : No .", "A : You can compute , uh , the posterior probability of one subset of the nodes , given some other nodes ,", "A : but totally ignore some other nodes , also .", "A : Basically , things you ignore get marginalized over .", "B : Yeah , but that 's that 's just shifting the problem .", "B : Then you would have to make a decision ,", "B : \" OK , if it 's a Where - Is question , which decision nodes do I query ? \"", "A : Yeah . So you have to make ", "A : Yeah .", "A : Yes .", "B : That 's un", "A : But I would think that 's what you want to do .", "A : Right ?", "B : Mmm .", "D : Well , eventually , you still have to pick out which ones you look at .", "B : Yeah .", "D : So it 's pretty much the same problem ,", "B : Yeah it 's it 's it 's apples and oranges .", "D : isn't it ?", "B : Nuh ?", "B : I mean , maybe it does make a difference in terms of performance , computational time .", "B : So either you always have it compute all the posterior possibilities for all the values for all nodes , and then prune the ones you think that are irrelevant ,", "A : Mm - hmm .", "A : Mmm .", "B : or you just make a p @ @ a priori estimate of what you think might be relevant and query those .", "A : Yeah .", "C : So basically , you 'd have a decision tree query , Go - there .", "C : If k if that 's false , query this one . If that 's true , query that one .", "C : And just basically do a binary search through the ?", "A : I don't know if it would necessarily be that , uh , complicated .", "A : But , uh ", "C : Well , in the case of Go - there , it would be .", "A : I mean , it w", "C : In the case ", "C : Cuz if you needed an If y If Go - there was true , you 'd wanna know what endpoint was .", "C : And if it was false , you 'd wanna d look at either Lo - Income Info - on or History .", "A : Yeah .", "A : That 's true , I guess .", "A : Yeah , so , in a way you would have that .", "C : Also , I 'm somewhat boggled by that Hugin software .", "A : OK , why 's that ?", "C : I can't figure out how to get the probabilities into it .", "C : Like , I 'd look at ", "A : Mm - hmm .", "C : It 's somewha It 's boggling me .", "A : OK .", "A : Alright .", "A : Well , hopefully it 's fixable .", "C : Ju", "C : Oh yeah , yeah . I d I just think I haven't figured out what the terms in Hugin mean , versus what Java Bayes terms are .", "A : It 's there 's a ", "A : OK .", "B : Um , by the way , are Do we know whether Jerry and Nancy are coming ?", "A : So we can figure this out .", "B : Or ?", "A : They should come when they 're done their stuff , basically , whenever that is .", "A : So .", "C : What d what do they need to do left ?", "A : Um ,", "A : I guess , Jerry needs to enter marks ,", "A : but I don't know if he 's gonna do that now or later .", "A : But , uh , if he 's gonna enter marks , it 's gonna take him awhile , I guess ,", "A : and he won't be here .", "C : And what 's Nancy doing ?", "A : Nancy ?", "A : Um , she was sorta finishing up the , uh , calculation of marks and assigning of grades ,", "A : but I don't know if she should be here .", "A : Well or , she should be free after that ,", "A : so ", "A : assuming she 's coming to this meeting .", "A : I don't know if she knows about it .", "C : She 's on the email list ,", "C : right ?", "A : Is she ?", "A : OK .", "B : Mm - hmm .", "B : OK .", "B : Because", "B : basically , what where we also have decided , prior to this meeting is that we would have a rerun of the three of us sitting together", "D : OK .", "B : sometime this week again", "A : OK .", "B : and finish up the , uh , values of this .", "B : So we have , uh Believe it or not , we have all the bottom ones here .", "C : Well , I ", "D : You added a bunch of nodes , for ?", "B : Yep .", "D : OK .", "B : We we we have Actually what we have is this line .", "B : Right ?", "C : Uh , what do the , uh , structures do ?", "B : Hmm ?", "C : So the the the For instance , this Location node 's got two inputs ,", "C : that one you ", "A : Four inputs .", "B : Hmm .", "B : Four .", "A : Those are The bottom things are inputs , also .", "C : Oh , I see .", "A : Yeah .", "C : OK , that was", "C : OK . That makes a lot more sense to me now .", "B : Yep .", "C : Cuz I thought it was like , that one in Stuart 's book about , you know , the ", "A : Alarm in the dog ?", "C : U Yeah .", "A : Yeah .", "C : Or the earthquake and the alarm .", "A : Sorry . Yeah , I 'm confusing two .", "C : Yeah , there 's a dog one , too ,", "C : but that 's in Java Bayes ,", "C : isn't it ?", "A : Right .", "A : Maybe .", "C : But there 's something about bowel problems or something with the dog .", "A : Yeah .", "B : And we have all the top ones ,", "B : all the ones to which no arrows are pointing .", "B : What we 're missing are the these ,", "B : where arrows are pointing ,", "B : where we 're combining top ones .", "B : So , we have to come up with values for this , and this , this , this , and so forth .", "B : And maybe just fiddle around with it a little bit more .", "B : And , um .", "B : And then it 's just , uh , edges ,", "B : many of", "B : edges .", "B : And , um , we won't meet next Monday .", "B : So .", "C : Cuz of Memorial Day ?", "B : Yep .", "A : We 'll meet next Tuesday , I guess .", "B : Yeah .", "C : When 's Jerry leaving for Italia ?", "B : On on Friday .", "A : Which Friday ?", "B : This this Friday .", "A : OK .", "D : Oh . This Friday ?", "C : Ugh .", "B : This Friday .", "C : As in , four days ?", "B : Yep .", "C : Or , three days ?", "A : Is he How long is he gone for ?", "B : Two weeks .", "A : Italy , huh ?", "A : What 's , uh what 's there ?", "B : Well , it 's a country .", "B : Buildings .", "B : People .", "C : But it 's not a conference or anything .", "A : Pasta .", "C : He 's just visiting .", "B : Hmm ?", "A : Right .", "A : Just visiting .", "B : Vacation .", "A : It 's a pretty nice place ,", "A : in my brief , uh , encounter with it .", "B : Do you guys ", "B : Oh , yeah . So . Part of what we actually want to do is sort of schedule out what we want to surprise him with when when he comes back .", "B : Um , so ", "C : Oh , I think we should disappoint him .", "B : Yeah ?", "B : You or have a finished construction parser and a working belief - net ,", "B : and uh ", "C : That wouldn't be disappointing .", "C : I think w we should do absolutely no work for the two weeks that he 's gone .", "B : Well , that 's actually what I had planned , personally .", "B : I had I I had sort of scheduled out in my mind that you guys do a lot of work , and I do nothing .", "B : And then , I sort of ", "C : Oh , yeah , that sounds good , too .", "B : sort of bask in in your glory .", "B : But , uh , i do you guys have any vacation plans ,", "B : because I myself am going to be , um , gone ,", "B : but this is actually not really important .", "B : Just this weekend we 're going camping .", "C : Yeah , I 'm wanna be this gone this weekend , too .", "B : Ah .", "B : But we 're all going to be here on Tuesday again ?", "B : Looks like it ?", "D : Yeah .", "B : OK , then . Let 's meet meet again next Tuesday .", "B : And , um , finish up this Bayes - net .", "B : And once we have finished it , I guess we can , um ", "B : and that 's going to be more just you and me ,", "B : because Bhaskara is doing probabilistic , recursive , structured , object - oriented , uh ,", "C : Killing machines !", "B : reasoning machines .", "A : Yes .", "B : And , um ", "C : Killing , reasoning . What 's the difference ?", "D : Wait . So you 're saying , next Tuesday , is it the whole group meeting ,", "D : or just us three working on it , or or ?", "B : Uh .", "B : The whole group .", "B : And we present our results ,", "B : our final ,", "D : OK .", "B : definite ", "D : So , when you were saying we need to do a re - run of , like ", "A : h What ?", "D : What Like , just working out the rest of the ", "B : Yeah . We should do this th the upcoming days .", "D : This week ?", "B : So , this week , yeah .", "D : OK .", "C : When you say , \" the whole group \" , you mean the four of us , and Keith ?", "B : And , Ami might .", "C : Ami might be here ,", "C : and it 's possible that Nancy 'll be here ?", "B : Yep .", "C : So ,", "C : yeah .", "B : Because , th you know ,", "B : once we have the belief - net done ", "C : You 're just gonna have to explain it to me , then , on Tuesday ,", "C : how it 's all gonna work out .", "C : You know .", "B : We will .", "B : OK . Because then , once we have it sort of up and running , then we can start you know , defining the interfaces", "B : and then feed stuff into it", "B : and get stuff out of it ,", "B : and then hook it up to some fake construction parser", "B : and ", "C : That you will have in about nine months or so .", "C : Yeah .", "B : Yeah .", "B : And , um ,", "C : The first bad version 'll be done in nine months .", "B : Yeah , I can worry about the ontology interface", "B : and you can Keith can worry about the discourse .", "B : I mean , this is pretty Um , I mean , I I I hope everybody uh knows that these are just going to be uh dummy values ,", "B : right ?", "A : Which ", "B : where the ", "A : Which ones ?", "B : S so so if the endpoint if the Go - there is Yes and No , then Go - there - discourse will just be fifty - fifty .", "B : Right ?", "A : Um , what do you mean ? If the Go - there says No , then the Go - there is ", "D : I don't get it .", "A : I don't u understand .", "B : Um .", "A : Like , the Go - there depends on all those four things .", "B : Yep .", "A : Yeah .", "B : But , what are the values of the Go - there - discourse ?", "A : Well , it depends on the situation .", "A : If the discourse is strongly indicating that ", "B : Yeah , but , uh , we have no discourse input .", "A : Oh , I see . The d See , uh , specifically in our situation , D and O are gonna be , uh ", "A : Yeah .", "A : Sure .", "A : So , whatever .", "D : So , so far we have ", "D : Is that what the Keith node is ?", "B : Yep .", "D : OK . And you 're taking it out ? for now ?", "D : Or ?", "B : Well , this is D ", "B : OK , this , I can I can get it in here .", "D : All the D 's are ", "B : I can get it in here ,", "B : so th We have the , uh , um ,", "B : sk let 's let 's call it \" Keith - Johno", "B : node \" .", "A : Johno ?", "B : There is an H somewhere printed .", "C : There you go .", "A : Yeah . People have the same problem with my name .", "B : Yeah .", "A : Oops .", "B : And , um ,", "C : Does th th does the H go b before the A or after the A ?", "A : Oh , in my name ?", "C : Yeah .", "A : Before the A .", "C : OK , good .", "C : Cuz you kn When you said people have the same problem , I thought ", "C : Cuz my H goes after the uh e e e the v", "A : People have the inverse problem with my name .", "C : OK .", "C : I always have to check , every time y I send you an email , a past email of yours , to make sure I 'm spelling your name correctly .", "A : Yeah .", "A : That 's good .", "C : I worry about you .", "A : I appreciate that .", "B : But , when you abbreviate yourself as the \" Basman \" , you don't use any H 's .", "A : \" Basman \" ?", "A : Yeah , it 's because of the chessplayer named Michael Basman ,", "A : who is my hero .", "B : OK .", "C : You 're a geek .", "C : It 's O K .", "C : I", "C : How do you pronou", "B : OK .", "C : How do you pronounce your name ?", "D : Eva .", "C : Eva ?", "D : Yeah .", "A : Not Eva ?", "C : What if I were What if I were to call you Eva ?", "D : I 'd probably still respond to it .", "D : I 've had people call me Eva ,", "D : but I don't know .", "C : No , not just Eva , Eva .", "C : Like if I u take the V and s pronounce it like it was a German V ?", "B : Which is F .", "C : Yeah .", "D : Um , no idea then .", "B : Voiced .", "D : What ?", "C : It sounds like an F .", "D : I ", "C : There 's also an F in German ,", "D : OK .", "C : which is why I ", "B : Well , it 's just the difference between voiced and unvoiced .", "C : Yeah .", "D : OK .", "C : As long as that 's O K .", "D : Um .", "C : I mean , I might slip out and say it accidentally .", "C : That 's all I 'm saying .", "D : That 's fine .", "A : Yeah . It doesn't matter what those nodes are , anyway ,", "A : because we 'll just make the weights \" zero \" for now .", "B : Yep .", "B : We 'll make them zero for now ,", "B : because it who who knows what they come up with ,", "B : what 's gonna come in there .", "B : OK .", "B : And , um ,", "B : then", "B : should we start on Thursday ?", "A : OK .", "B : And not meet tomorrow ?", "A : Sure .", "B : OK .", "B : I 'll send an email ,", "B : make a time suggestion .", "C : Wait ,", "C : maybe it 's OK , so that that that we can that we have one node per construction .", "C : Cuz even in people , like , they don't know what you 're talking about if you 're using some sort of strange construction .", "B : Yeah , they would still c sort of get the closest , best fit .", "C : Well , yeah , but I mean , the uh , I mean , that 's what the construction parser would do .", "C : Uh , I mean , if you said something completely arbitrary , it would f find the closest construction ,", "B : Mm - hmm .", "B : OK .", "C : right ?", "C : But if you said something that was completel er ", "C : h theoretically the construction parser would do that ", "C : But if you said something for which there was no construction whatsoever , n people wouldn't have any idea what you were talking about .", "B : Mm - hmm .", "C : Like \" Bus dog fried egg . \" I mean .", "C : You know .", "B : Or , if even something Chinese , for example .", "C : Or , something in Mandarin , yeah .", "C : Or Cantonese , as the case may be .", "C : What do you think about that , Bhaskara ?", "A : I mean ", "A : Well ", "A : But how many constructions do could we possibly have nodes for ?", "C : In this system , or in r", "A : No , we .", "A : Like , when people do this kind of thing .", "C : Oh , when p How many constructions do people have ?", "A : Yeah .", "C : I have not the slightest idea .", "A : Is it considered to be like in are they considered to be like very , uh , sort of s abstract things ?", "C : Every noun is a construction .", "A : OK , so it 's like in the thousands .", "C : The ", "C : Yeah .", "C : Any any form - meaning pair , to my understanding , is a construction .", "A : OK .", "B : So .", "C : And form u starts at the level of noun Or actually , maybe even sounds .", "B : Phoneme .", "B : Yep .", "C : Yeah .", "C : And goes upwards", "C : until you get the ditransitive construction .", "C : And then , of course , the c I guess , maybe there can be the ", "A : S", "C : Can there be combinations of the dit", "A : Discourse - level constructions .", "C : Yeah .", "C : The \" giving a speech \" construction ,", "B : Rhetorical constructions .", "B : Yeah .", "A : Yes .", "B : But , I mean , you know , you can probably count count the ways .", "B : I mean .", "C : It 's probab Yeah , I would s definitely say it 's finite .", "B : Yeah .", "C : And at least in compilers , that 's all that really matters ,", "C : as long as your analysis is finite .", "A : How 's that ? How it can be finite , again ?", "C : Nah , I can't think of a way it would be infinite .", "B : Well , you can come up with new constructions .", "C : Yeah . If the if your if your brain was totally non - deterministic , then perhaps there 's a way to get , uh , infin an infinite number of constructions that you 'd have to worry about .", "A : But , I mean , in the practical sense , it 's impossible .", "C : Right . Cuz if we have a fixed number of neurons ?", "A : Yeah .", "C : So the best - case scenario would be the number of constructions ", "C : or , the worst - case scenario is the number of constructions equals the number of neurons .", "A : Well , two to the power of the number of neurons .", "C : Right .", "C : But still finite .", "B : OK .", "C : No , wait .", "C : Not necessarily , is it ?", "C : We can end the meeting .", "C : I just ", "C : Can't you use different var different levels of activation ?", "C : across , uh lots of different neurons , to specify different values ?", "B : Mm - hmm .", "A : Um , yeah , but there 's , like , a certain level of ", "C : There 's a bandwidth issue ,", "C : right ?", "A : Bandw - Yeah , so you can't do better than something .", "C : Yeah .", "B : Turn off the mikes .", "B : Otherwise it gets really tough for the tr" ], "abstractive summary": [ { "id": "Bed012.s.1", "text": "The focus of the meeting was on a presentation of the work done already on the building of the Bayes-net.", "type": "abstract" }, { "id": "Bed012.s.2", "text": "The input layer deriving information from things like the user and situation models, feeds into a set of decision nodes, such as the Enter/View/Approach (EVA) endpoint.", "type": "abstract" }, { "id": "Bed012.s.3", "text": "In any particular situation, most of the outputs will not be relevant to the given context.", "type": "abstract" }, { "id": "Bed012.s.4", "text": "Therefore, they will either have to be pruned a posteriori, or only a subset of the possible decision nodes will be computed in each occasion.", "type": "abstract" }, { "id": "Bed012.s.5", "text": "The latter option could could follow a binary search-tree approach and it could also be better in computational terms.", "type": "abstract" }, { "id": "Bed012.s.6", "text": "In any case, on what basis the \"winner\" output is chosen is not clear.", "type": "abstract" }, { "id": "Bed012.s.7", "text": "One suggestion was discussed: the particular constructions used can determine the pertinent decision (output) nodes.", "type": "abstract" }, { "id": "Bed012.s.8", "text": "The complete prototype of the Bayes-net will be presented in the next meeting.", "type": "abstract" }, { "id": "Bed012.s.9", "text": "After that, it will be possible to define interfaces and a dummy construction parser, in order to test and link modules together.", "type": "abstract" }, { "id": "Bed012.s.10", "text": "The suggestion that the most appropriate decision node of the belief-net in each situation could be chosen as a function of what construction was used, was deemed unsuitable at this stage.", "type": "decisions" }, { "id": "Bed012.s.11", "text": "There are many interdependencies between the output nodes that this approach could not take into account.", "type": "decisions" }, { "id": "Bed012.s.12", "text": "The rest of the values for the Bayes-net nodes will be built in within the week.", "type": "decisions" }, { "id": "Bed012.s.13", "text": "The finished prototype will be presented during the next meeting.", "type": "decisions" }, { "id": "Bed012.s.14", "text": "Any set of inputs will provide either the whole range of output values of the Bayes-net or an a priori selection of those outputs.", "type": "problems" }, { "id": "Bed012.s.15", "text": "In both cases, what is needed is a way to single out the appropriate outputs for any given context.", "type": "problems" }, { "id": "Bed012.s.16", "text": "For example, in the case of a \"where is\" question, whether the prevalent output should be \"Go-there\" or \"Info-on\" or even a third option has to be computed somehow.", "type": "problems" }, { "id": "Bed012.s.17", "text": "In any case, there are many input values that have not been entered in the Bayes-net at this stage.", "type": "problems" }, { "id": "Bed012.s.18", "text": "Furthermore, no inputs for the Ontology and Discourse can be built in yet, as they involve research that will be carried out in the future.", "type": "problems" }, { "id": "Bed012.s.19", "text": "Finally, there has been a problem with adding probabilities in a net created with the Hugin software, but this should be overcome very shortly.", "type": "problems" }, { "id": "Bed012.s.20", "text": "The presented Bayes-net takes inputs from the Situation, User, Discourse and Ontology models.", "type": "progress" }, { "id": "Bed012.s.21", "text": "There are several values (elements) defined in each of these models.", "type": "progress" }, { "id": "Bed012.s.22", "text": "The inputs are fed into the belief-net, which, in turn, outputs the posterior probabilities for the values of all the decision nodes.", "type": "progress" }, { "id": "Bed012.s.23", "text": "These comprise \"Go-there\", \"EVA\", \"Info-on\", \"Location\", \"Timing\", etc.", "type": "progress" }, { "id": "Bed012.s.24", "text": "At this stage, all the decision nodes are evenly weighted: regardless of the context, each output is trusted equally.", "type": "progress" }, { "id": "Bed012.s.25", "text": "Input and output node structure was presented in XML, as this is the format that will be used for the system.", "type": "progress" }, { "id": "Bed012.s.26", "text": "A large number of the value probabilities have already been set.", "type": "progress" } ], "extractive summary": [ "B : What 's the situation like at the entity that is mentioned ?", "B : if we know anything about it ?", "B : That 's basically just specifying the the input for the w what 's", "C : y Why are you specifying it in XML ?", "B : And , also , I mean , this is a what the input is going to be .", "B : I mean , once Once you have this sort of as running as a module ", "B : What you want is You wanna say , \" OK , give me the posterior probabilities of the Go - there node , when this is happening . \"", "C : So this is just a specification of all the possible inputs ?", "B : Yep .", "B : As , if I understand it correctly , it always gives you all the posterior probabilities for all the values of all decision nodes .", "B : and the question is what to do with it ,", "B : how to interpret it ,", "B : So the person said , um , \" Where is X ? \"", "B : we want to know , um , is Does he want info ?", "B : or know the location ?", "B : Or does he want to go there ?", "B : I mean . It 's always gonna give us a value of how likely we think i it is that he wants to go there and doesn't want to go there ,", "B : or how likely it is that he wants to get information .", "B : So , it n I mean a All I 'm saying is , whatever our input is , we 're always gonna get the full output .", "B : And some some things will always be sort of too not significant enough .", "C : You won't it 'll be hard to decide .", "C : you could say , \" Well , there Here 's the Where - Is construction . \"", "C : And for the Where - Is construction , we know we need to l look at this node , that merges these three things together", "C : I guess they are evenly weighted .", "A : The different decision nodes , you mean ?", "A : If we trusted the Go - there node more th much more than we trusted the other ones , then we would conclude , even in this situation , that he wanted to go there .", "B : What what what I am thinking , or what we 're about to propose here is we 're always gonna get the whole list of values and their posterior probabilities .", "B : And now we need an expert system or belief - net or something that interprets that ,", "C : Yeah I know . But how do we weight what we get out ?", "C : As , which one i Which ones are important ?", "C : So my i So , if we were to it with a Bayes - net , we 'd have to have a node for every question that we knew how to deal with ,", "C : that would take all of the inputs and weight them appropriately for that question .", "C : Are we going to make a node for every question ?", "C : Every construction .", "A : I mean , in general , like , we won't just have those three ,", "A : We 'll have , uh , like , many , many nodes .", "A : So we have to , like So that it 's no longer possible to just look at the nodes themselves and figure out what the person is trying to say .", "B : Because there are interdependencies ,", "A : I mean , it 's not based on constructions ,", "A : it 's based on things like , uh , there 's gonna be a node for Go - there or not , and there 's gonna be a node for Enter , View , Approach .", "C : Oh , so if Let 's say I had a construction parser , and I plug this in , I would know what each construction the communicative intent of the construction was", "C : and so then I would know how to weight the nodes appropriately , in response .", "C : So then , the Bayes - net that would merge there , that would make the decision between Go - there , Info - on , and Location , would have a node to tell you which one of those three you wanted ,", "A : Like , if you 're asked a Where - Is question , you may not even look like , ask for the posterior probability of the , uh , EVA node ,", "A : Cuz , that 's what I mean , in the Bayes - net you always ask for the posterior probability of a specific node .", "A : you may not even bother to compute things you don't need .", "B : Yeah , but that 's that 's just shifting the problem .", "B : Then you would have to make a decision ,", "B : \" OK , if it 's a Where - Is question , which decision nodes do I query ? \"", "D : Well , eventually , you still have to pick out which ones you look at .", "B : I mean , maybe it does make a difference in terms of performance , computational time .", "C : So basically , you 'd have a decision tree query , Go - there .", "C : If k if that 's false , query this one . If that 's true , query that one .", "C : And just basically do a binary search through the ?", "C : Also , I 'm somewhat boggled by that Hugin software .", "C : I can't figure out how to get the probabilities into it .", "B : basically , what where we also have decided , prior to this meeting is that we would have a rerun of the three of us sitting together", "B : sometime this week again", "B : and finish up the , uh , values of this .", "B : So we have , uh Believe it or not , we have all the bottom ones here .", "B : So , we have to come up with values for this , and this , this , this , and so forth .", "B : And maybe just fiddle around with it a little bit more .", "B : And , um , we won't meet next Monday .", "A : We 'll meet next Tuesday , I guess .", "B : OK , then . Let 's meet meet again next Tuesday .", "B : And , um , finish up this Bayes - net .", "B : And we present our results ,", "B : OK . Because then , once we have it sort of up and running , then we can start you know , defining the interfaces", "B : and then feed stuff into it", "B : and get stuff out of it ,", "B : and then hook it up to some fake construction parser", "B : Yeah , I can worry about the ontology interface", "B : and you can Keith can worry about the discourse .", "C : maybe it 's OK , so that that that we can that we have one node per construction .", "C : Cuz even in people , like , they don't know what you 're talking about if you 're using some sort of strange construction .", "B : Yeah , they would still c sort of get the closest , best fit .", "A : But how many constructions do could we possibly have nodes for ?", "C : Any any form - meaning pair , to my understanding , is a construction .", "B : But , I mean , you know , you can probably count count the ways .", "C : as long as your analysis is finite ." ] }, "log": [] }, "ICSI--train--2": { "original dialog id": "Bro013.json", "dialog index": 2, "original dialog info": { "dialog history": [ "A : We 're going ?", "A : OK .", "B : OK .", "A : Sh - Close your door on door on the way out ?", "B : Thanks .", "A : Thanks .", "B : Oh .", "A : Yeah .", "A : Probably wanna get this other door , too .", "A : OK .", "A : So .", "A : Um . What are we talking about today ?", "E : Uh , well , first there are perhaps these uh Meeting Recorder digits that we tested .", "A : Oh , yeah . That was kind of uh interesting .", "E : So .", "A : The both the uh the SRI System and the oth", "E : Um .", "A : And for one thing that that sure shows the difference between having a lot of uh training data or not ,", "E : Of data ?", "E : Yeah .", "A : uh , the uh The best kind of number we have on the English uh on near microphone only is is uh three or four percent .", "E : Mm - hmm .", "A : And uh it 's significantly better than that , using fairly simple front - ends on on the uh uh , with the SRI system .", "E : Mm - hmm .", "A : So I th I think that the", "A : uh ", "A : But that 's that 's using uh a a pretty huge amount of data ,", "A : mostly not digits , of course ,", "A : but but then again ", "A : Well , yeah .", "A : In fact , mostly not digits for the actual training the H M Ms whereas uh in this case we 're just using digits for training the H M", "E : Yeah .", "E : Right .", "A : Did anybody mention about whether the the SRI system is a is is doing the digits um the wor as a word model or as uh a sub s sub - phone states ?", "E : I guess it 's it 's uh allophone models ,", "A : Yeah .", "E : so , well ", "A : Probably .", "A : Huh ?", "E : Yeah . I think so ,", "E : because it 's their very d huge , their huge system .", "A : Yeah .", "E : And .", "E : But .", "E : So . There is one difference ", "E : Well , the SRI system the result for the SRI system that are represented here are with adaptation .", "E : So there is ", "E : It 's their complete system and including on - line uh unsupervised adaptation .", "A : That 's true .", "E : And if you don't use adaptation , the error rate is around fifty percent worse , I think , if I remember .", "E : Yeah .", "A : OK .", "A : It 's tha it 's that much , huh ?", "E : Nnn .", "E : It 's ", "E : Yeah .", "E : It 's quite significant .", "E : Yeah .", "A : Oh .", "A : OK .", "A : Still .", "E : Mm - hmm .", "A : But but uh what what I think I 'd be interested to do given that , is that we we should uh take ", "A : I guess that somebody 's gonna do this ,", "A : right ? ", "A : is to take some of these tandem things and feed it into the SRI system ,", "A : right ?", "E : Yeah .", "A : Yeah .", "E : We can do something like that .", "E : Yeah .", "A : Yeah . Because ", "E : But ", "E : But I guess the main point is the data because", "E : uh I am not sure .", "E : Our back - end is is fairly simple", "E : but until now , well , the attempts to improve it or have fail", "E : Ah , well , I mean uh what Chuck tried to to to do", "A : Yeah ,", "A : but he 's doing it with the same data , right ?", "E : Yeah .", "A : I mean so to So there 's there 's there 's two things being affected .", "E : So it 's ", "E : Yeah .", "A : I mean . One is that that , you know , there 's something simple that 's wrong with the back - end .", "A : We 've been playing a number of states", "E : Mm - hmm .", "A : uh I I don't know if he got to the point of playing with the uh number of Gaussians yet", "E : Mm - hmm .", "A : but but uh ,", "A : uh , you know .", "A : But , yeah , so far he hadn't gotten any big improvement ,", "E : Mm - hmm .", "A : but that 's all with the same amount of data which is pretty small .", "E : Yeah .", "E : Mmm .", "A : And um .", "E : So , yeah , we could retrain some of these tandem on on huge ", "A : Well , you could do that , but I 'm saying even with it not with that part not retrained ,", "E : Ah , yeah .", "A : just just using having the H M Ms much better H M", "E : Just ", "E : f for the HMM models .", "E : Yeah .", "A : Yeah .", "E : Mm - hmm .", "E : Mm - hmm .", "A : Um . But just train those H M Ms using different features ,", "A : the features coming from our Aurora stuff .", "E : Yeah .", "A : So .", "E : Yeah .", "E : But what would be interesting to see also is what what ", "E : perhaps it 's not related , the amount of data but the um recording conditions .", "E : I don't know .", "E : Because it 's probably not a problem of noise , because our features are supposed to be robust to noise .", "E : It 's not a problem of channel , because there is um normalization with respect to the channel .", "A : Well , yeah .", "E : So ", "A : I I I 'm sorry .", "A : What what is the problem that you 're trying to explain ?", "E : The the fact that the result with the tandem and Aurora system are uh so much worse .", "A : That the ", "A : Oh .", "A : So much worse ?", "E : Yeah .", "A : Oh .", "A : I uh but I 'm I 'm almost certain that it it I mean , that it has to do with the um amount of training data .", "E : It ", "A : It it 's it 's orders of magnitude off .", "E : Yeah but ", "E : Yeah .", "E : Yeah but we train only on digits and it 's it 's a digit task ,", "E : so .", "E : Well .", "A : But but having a huge ", "E : It ", "A : If if you look at what commercial places do , they use a huge amount of data .", "E : Mm - hmm .", "E : Alright .", "A : This is a modest amount of data .", "E : Yeah .", "E : Mm - hmm .", "A : So . I mean , ordinarily you would say \" well , given that you have enough occurrences of the digits , you can just train with digits rather than with , you know \" ", "E : Mm - hmm .", "A : But the thing is , if you have a huge ", "A : in other words , do word models ", "A : But if you have a huge amount of data then you 're going to have many occurrences of similar uh allophones .", "E : Right .", "E : Mmm .", "E : Yeah .", "A : And that 's just a huge amount of training for it .", "A : So it 's um I I think it has to be that , because , as you say , this is , you know , this is near - microphone ,", "E : Mm - hmm .", "A : it 's really pretty clean data .", "E : Mm - hmm .", "A : Um .", "A : Now , some of it could be the fact that", "A : uh ", "A : let 's see , in the in these multi - train things did we include noisy data in the training ?", "E : Yeah .", "A : I mean , that could be hurting us actually , for the clean case .", "E : Yeah .", "E : Well , actually we see that the clean train for the Aurora proposals are are better than the multi - train ,", "A : It is if ", "E : yeah .", "A : Yeah .", "A : Yeah .", "A : Cuz this is clean data ,", "A : and so that 's not too surprising .", "E : Mm - hmm .", "A : But um . Uh .", "A : So .", "E : Well , o I guess what I meant is that", "E : well , let 's say if we if we add enough data to train on the um on the Meeting Recorder digits , I guess we could have better results than this .", "A : Uh - huh .", "A : Mm - hmm .", "E : And .", "E : What I meant is that perhaps we can learn something uh from this ,", "E : what 's what 's wrong uh what what is different between TI - digits and these digits", "E : and ", "A : What kind of numbers are we getting on TI - digits ?", "E : It 's point eight percent ,", "E : so .", "A : Oh .", "A : I see .", "E : Four - Fourier .", "A : So in the actual TI - digits database we 're getting point eight percent ,", "E : Yeah .", "E : Yeah .", "A : and here we 're getting three or four three ,", "E : Mm - hmm .", "A : let 's see , three for this ?", "A : Yeah .", "A : Sure , but I mean ,", "A : um point eight percent is something like double uh or triple what people have gotten who 've worked very hard at doing that .", "E : Mm - hmm .", "A : And and also , as you point out , there 's adaptation in these numbers also .", "E : Mmm .", "A : So if you , you know , put the ad adap take the adaptation off , then it for the English - Near you get something like two percent .", "A : And here you had , you know , something like three point four .", "E : Mm - hmm .", "A : And I could easily see that difference coming from this huge amount of data that it was trained on .", "E : Mm - hmm .", "A : So it 's ", "A : You know , I don't think there 's anything magical here .", "E : Yeah .", "A : It 's , you know , we used a simple HTK system with a modest amount of data .", "A : And this is a a , you know , modern uh system", "E : Yeah .", "A : uh has has a lot of nice points to it .", "E : Mm - hmm .", "A : Um .", "A : So . I mean , the HTK is an older HTK , even .", "A : So .", "E : Mm - hmm .", "A : Yeah it it 's not that surprising .", "A : But to me it just it just meant a practical point that um if we want to publish results on digits that that people pay attention to we probably should uh ", "A : Cuz we 've had the problem before that you get show some nice improvement on something that 's that 's uh , uh it seems like too large a number ,", "E : Mm - hmm .", "A : and uh uh people don't necessarily take it so seriously .", "A : Um .", "A : Yeah .", "A : Yeah .", "A : So the three point four percent for this uh is is", "A : uh ", "A : So why is it ", "A : It 's an interesting question though , still .", "A : Why is why is it three point four percent for the d the digits recorded in this environment as opposed to the uh point eight percent for for for the original TI - digits database ?", "A : Um .", "E : Yeah .", "E : th that 's th that 's my point", "A : Given given the same ", "E : I I I don't I ", "A : Yeah . So ignore ignoring the the the SRI system for a moment ,", "E : Mm - hmm .", "A : just looking at the TI - di the uh tandem system ,", "A : if we 're getting point eight percent , which , yes , it 's high .", "A : It 's , you know , it it 's not awfully high ,", "E : Mm - hmm .", "A : but it 's , you know it 's it 's high .", "A : Um . Why is it uh four times as high , or more ?", "E : Yeah , I guess .", "A : Right ?", "A : I mean , there 's even though it 's close - miked there 's still there really is background noise .", "E : Mm - hmm .", "A : Um . And uh I suspect when the TI - digits were recorded if somebody fumbled or said something wrong or something that they probably made them take it over .", "E : Mm - hmm .", "A : It was not I mean there was no attempt to have it be realistic in any in any sense at all .", "E : Well .", "E : Yeah .", "E : And acoustically , it 's q it 's ", "E : I listened .", "E : It 's quite different .", "E : TI - digit is it 's very , very clean and it 's like studio recording", "A : Mm - hmm .", "E : whereas these Meeting Recorder digits sometimes you have breath noise", "E : and", "E : Mmm .", "A : Right .", "A : Yeah .", "A : So I think they were ", "E : It 's not controlled at all , I mean .", "A : Bless you .", "B : Thanks .", "A : I Yeah . I think it 's it 's ", "E : Mm - hmm .", "A : So . Yes .", "E : But", "A : It 's I think it 's it 's the indication it 's harder .", "E : Yeah .", "A : Uh . Yeah and again , you know , i that 's true either way .", "A : I mean so take a look at the uh um , the SRI results .", "A : I mean , they 're much much better ,", "A : but still you 're getting something like one point three percent for uh things that are same data as in T TI - digits the same same text .", "E : Mm - hmm .", "A : Uh . And uh , I 'm sure the same same system would would get , you know , point point three or point four or something on the actual TI - digits .", "A : So this I think , on both systems the these digits are showing up as harder .", "E : Mmm .", "A : Um .", "E : Mm - hmm .", "A : Which I find sort of interesting cause I think this is closer to ", "A : uh I mean it 's still read .", "A : But I still think it 's much closer to to what what people actually face , um when they 're they 're dealing with people saying digits over the telephone .", "A : I mean . I don't think uh ", "A : I mean , I 'm sure they wouldn't release the numbers ,", "A : but I don't think that uh the uh the the companies that that do telephone speech get anything like point four percent on their digits .", "A : I 'm I 'm I 'm sure they get ", "A : Uh , I mean , for one thing people do phone up who don't have uh uh Middle America accents and", "E : Mm - hmm .", "A : it 's a we we it 's it 's it 's US .", "A : it has has many people who sound in many different ways .", "A : So .", "A : Um .", "A : I mean .", "A : OK .", "A : That was that topic .", "A : What else we got ?", "E : Um .", "A : Did we end up giving up on on , any Eurospeech submissions ,", "E : But ", "A : or ?", "A : I know Thilo and Dan Ellis are are submitting something ,", "A : but uh .", "E : Yeah .", "E : I I guess e the only thing with these the Meeting Recorder and , well , ", "E : So , I think , yeah I think we basically gave up .", "A : Um . Now , actually for the for the Aur - uh", "E : But ", "A : we do have stuff for Aurora , right ?", "A : Because because we have ano an extra month or something .", "E : Yeah .", "E : Yeah .", "E : Yeah .", "E : So .", "E : Yeah , for sure we will do something for the special session .", "A : Yeah .", "A : Well , that 's fine .", "A : So th so so we have a couple a couple little things on Meeting Recorder", "E : Yeah .", "E : Mm - hmm .", "A : and we have We don't we don't have to flood it with papers .", "A : We 're not trying to prove anything to anybody .", "A : so .", "A : That 's fine .", "A : Um .", "A : Anything else ?", "E : Yeah .", "E : Well . So .", "E : Perhaps the point is that we 've been working on is ,", "E : yeah , we have put the um the good VAD in the system", "E : and it really makes a huge difference .", "E : Um .", "E : So ,", "E : yeah .", "E : I think , yeah , this is perhaps one of the reason why our system was not not the best ,", "E : because with the new VAD , it 's very the results are similar to the France Telecom results and perhaps even better sometimes .", "A : Hmm .", "B : Huh .", "E : Um . So there is this point .", "E : Uh . The problem is that it 's very big and we still have to think how to where to put it", "E : and um ,", "A : Mm - hmm .", "E : because it it ", "E : well , this VAD", "E : uh either some delay", "E : and we if we put it on the server side , it doesn't work ,", "E : because on the server side features you already have LDA applied from the f from the terminal side", "E : and so you accumulate the delay", "E : so the VAD should be before the LDA", "E : which means perhaps on the terminal side", "E : and then smaller and", "A : So wha where did this good VAD come from ?", "E : So .", "E : It 's um from OGI .", "E : So it 's the network trained it 's the network with the huge amounts on hidden of hidden units ,", "E : and um nine input frames compared to the VAD that was in the proposal which has a very small amount of hidden units and fewer inputs .", "A : This is the one they had originally ?", "E : Yeah .", "A : Oh .", "A : Yeah , but they had to get rid of it because of the space ,", "A : didn't they ?", "E : Yeah .", "E : So .", "E : Yeah .", "E : But the abso assumption is that we will be able to make a VAD that 's small and that works fine .", "E : And . So we can ", "A : Well . So that 's a problem .", "A : Yeah .", "E : Yeah but ", "E : nnn .", "A : But the other thing is uh to use a different VAD entirely .", "A : I mean , uh i if if there 's a if if ", "A : I I don't know what the thinking was amongst the the the the ETSI folk", "A : but um if everybody agreed sure let 's use this VAD and take that out of there ", "E : Mm - hmm .", "E : Mm - hmm .", "E : They just want , apparently they don't want to fix the VAD because they think there is some interaction between feature extraction and and VAD or frame dropping", "E : But they still want to just to give some um requirement for this VAD", "E : because it 's it will not be part of they don't want it to be part of the standard .", "A : OK .", "E : So . So it must be at least uh somewhat fixed but not completely .", "E : So there just will be some requirements that are still not uh not yet uh ready I think .", "A : Determined .", "A : I see .", "E : Nnn .", "A : But I was thinking that that uh s \" Sure , there may be some interaction ,", "A : but I don't think we need to be stuck on using our or OGI 's VAD .", "A : We could use somebody else 's if it 's smaller", "A : or ", "E : Yeah .", "A : You know , as long as it did the job .", "E : Mm - hmm .", "A : So that 's good .", "E : Uh . So there is this thing .", "E : There is", "E : um ", "E : Yeah .", "E : Uh I designed a new a new filter", "E : because when I designed other filters with shorter delay from the LDA filters , there was one filter with fif sixty millisecond delay and the other with ten milliseconds", "E : and uh Hynek suggested that both could have sixty - five sixty - s", "A : Right .", "E : I think it 's sixty - five .", "E : Yeah .", "A : Yeah .", "E : Both should have sixty - five because ", "E : Yeah .", "A : You didn't gain anything , right ?", "E : And . So I did that", "E : and uh it 's running .", "E : So , let 's see what will happen .", "E : Uh but the filter is of course closer to the reference filter .", "A : Mm - hmm .", "E : Mmm .", "E : Um . Yeah .", "E : I think ", "A : So that means logically , in principle , it should be better .", "A : So probably it 'll be worse .", "E : Yeah", "A : Or in the basic perverse nature uh of reality .", "A : Yeah .", "A : OK .", "E : Yeah .", "E : Sure .", "C : Yeah .", "A : OK .", "E : Yeah , and then we 've started to work with this of um voiced - unvoiced stuff .", "A : Mm - hmm .", "E : And next week I think we will perhaps try to have um a new system with uh uh MSG stream also", "E : see what what happens .", "E : So , something that 's similar to the proposal too , but with MSG stream .", "A : Mm - hmm .", "A : Mm - hmm .", "E : Mmm .", "A : OK .", "D : No , I w I begin to play with Matlab and to found some parameter robust for voiced - unvoiced decision .", "D : But only to play .", "D : And we they we found that maybe w is a classical parameter , the sq the variance between the um FFT of the signal and the small spectrum of time we after the um mel filter bank .", "A : Uh - huh .", "D : And , well , is more or less robust .", "D : Is good for clean speech .", "D : Is quite good for noisy speech .", "A : Huh ?", "A : Mm - hmm .", "D : but um we must to have bigger statistic with TIMIT ,", "A : Mm - hmm .", "D : and is not ready yet to use on ,", "A : Yeah .", "D : well , I don't know .", "A : Yeah .", "E : Yeah .", "E : So , basically we wa want to look at something like the ex the ex excitation signal and ", "A : Right .", "D : Mm - hmm .", "E : which are the variance of it and ", "D : I have here .", "D : I have here for one signal , for one frame .", "E : Mmm .", "A : Yeah .", "D : The the mix of the two , noise and unnoise , and the signal is this .", "A : Uh - huh .", "D : Clean , and this noise .", "A : Uh .", "D : These are the two ", "D : the mixed , the big signal is for clean .", "A : Well , I 'm s uh ", "A : There 's ", "A : None of these axes are labeled ,", "A : so I don't know what this ", "A : What 's this axis ?", "D : Uh this is uh this axis is nnn , \" frame \" .", "A : Frame .", "D : Mm - hmm .", "A : And what 's th what this ?", "D : Uh , this is uh energy , log - energy of the spectrum .", "D : Of the", "D : this is the variance ,", "D : the difference between the spectrum of the signal and FFT of each frame of the signal", "D : and this mouth spectrum of time after the f may fit for the two ,", "A : For this one .", "A : For the noi", "D : this big , to here , they are to signal .", "D : This is for clean and this is for noise .", "A : Oh .", "A : There 's two things on the same graph .", "D : Yeah .", "D : I don't know .", "D : I I think that I have d another graph ,", "D : but I 'm not sure .", "E : Yeah .", "A : So w which is clean and which is noise ?", "E : I think the lower one is noise .", "D : The lower is noise and the height is clean .", "A : OK .", "A : So it 's harder to distinguish", "D : It 's height .", "A : but it but it g", "E : Yeah .", "A : with noise of course", "A : but but ", "D : Oh . I must to have .", "A : Uh .", "D : Pity , but I don't have two different", "A : And presumably when there 's a a ", "E : So this should the the the t voiced portions .", "A : Uh - huh .", "D : Yeah , it is the height is voiced portion .", "E : The p the peaks should be voiced portion .", "D : And this is the noise portion .", "A : Uh - huh .", "D : And this is more or less like this .", "D : But I meant to have see @ @ two two the picture .", "A : Yeah .", "A : Yeah .", "D : This is , for example , for one frame .", "A : Yeah", "D : the the spectrum of the signal .", "D : And this is the small version of the spectrum after ML mel filter bank .", "A : Yeah .", "A : And this is the difference ?", "D : And this is", "D : I don't know .", "D : This is not the different .", "D : This is trying to obtain with LPC model the spectrum", "D : but using Matlab without going factor and s", "A : No pre - emphasis ?", "A : Yeah .", "D : Not pre - emphasis .", "D : Nothing .", "A : Yeah so it 's ", "D : And the ", "A : doesn't do too well there .", "D : I think that this is good .", "D : This is quite similar .", "D : this is this is another frame .", "D : ho how I obtained the envelope , this envelope , with the mel filter bank .", "A : Right .", "A : So now I wonder ", "A : I mean , do you want to ", "A : I know you want to get at something orthogonal from what you get with the smooth spectrum", "A : Um . But if you were to really try and get a voiced - unvoiced , do you do you want to totally ignore that ?", "A : I mean , do you do you ", "A : I mean , clearly a a very big very big cues for voiced - unvoiced come from uh spectral slope and so on ,", "E : Mm - hmm .", "A : right ?", "A : Um .", "E : Yeah .", "E : Well , this would be this would be perhaps an additional parameter ,", "E : simply isn't ", "A : Yeah .", "A : I see .", "E : Yeah .", "D : Yeah because when did noise clear in these section is clear", "E : Uh .", "A : Mm - hmm .", "D : if s @ @ val value is indicative that is a voice frame and it 's low values", "A : Yeah .", "A : Yeah .", "A : Well , you probably want ", "A : I mean , certainly if you want to do good voiced - unvoiced detection , you need a few features .", "A : Each each feature is by itself not enough .", "A : But , you know , people look at at slope and uh first auto - correlation coefficient , divided by power .", "E : Mmm .", "A : Or or uh", "A : um", "A : there 's uh ", "A : I guess we prob probably don't have enough computation to do a simple pitch detector or something ?", "A : I mean with a pitch detector you could have a have a an estimate of of what the ", "E : Mmm .", "A : Uh .", "A : Or maybe you could you just do it going through the P FFT 's figuring out some um probable um harmonic structure .", "A : Right .", "E : Mmm .", "A : And and uh .", "D : you have read up and you have a paper , the paper that you s give me yesterday .", "E : Oh , yeah .", "D : they say that yesterday they are some problem", "E : But ", "E : Yeah , but it 's not ", "E : it 's ,", "E : yeah , it 's it 's another problem .", "D : and the ", "E : Yeah", "D : Is another problem .", "E : Um .", "E : Yeah , there is th this fact actually .", "E : If you look at this um spectrum ,", "A : Yeah .", "E : What 's this again ?", "E : Is it the mel - filters ?", "D : Yeah like this .", "E : Yeah .", "D : Of kind like this .", "E : OK .", "E : So the envelope here is the output of the mel - filters", "A : Mm - hmm .", "E : and what we clearly see is that in some cases ,", "E : and it clearly appears here ,", "E : and the the harmonics are resolved by the f", "E : Well , there are still appear after mel - filtering ,", "A : Mm - hmm .", "E : and it happens for high pitched voice because the width of the lower frequency mel - filters is sometimes even smaller than the pitch .", "A : Yeah .", "E : It 's around one hundred , one hundred and fifty hertz Nnn .", "A : Right .", "E : And so what happens is that this uh , add additional variability to this envelope", "E : and um", "A : Yeah .", "E : so we were thinking to modify the mel - spectrum to have something that that 's smoother on low frequencies .", "A : That 's as as a separate thing .", "E : i", "A : Yeah .", "E : Yeah .", "E : This is a separate thing .", "D : Yeah .", "A : Separate thing ?", "A : Yeah .", "E : And .", "A : Yeah .", "A : Maybe so .", "A : Um .", "A : Yeah .", "A : So , what ", "A : Yeah .", "A : What I was talking about was just , starting with the FFT", "A : you could you could uh do a very rough thing to estimate estimate uh pitch .", "E : Yeah .", "E : Mm - hmm .", "A : And uh uh , given you know , given that , uh you could uh uh come up with some kind of estimate of how much of the low frequency energy was was explained by by uh uh those harmonics .", "E : Mm - hmm .", "A : Uh .", "A : It 's uh a variant on what you 're s what you 're doing .", "A : The I mean , the the the mel does give a smooth thing .", "A : But as you say it 's not that smooth here .", "A : And and so if you if you just you know subtracted off uh your guess of the harmonics then something like this would end up with quite a bit lower energy in the first fifteen hundred hertz or so", "A : and and our first kilohertz , even .", "E : Mm - hmm .", "A : And um if was uh noisy , the proportion that it would go down would be", "E : Mm - hmm .", "A : if it was if it was unvoiced or something .", "A : So you oughta be able to pick out voiced segments .", "A : At least it should be another another cue .", "E : Mm - hmm .", "A : So . Anyway .", "A : OK ?", "A : That 's what 's going on .", "A : Uh .", "A : What 's up with you ?", "B : Um our t I went to talk with uh Mike Jordan this this week", "A : Mm - hmm .", "B : um and uh shared with him the ideas about um extending the Larry Saul work", "B : and um I asked him some questions about factorial H M", "B : so like later down the line when we 've come up with these these feature detectors , how do we how do we uh you know , uh model the time series that that happens", "B : um and and we talked a little bit about factorial H M Ms and how um when you 're doing inference or w when you 're doing recognition , there 's like simple Viterbi stuff that you can do for for these H M", "B : and the uh the great advantages that um a lot of times the factorial H M Ms don't um don't over - alert the problem", "B : there they have a limited number of parameters and they focus directly on on uh the sub - problems at hand", "B : so you can imagine um five or so parallel um features um transitioning independently", "B : and then at the end you you uh couple these factorial H M Ms with uh with uh undirected links um based on based on some more data .", "A : Hmm .", "B : So he he seemed he seemed like really interested in in um in this", "B : and said said this is this is something very do - able and can learn a lot", "B : and um", "B : yeah , I 've just been continue reading um about certain things .", "A : Mm - hmm .", "B : um", "B : thinking of maybe using um um m modulation spectrum stuff to um as features um also in the in the sub - bands", "A : Mm - hmm .", "B : because it seems like the modulation um spectrum tells you a lot about the intelligibility of of certain um words and stuff", "B : So , um .", "B : Yeah .", "B : Just", "B : that 's about it .", "A : OK .", "C : OK .", "C : And", "C : um", "C : so I 've been looking at Avendano 's work", "C : and um", "C : uh I 'll try to write up in my next stat status report a nice description of what he 's doing ,", "C : but it 's it 's an approach to deal with reverberation or that the aspect of his work that I 'm interested in", "C : the idea is that um normally an analysis frames are um too short to encompass reverberation effects um in full .", "C : You miss most of the reverberation tail in a ten millisecond window", "C : and so you you 'd like it to be that um the reverberation responses um simply convolved um in ,", "C : but it 's not really with these ten millisecond frames", "C : cuz you j", "C : But if you take , say , a two millisecond um window ", "C : I 'm sorry", "C : a two second window", "C : then in a room like this , most of the reverberation response is included in the window", "C : and the then it um then things are l more linear .", "C : It is it is more like the reverberation response is simply c convolved", "C : and um and you can use channel normalization techniques like", "C : uh in his thesis he 's assuming that the reverberation response is fixed .", "C : He just does um mean subtraction ,", "C : which is like removing the DC component of the modulation spectrum", "C : and that 's supposed to d um deal uh deal pretty well with the um reverberation", "C : and um the neat thing is you can't take these two second frames and feed them to a speech recognizer", "C : um so he does this um method training trading the um the spectral resolution for time resolution and um come ca uh synthesizes a new representation which is with say ten second frames but a lower s um frequency resolution .", "C : So I don't really know the theory .", "C : I guess it 's these are called \" time frequency representations \"", "C : and h he 's making the the time sh um finer grained and the frequency resolution um less fine grained .", "E : Mm - hmm .", "C : s so I 'm I guess my first stab actually in continuing his work is to um re - implement this this thing which um changes the time and frequency resolutions", "C : cuz he doesn't have code for me .", "C : So that that 'll take some reading about the theory .", "C : I don't really know the theory .", "E : Mm - hmm .", "C : Oh , and um , another f first step is", "C : um ,", "C : so the the way I want to extend his work is make it able to deal with a time varying reverberation response", "C : um and um", "C : we don't really know how fast the um the reverberation response is varying the Meeting Recorder data", "C : um so um we we have this um block least squares um imp echo canceller implementation", "C : and um I want to try finding the the response , say , between a near mike and the table mike for someone using the echo canceller", "C : and looking at the echo canceller taps and then see how fast that varies from block to block .", "E : Mm - hmm .", "C : That should give an idea of how fast the reverberation response is changing .", "E : Mm - hmm .", "A : OK .", "A : Um . I think we 're sort of done .", "E : Yeah .", "A : So let 's read our digits and go home .", "C : Um .", "C : S so um", "C : y you do I think you read some of the the zeros as O 's and some as zeros .", "A : Yeah .", "C : Is there a particular way we 're supposed to read them ?", "E : There are only zeros here .", "E : Well .", "A : No . \" O \" \" O \" \" O \"", "A : \" O \" \" O \" \" O \" and \" zero \" are two ways that we say that digit .", "E : Eee .", "E : Yeah .", "A : So it 's ", "E : But ", "B : Ha !", "A : so it 's i", "E : Perhaps in the sheets there should be another sign for the ", "E : if we want to the the guy to say \" O \"", "E : or", "A : No .", "A : I mean . I think people will do what they say .", "E : It 's ", "E : Yeah .", "A : It 's OK .", "E : OK .", "C : Alright .", "A : I mean in digit recognition we 've done before , you have you have two pronunciations for that value , \" O \" and \" zero \" .", "C : OK .", "E : But it 's perhaps more difficult for the people to prepare the database then , if ", "E : because here you only have zeros", "A : No , they just write ", "E : and and people pronounce \" O \" or zero ", "A : they they write down OH .", "A : or they write down ZERO", "A : a and they and they each have their own pronunciation .", "E : Yeah but if the sh the sheet was prepared with a different sign for the \" O \" .", "A : But people wouldn't know what that wa", "A : I mean there is no convention for it .", "E : OK .", "E : Yeah .", "E : OK .", "A : See . I mean , you 'd have to tell them \" OK when we write this , say it tha \" ,", "A : you know ,", "A : and you just They just want people to read the digits as you ordinarily would", "E : Mm - hmm .", "E : Yeah .", "E : Yep .", "A : and and people say it different ways .", "C : OK .", "C : Is this a change from the last batch of of um forms ?", "C : Because in the last batch it was spelled out which one you should read .", "E : Yeah , it was orthographic ,", "E : so .", "A : Yes .", "A : That 's right .", "A : It was it was spelled out , and they decided they wanted to get at more the way people would really say things .", "C : Oh . OK .", "A : That 's also why they 're they 're bunched together in these different groups .", "C : OK .", "A : So so it 's ", "A : Yeah .", "A : So it 's it 's ", "A : Everything 's fine .", "C : OK .", "A : OK .", "A : Actually , let me just s since since you brought it up , I was just it was hard not to be self - conscious about that when it after we since we just discussed it .", "A : But I realized that that um when I 'm talking on the phone , certainly , and and saying these numbers , I almost always say zero .", "A : And uh cuz because uh i it 's two syllables .", "A : It 's it 's more likely they 'll understand what I said .", "A : So that that that 's the habit I 'm in ,", "A : but some people say \" O \" and ", "B : Yeah I normally say \" O \" cuz it 's easier to say .", "A : Yeah it 's shorter .", "A : Yeah .", "A : So it 's So . So uh .", "B : \" O \"", "A : Now , don't think about it .", "B : Oh , no !", "A : OK .", "A : We 're done ." ], "abstractive summary": [ { "id": "Bro013.s.1", "text": "The Main purpose of the meeting of ICSI's Meeting Recorder Group at Berkeley was to discuss the recent progress of it's members.", "type": "abstract" }, { "id": "Bro013.s.2", "text": "This includes reports on the progress of the groups main digit recogniser project, with interest on voice-activity detectors and voiced/unvoiced detection, work on acoustic feature detection, and research into dealing with reverberation.", "type": "abstract" }, { "id": "Bro013.s.3", "text": "There was also talk of comparing different recognition systems and training datasets, and a discussion of the pronunciation of the digit zero for the recording at the end of the meeting.", "type": "abstract" }, { "id": "Bro013.s.4", "text": "In his next status report, me026 will summarise the work he has been researching.", "type": "decisions" }, { "id": "Bro013.s.5", "text": "The digit recognition system is still not working well enough, they must get better results if they want to publish and be noticed.", "type": "problems" }, { "id": "Bro013.s.6", "text": "They have not really made many improvements, which may be due to their comparatively small training set, or the conditions the data is recorded under.", "type": "problems" }, { "id": "Bro013.s.7", "text": "The new VAD is quite a large network, and adds a delay to the process.", "type": "problems" }, { "id": "Bro013.s.8", "text": "This caused OGI to drop it, though speaker mn007 is assuming that a smaller and equally effective system can be developed.", "type": "problems" }, { "id": "Bro013.s.9", "text": "The alternative is to get yet another VAD form somewhere else, though it's not clear if they will even be required in the final system.", "type": "problems" }, { "id": "Bro013.s.10", "text": "There are some problems with the voiced/unvoiced feature detection, because some pitches are slipping through the filtering.", "type": "problems" }, { "id": "Bro013.s.11", "text": "The group have been comparing their recognition system to a few others, and theirs has not come off favourably.", "type": "progress" }, { "id": "Bro013.s.12", "text": "There could be many reasons for this, including smaller training set, more realistic data, or older technology.", "type": "progress" }, { "id": "Bro013.s.13", "text": "Speaker mn007 has put the best voice activity detector into the system, to great improvements along with designing new filters that run at the correct latency.", "type": "progress" }, { "id": "Bro013.s.14", "text": "Speaker fn002 has started to find parameters for voiced/unvoiced feature detection, and has found some classic ones, although there are other things she wishes to look at.", "type": "progress" }, { "id": "Bro013.s.15", "text": "Me013 offers a few ideas of simple things she may want to try, as he is not confident with everything she is trying.", "type": "progress" }, { "id": "Bro013.s.16", "text": "Speaker me006 is continuing with the idea of extending work on acoustic feature detection.", "type": "progress" }, { "id": "Bro013.s.17", "text": "He is continuing to read, and has discussed the suitability of factorial HMMs with a colleague.", "type": "progress" }, { "id": "Bro013.s.18", "text": "Speaker me026 has been learning more about previous work on reverberation, and is ready to start with a re-implementation of the theory.", "type": "progress" }, { "id": "Bro013.s.19", "text": "From there he wants to extend the work to look at time-varying reverb.", "type": "progress" } ], "extractive summary": [ "A : Um . What are we talking about today ?", "E : Uh , well , first there are perhaps these uh Meeting Recorder digits that we tested .", "A : The both the uh the SRI System and the oth", "A : And for one thing that that sure shows the difference between having a lot of uh training data or not ,", "A : uh , the uh The best kind of number we have on the English uh on near microphone only is is uh three or four percent .", "A : And uh it 's significantly better than that , using fairly simple front - ends on on the uh uh , with the SRI system .", "A : But that 's that 's using uh a a pretty huge amount of data ,", "A : In fact , mostly not digits for the actual training the H M Ms whereas uh in this case we 're just using digits for training the H M", "E : because it 's their very d huge , their huge system .", "A : But but uh what what I think I 'd be interested to do given that , is that we we should uh take ", "A : is to take some of these tandem things and feed it into the SRI system ,", "E : But I guess the main point is the data because", "E : Our back - end is is fairly simple", "E : but until now , well , the attempts to improve it or have fail", "A : I mean so to So there 's there 's there 's two things being affected .", "A : I mean . One is that that , you know , there 's something simple that 's wrong with the back - end .", "A : We 've been playing a number of states", "A : uh I I don't know if he got to the point of playing with the uh number of Gaussians yet", "A : But , yeah , so far he hadn't gotten any big improvement ,", "A : but that 's all with the same amount of data which is pretty small .", "E : So , yeah , we could retrain some of these tandem on on huge ", "A : Well , you could do that , but I 'm saying even with it not with that part not retrained ,", "A : just just using having the H M Ms much better H M", "E : perhaps it 's not related , the amount of data but the um recording conditions .", "E : The the fact that the result with the tandem and Aurora system are uh so much worse .", "A : What what is the problem that you 're trying to explain ?", "A : I uh but I 'm I 'm almost certain that it it I mean , that it has to do with the um amount of training data .", "A : It it 's it 's orders of magnitude off .", "A : let 's see , in the in these multi - train things did we include noisy data in the training ?", "A : I mean , that could be hurting us actually , for the clean case .", "A : You know , I don't think there 's anything magical here .", "A : It 's , you know , we used a simple HTK system with a modest amount of data .", "A : And this is a a , you know , modern uh system", "A : uh has has a lot of nice points to it .", "A : So . I mean , the HTK is an older HTK , even .", "A : But to me it just it just meant a practical point that um if we want to publish results on digits that that people pay attention to we probably should uh ", "A : Cuz we 've had the problem before that you get show some nice improvement on something that 's that 's uh , uh it seems like too large a number ,", "A : and uh uh people don't necessarily take it so seriously .", "A : I mean , there 's even though it 's close - miked there 's still there really is background noise .", "A : Um . And uh I suspect when the TI - digits were recorded if somebody fumbled or said something wrong or something that they probably made them take it over .", "A : It was not I mean there was no attempt to have it be realistic in any in any sense at all .", "E : TI - digit is it 's very , very clean and it 's like studio recording", "E : whereas these Meeting Recorder digits sometimes you have breath noise", "E : Perhaps the point is that we 've been working on is ,", "E : yeah , we have put the um the good VAD in the system", "E : and it really makes a huge difference .", "E : I think , yeah , this is perhaps one of the reason why our system was not not the best ,", "E : because with the new VAD , it 's very the results are similar to the France Telecom results and perhaps even better sometimes .", "E : Uh . The problem is that it 's very big and we still have to think how to where to put it", "E : uh either some delay", "E : and we if we put it on the server side , it doesn't work ,", "E : because on the server side features you already have LDA applied from the f from the terminal side", "E : and so you accumulate the delay", "A : So wha where did this good VAD come from ?", "E : It 's um from OGI .", "A : This is the one they had originally ?", "A : Yeah , but they had to get rid of it because of the space ,", "E : But the abso assumption is that we will be able to make a VAD that 's small and that works fine .", "A : But the other thing is uh to use a different VAD entirely .", "A : I I don't know what the thinking was amongst the the the the ETSI folk", "A : but um if everybody agreed sure let 's use this VAD and take that out of there ", "E : They just want , apparently they don't want to fix the VAD because they think there is some interaction between feature extraction and and VAD or frame dropping", "E : But they still want to just to give some um requirement for this VAD", "E : because it 's it will not be part of they don't want it to be part of the standard .", "E : So there just will be some requirements that are still not uh not yet uh ready I think .", "A : but I don't think we need to be stuck on using our or OGI 's VAD .", "A : We could use somebody else 's if it 's smaller", "A : You know , as long as it did the job .", "E : Uh I designed a new a new filter", "E : because when I designed other filters with shorter delay from the LDA filters , there was one filter with fif sixty millisecond delay and the other with ten milliseconds", "E : and uh Hynek suggested that both could have sixty - five sixty - s", "E : Both should have sixty - five because ", "A : You didn't gain anything , right ?", "E : and uh it 's running .", "E : Uh but the filter is of course closer to the reference filter .", "E : Yeah , and then we 've started to work with this of um voiced - unvoiced stuff .", "D : No , I w I begin to play with Matlab and to found some parameter robust for voiced - unvoiced decision .", "D : And we they we found that maybe w is a classical parameter , the sq the variance between the um FFT of the signal and the small spectrum of time we after the um mel filter bank .", "D : And , well , is more or less robust .", "D : Is good for clean speech .", "E : So , basically we wa want to look at something like the ex the ex excitation signal and ", "E : which are the variance of it and ", "A : So now I wonder ", "A : I know you want to get at something orthogonal from what you get with the smooth spectrum", "A : Um . But if you were to really try and get a voiced - unvoiced , do you do you want to totally ignore that ?", "A : I mean , clearly a a very big very big cues for voiced - unvoiced come from uh spectral slope and so on ,", "E : Well , this would be this would be perhaps an additional parameter ,", "D : Yeah because when did noise clear in these section is clear", "A : I mean , certainly if you want to do good voiced - unvoiced detection , you need a few features .", "A : But , you know , people look at at slope and uh first auto - correlation coefficient , divided by power .", "A : Each each feature is by itself not enough .", "A : Or maybe you could you just do it going through the P FFT 's figuring out some um probable um harmonic structure .", "E : yeah , it 's it 's another problem .", "E : If you look at this um spectrum ,", "E : Is it the mel - filters ?", "E : and what we clearly see is that in some cases ,", "E : and the the harmonics are resolved by the f", "E : Well , there are still appear after mel - filtering ,", "E : and it happens for high pitched voice because the width of the lower frequency mel - filters is sometimes even smaller than the pitch .", "E : so we were thinking to modify the mel - spectrum to have something that that 's smoother on low frequencies .", "A : What I was talking about was just , starting with the FFT", "A : you could you could uh do a very rough thing to estimate estimate uh pitch .", "A : And uh uh , given you know , given that , uh you could uh uh come up with some kind of estimate of how much of the low frequency energy was was explained by by uh uh those harmonics .", "A : It 's uh a variant on what you 're s what you 're doing .", "A : But as you say it 's not that smooth here .", "A : And and so if you if you just you know subtracted off uh your guess of the harmonics then something like this would end up with quite a bit lower energy in the first fifteen hundred hertz or so", "A : What 's up with you ?", "B : Um our t I went to talk with uh Mike Jordan this this week", "B : um and uh shared with him the ideas about um extending the Larry Saul work", "B : and um I asked him some questions about factorial H M", "B : so like later down the line when we 've come up with these these feature detectors , how do we how do we uh you know , uh model the time series that that happens", "B : um and and we talked a little bit about factorial H M Ms and how um when you 're doing inference or w when you 're doing recognition , there 's like simple Viterbi stuff that you can do for for these H M", "B : and the uh the great advantages that um a lot of times the factorial H M Ms don't um don't over - alert the problem", "B : there they have a limited number of parameters and they focus directly on on uh the sub - problems at hand", "B : So he he seemed he seemed like really interested in in um in this", "B : and said said this is this is something very do - able and can learn a lot", "B : yeah , I 've just been continue reading um about certain things .", "B : thinking of maybe using um um m modulation spectrum stuff to um as features um also in the in the sub - bands", "B : because it seems like the modulation um spectrum tells you a lot about the intelligibility of of certain um words and stuff", "C : so I 've been looking at Avendano 's work", "C : uh I 'll try to write up in my next stat status report a nice description of what he 's doing ,", "C : but it 's it 's an approach to deal with reverberation or that the aspect of his work that I 'm interested in", "C : s so I 'm I guess my first stab actually in continuing his work is to um re - implement this this thing which um changes the time and frequency resolutions", "C : cuz he doesn't have code for me .", "C : So that that 'll take some reading about the theory .", "C : I don't really know the theory .", "C : so the the way I want to extend his work is make it able to deal with a time varying reverberation response", "C : we don't really know how fast the um the reverberation response is varying the Meeting Recorder data", "C : y you do I think you read some of the the zeros as O 's and some as zeros .", "C : Is there a particular way we 're supposed to read them ?", "E : Perhaps in the sheets there should be another sign for the ", "A : I mean . I think people will do what they say .", "A : I mean in digit recognition we 've done before , you have you have two pronunciations for that value , \" O \" and \" zero \" .", "A : No , they just write ", "E : and and people pronounce \" O \" or zero ", "A : and you just They just want people to read the digits as you ordinarily would" ] }, "log": [] }, "ICSI--train--3": { "original dialog id": "Bed015.json", "dialog index": 3, "original dialog info": { "dialog history": [ "B : what things to talk about .", "F : I 'm What ?", "F : Really ?", "F : Oh , that 's horrible !", "F : Disincentive !", "A : OK , we 're recording .", "F : Hello ?", "B : Check check check check .", "F : Hello ?", "D : Uh , yeah .", "F : Which am I ?", "C : Oh right .", "B : Alright . Good .", "F : Channel fi OK .", "F : OK .", "F : Are you doing something ?", "F : OK , then I guess I 'm doing something .", "F : So , um ,", "F : So basically the result of m much thinking since the last time we met , um , but not as much writing , um , is a sheet that I have a lot of , like , thoughts and justification of comments on", "F : but I 'll just pass out as is right now .", "F : So , um ,", "F : here .", "F : If you could pass this around ?", "F : And there 's two things .", "F : And so one on one side is on one side is a sort of the revised sort of updated semantic specification .", "D : Um ", "F : And the other side is , um , sort of a revised construction formalism .", "D : The ", "D : wait .", "E : This is just one sheet , right ?", "D : Ah !", "F : It 's just one sheet .", "D : Just one sheet .", "F : It 's just a Nothing else .", "D : OK .", "D : Front , back .", "F : Um ,", "F : Enough to go around ?", "F : OK .", "F : And in some ways it 's it 's it 's very similar to ", "F : There are very few changes in some ways from what we 've , um , uh , b done before", "F : but I don't think everyone here has seen all of this .", "F : So , uh , I 'm not sure where to begin .", "F : Um , as usual the disclaimers are there are all these things are ", "F : it 's only slightly more stable than it was before .", "E : Mm - hmm .", "F : And , um , after a little bit more discussion and especially like Keith and I I have more linguistic things to settle in the next few days , um , it 'll probably change again some more .", "E : Yeah .", "F : Um , maybe I will ", "F : let 's start b let 's start on number two actually on the notation ,", "F : um , because that 's ,", "F : I 'm thinking , possibly a little more familiar to , um to people .", "F : OK , so the top block is just sort of a sort of abstract nota", "F : it 's sort of like , um , listings of the kinds of things that we can have .", "F : And certain things that have , um , changed , have changed back to this .", "F : There there 's been a little bit of , um , going back and forth .", "F : But basically obviously all constructions have some kind of name .", "F : I forgot to include that you could have a type included in this line .", "C : What I was gonna ", "F : So something like , um ", "C : Right .", "F : Well , there 's an example ", "F : the textual example at the end has clausal construction .", "F : So ,", "F : um , just to show it doesn't have to be beautiful", "F : It could be , you know , simple old text as well .", "F : Um , there are a couple of ", "F : Uh , these three have various ways of doing certain things .", "F : So I 'll just try to go through them .", "F : So they could all have a type at the beginning .", "F : Um , and then they say the key word construction", "C : Oh , I see .", "F : and they have some name .", "C : So so the current syntax is if it s if there 's a type it 's before construct", "F : Yeah , right .", "C : OK , that 's fine .", "F : OK , and then it has a block that is constituents .", "F : And as usual I guess all the constructions her all the examples here have only , um , tsk one type of constituent ,", "F : that is a constructional constituent .", "F : I think that 's actually gonna turn out to m be certainly the most common kind .", "F : But in general instead of the word \" construct \" , th here you might have \" meaning \" or \" form \" as well .", "F : OK ?", "F : So if there 's some element that doesn't that isn't yet constructional in the sense that it maps form and meaning .", "F : OK ,", "F : um , the main change with the constructs which each of which has , um , the key word \" construct \" and then some name , and then some type specification , is that it 's ", "F : it 's pro it 's often sometimes the case in the first case here that you know what kind of construction it is .", "F : So for example whatever I have here is gonna be a form of the word \" throw \" ,", "F : or it 's gonna be a form of the word , you know , I don't know , \" happy \" , or something like that .", "F : Or , you know , some it 'll be a specific word", "F : or maybe you 'll have the type .", "F : You 'll say \" I need a p uh spatial relation phrase here \"", "F : or \" I need a directional specifier here \" .", "F : So - uh you could have a j a actual type here .", "F : Um , or you could just say in the second case that you only know the meaning type .", "F : So a very common example of this is that , you know , in directed motion , the first person to do something should be an agent of some kind ,", "F : often a human .", "F : Right ?", "F : So if I you know , the um , uh , run down the street then I ", "F : I I run down the street ,", "F : it 's typed , uh , \" I \" ,", "F : meaning category is what 's there .", "F : The the new kind is this one that is sort of a pair", "F : and , um ,", "F : sort of skipping fonts and whatever .", "F : The idea is that sometimes there are , um , general constructions that you know , that you 're going to need .", "F : It 's it 's the equivalent of a noun phrase or a prepositional phrase , or something like that there .", "E : Mm - hmm .", "F : And usually it has formal um , considerations that will go along with it .", "C : Mm - hmm .", "F : And then uh , you might know something much more specific depending on what construction you 're talking about ,", "F : about what meaning what specific meaning you want .", "F : So the example again at the bottom , which is directed motion ,", "F : you might need a nominal expression to take the place of , you know , um , \" the big th \" , you you know , \" the big the tall dark man \" , you know , \" walked into the room \" .", "E : Mm - hmm .", "F : But because of the nature of this particular construction you know not just that it 's nominal of some kind", "F : but in particular , that it 's some kind of animate nominal ,", "F : and which will apply just as well to like , you know , a per you know , a simple proper noun", "F : or to some complicated expression .", "F : Um ,", "F : so I don't know if the syntax will hold but something that gives you a way to do both constructional and meaning types .", "F : So .", "F : OK , then I don't think the , um ", "F : at least ", "F : Yeah . None of these examples have anything different for formal constraints ?", "F : But you can refer to any of the , um , sort of available elements and scope ,", "F : right ? which here are the constructs , to say something about the relation .", "F : And I think i if you not if you compare like the top block and the textual block , um , we dropped like the little F subscript .", "F : The F subscripts refer to the \" form \" piece of the construct .", "C : Good .", "F : And I think that , um , in general it 'll be unambiguous .", "F : Like if you were giving a formal constraint then you 're referring to the formal pole of that .", "F : So so by saying if I just said \" Name one \" then that means name one formal", "F : and we 're talking about formal struc Which which makes sense .", "F : Uh , there are certain times when we 'll have an exception to that ,", "F : in which case you could just indicate \" here I mean the meaningful for some reason \" .", "F : Right ?", "F : Or Actually it 's more often that , only to handle this one special case of , you know , \" George and Jerry walk into the room in that order \" .", "F : So we have a few funny things where something in the meaning might refer to something in the form .", "E : Mm - hmm .", "F : But but s we 're not gonna really worry about that for right now", "F : and there are way We can be more specific if we have to later on .", "F : OK , and so in terms of the the relations ,", "F : you know , as usual they 're before and ends .", "F : I should have put an example in of something that isn't an interval relation", "F : but in form you might also have a value binding .", "F : You know , you could say that , um , you know , \" name - one dot \" , t you know , \" number equals \" , you know , a plural or something like that .", "E : Mm - hmm .", "F : There are certain things that are attribute - value , similar to the bindings below", "F : but I mean they 're just us usually they 're going to be value value fillers , right ?", "F : OK , and then again semantic constraints here are just are just bindings .", "F : There was talk of changing the name of that .", "F : And Johno and I I you you and I can like fight about that if you like ?", "F : but about changing it to \" semantic n effects \" ,", "F : which I thought was a little bit too order - biased", "B : Well ", "B : Th", "F : and \" semantic bindings \" ,", "F : which I thought might be too restrictive in case we don't have only bindings .", "F : And so it was an issue whether constraints um , there were some linguists who reacted against \" constraints \" , saying , \" oh , if it 's not used for matching , then it shouldn't be called a constraint \" .", "F : But I think we want to be uncommitted about whether it 's used for matching or not . Right ?", "F : Cuz there are I think we thought of some situations where it would be useful to use whatever the c bindings are , for", "F : actual , you know , sort of like modified constraining purposes .", "C : Well , you definitely want to de - couple the formalism from the parsing strategy .", "E : Yeah .", "C : So that whether or not it 's used for matching or only for verification , I ", "F : Yeah , yeah .", "F : It 's used shouldn't matter , right ?", "F : Mm - hmm .", "C : s For sure .", "F : Mm - hmm .", "C : I mean , I don't know what , uh , term we want to use", "C : but we don't want to ", "F : Yeah , uh , there was one time when when Hans explained why \" constraints \" was a misleading word for him .", "C : Yep .", "F : And I think the reason that he gave was similar to the reason why Johno thought it was a misleading term ,", "F : which was just an interesting coincidence .", "F : Um , but , uh ", "F : And so I was like , \" OK , well both of you don't like it ?", "F : Fine , we can change it \" .", "C : It 's g it 's gone .", "F : But I I I 'm starting to like it again .", "F : So that that 's why That 's why I 'll stick with it .", "B : But ", "F : So ", "A : Well , you know what ?", "A : If you have an \" if - then \" phrase , do you know what the \" then \" phrase is called ?", "C : Th", "F : What ?", "F : Con - uh , a consequent ?", "A : Yeah .", "F : Yeah , but it 's not an \" if - then \" .", "C : I know .", "C : Anyway , so the other the other strategy you guys could consider is when you don't know what word to put , you could put no word ,", "A : No , but ", "F : Mm - hmm .", "C : just meaning .", "C : OK ?", "E : Yeah .", "C : And the then let ", "F : Yeah , that 's true .", "B : So that 's why you put semantic constraints up top and meaning bindings down down here ?", "F : Oh , oops !", "F : No . That was just a mistake of cut and paste from when I was going with it .", "B : OK .", "C : OK .", "F : So , I 'm sorry .", "F : I didn't mean that one 's an in unintentional .", "B : So this should be semantic and ", "F : Sometimes I 'm intentionally inconsistent", "F : cuz I 'm not sure yet .", "F : Here , I actually it was just a mistake .", "B : Th - so this definitely should be \" semantic constraints \" down at the bottom ?", "E : Sure .", "F : Yeah .", "B : OK .", "F : Well , unless I go with \" meaning \"", "F : but i I mean , I kind of like \" meaning \" better than \" semantic \"", "B : Or ", "C : Oh , whatever .", "F : but I think there 's vestiges of other people 's biases .", "C : Or wh", "F : Like ", "C : That - b", "C : Right .", "C : Minor min problem ", "F : Minor point .", "E : Extremely .", "C : OK .", "F : OK , um ,", "F : so I think the middle block doesn't really give you any more information , ex than the top block .", "F : And the bottom block similarly only just illus you know , all it does is illustrate that you can drop the subscripts and and that you can drop the , um uh , that you can give dual types .", "F : Oh , one thing I should mention is about \" designates \" .", "F : I think I 'm actually inconsistent across these as well .", "F : So , um , strike out the M subscript on the middle block .", "C : Mm - hmm .", "F : So basically now , um , this is actually this little change actually goes along with a big linguistic change ,", "F : which is that \" designates \" isn't only something for the semantics to worry about now .", "C : Good .", "F : So we want s \" designates \" to actually know one of the constituents which acts like a head in some respects", "F : but is sort of , um , really important for say composition later on .", "F : So for instance , if some other construction says , you know , \" are you of type is this part of type whatever \" , um , the \" designates \" tells you which sort of part is the meaning part .", "F : OK , so if you have like \" the big red ball \" , you know , you wanna know if there 's an object or a noun .", "F : Well , ball is going to be the designated sort of element of that kind of phrase .", "E : Mmm .", "F : Um ,", "F : there is a slight complication here which is that when we talk about form it 's useful sometimes to talk about , um to talk about there also being a designated object", "F : and we think that that 'll be the same one , right ?", "F : So the ball is the head of the phrase ,", "F : \" the r the \" , um , \" big red ball \" ,", "F : and the entity denoted by the word \" ball \" is sort of the semantic head in some ways of of this sort of , um , in interesting larger element .", "C : A a and the ", "C : Yeah .", "C : And there 's uh there 's ca some cases where the grammar depends on some form property of the head .", "F : Mm - hmm .", "E : Yeah .", "C : And and this enables you to get that , if I understand you right .", "E : Yeah .", "F : Right , right .", "E : That 's the idea .", "E : Yeah .", "C : Yeah yeah .", "F : And , uh , you might be able to say things like if the head has to go last in a head - final language , you can refer to the head as a p the , you know the formal head as opposed to the rest of the form having to be at the end of that decision .", "C : Right .", "F : So that 's a useful thing", "F : so that you can get some internal structural constraints in .", "C : OK , so that all looks good .", "C : Let me ", "C : Oh , w Oh .", "C : I don't know .", "C : Were you finished ?", "F : Um , there was a list of things that isn't included", "F : but you you can you can ask a question .", "F : That might @ @ it .", "C : OK .", "C : So , i if I understand this the ", "C : aside from , uh , construed and all that sort of stuff , the the differences are mainly that , we 've gone to the possibility of having form - meaning pairs for a type", "F : Mm - hmm .", "C : or actually gone back to ,", "F : Right .", "C : if we go back far enough ", "F : Well , except for their construction meaning ,", "F : so it 's not clear that ,", "F : uh ", "F : Well , right now it 's a c uh contr construction type and meaning type .", "F : So I don't know what a form type is .", "C : Oh , I see .", "C : Yeah , yeah , yeah .", "C : I 'm sorry ,", "F : Yeah .", "C : you 're right .", "C : A construction type .", "C : Uh , that 's fine .", "F : Right .", "C : But it , um ", "F : A well , and a previous , um , you know , version of the notation certainly allowed you to single out the meaning bit by it .", "F : So you could say \" construct of type whatever designates something \" .", "C : Yeah .", "F : But that was mostly for reference purposes ,", "F : just to refer to the meaning pole .", "F : I don't think that it was often used to give an extra meaning const type constraint on the meaning ,", "F : which is really what we want most of the time I think .", "C : Mm - hmm .", "F : Um ,", "F : I I don't know if we 'll ever have a case where we actually h", "F : if there is a form category constraint , you could imagine having a triple there that says , you know that 's kind of weird .", "C : No , no , no , I don't think so .", "C : I think that you 'll you 'll do fine .", "E : I ", "C : In fact , these are , um ,", "C : as long as as Mark isn't around , these are form constraints .", "C : So a nominal expression is uh , the fact that it 's animate , is semantic .", "C : The fact that it 's n uh , a nominal expression I would say on most people 's notion of of f you know , higher form types , this i this is one .", "F : Mm - hmm .", "E : Yeah .", "F : Right , right .", "E : Yeah , yeah .", "C : And I think that 's just fine .", "F : Which is fine , yeah .", "E : It 's that now , um , I 'm mentioned this ,", "C : Yeah .", "E : I I don't know if I ever explained this", "E : but the point of , um ,", "E : I mentioned in the last meeting , the point of having something called \" nominal expression \" is , um ,", "E : because it seems like having the verb subcategorize for , you know , like say taking as its object just some expression which , um , designates an object or designates a thing , or whatever , um , that leads to some syntactic problems basically ?", "E : So you wanna , you know you sort of have this problem like \" OK , well , I 'll put the word \" , uh , let 's say , the word \" dog \" ,", "E : you know . And that has to come right after the verb", "F : Mm - hmm .", "E : cuz we know verb meets its object .", "E : And then we have a construction that says , oh , you can have \" the \" preceding a noun .", "E : And so you 'd have this sort of problem that the verb has to meet the designatum .", "C : Right .", "E : And you could get , you know , \" the kicked dog \" or something like that , meaning \" kicked the dog \" .", "C : Right .", "E : Um , so you kind of have to let this phrase idea in there", "E : but ", "C : That I I have no problem with it at all .", "E : It - it", "E : Yeah .", "F : Yeah .", "C : I think it 's fine .", "F : Right , n s", "F : you may be you may not be like everyone else in in Berkeley ,", "E : Yeah .", "E : Yeah .", "F : but that 's OK .", "E : I mean , we we we sort of thought we were getting away with , uh with , a p", "F : Uh , we don't mind either ,", "F : so ", "E : I mean , this is not reverting to the X - bar theory of of phrase structure .", "C : Right .", "E : But , uh ,", "F : Right .", "E : I just know that this is Like , we didn't originally have in mind that , uh that verbs would subcategorize for a particular sort of form .", "F : Mm - hmm .", "C : But they do .", "E : Um ,", "E : but they does .", "F : Well , there 's an alternative to this", "E : At least in English .", "F : which is , um The question was did we want directed motion ,", "C : Yeah .", "F : which is an argument structure construction ", "E : Yeah .", "C : Mm - hmm .", "F : did we want it to worry about , um , anything more than the fact that it , you know , has semantic ", "F : You know , it 's sort of frame - based construction .", "F : So one option that , you know , Keith had mentioned also was like , well if you have more abstract constructions such as subject , predicate , basically things like grammatical relations ,", "E : Mm - hmm .", "F : those could intersect with these", "F : in such a way that subject , predicate , or subject , predicate , subject , verb , ob you know , verb object would require that those things that f fill a subject and object are NOM expressions .", "C : Right .", "F : And that would be a little bit cleaner in some way .", "F : But you know , for now , I mean ,", "C : Yeah .", "C : But it y y it 's yeah , just moving it moving the c the cons the constraints around .", "F : uh , you know .", "F : M moving it to another place , right .", "E : Yeah .", "C : OK , so that 's ", "F : But there does basically , the point is there has to be that constraint somewhere , right ?", "C : Right .", "F : So ,", "F : yeah .", "C : And so that was the ", "F : Robert 's not happy now ?", "A : No !", "F : Oh , OK .", "C : OK , and sort of going with that is that the designatum also now is a pair .", "F : Yes .", "C : Instead of just the meaning .", "F : Mm - hmm .", "C : And that aside from some terminology , that 's basically it .", "F : Right .", "C : I just want to b I 'm I 'm asking .", "F : Yep .", "E : Mm - hmm .", "E : Yeah .", "F : Yeah ,", "F : um , the un sort of the un - addressed questions in this , um , definitely would for instance be semantic constraints we talked about .", "C : Yeah .", "F : Here are just bindings", "F : but ,", "F : right ?", "F : we might want to introduce mental spaces ", "F : You know , there 's all these things that we don't ", "C : The whole the mental space thing is clearly not here .", "F : Right ?", "F : So there 's going to be some extra you know , definitely other notation we 'll need for that which we skip for now .", "E : Mm - hmm .", "C : By the way , I do want to get on that as soon as Robert gets back .", "F : Uh", "F : Yeah .", "F : OK .", "C : So , uh , the the mental space thing .", "E : Mm - hmm .", "C : Um , obviously , construal is a b is a b is a big component of that", "C : so this probably not worth trying to do anything till he gets back .", "C : But sort of as soon as he gets back I think um , we ought to ", "F : Mm - hmm .", "F : Mm - hmm .", "E : So what 's the what 's the time frame ?", "E : I forgot again when", "E : you 're going away for how long ?", "A : Just , uh , as a sort of a mental bridge , I 'm not I 'm skipping fourth of July .", "E : OK .", "A : So , uh , right afterwards I 'm back .", "E : OK .", "F : What ?", "F : You 're missing like the premier American holiday ?", "F : What 's the point of spending a year here ?", "A : Uh , I 've had it often enough .", "F : So , anyway .", "B : Well he w he went to college here .", "F : Oh , yeah , I forgot .", "F : Oops . Sorry .", "C : Yeah .", "F : OK .", "C : And furthermore it 's well worth missing .", "F : Not in California .", "F : Yeah , that 's true .", "E : Yes .", "F : I like I I like spending fourth of July in other countries , whenever I can .", "C : Right .", "F : Um ", "C : OK , so that 's great .", "F : Construal ,", "F : OK , so ", "F : Oh , so there was one question that came out .", "F : I hate this thing .", "F : Sorry .", "F : Um , which is ,", "F : so something like \" past \" which i you know , we think is a very simple uh , we 've often just stuck it in as a feature ,", "C : Right .", "F : you know , \" oh , this event takes place before speech time \" , OK , is what this means .", "C : Right .", "F : Um , it 's often thought of as it is also considered a mental space ,", "C : Right .", "F : you know , by , you know , lots of people around here .", "C : Right .", "F : So there 's this issue of well sometimes there are really exotic explicit space builders that say \" in France , blah - blah - blah \" ,", "E : Mm - hmm .", "F : and you have to build up you ha you would imagine that would require you , you know , to be very specific about the machinery ,", "F : whereas past is a very conventionalized one", "F : and we sort of know what it means", "F : but it we doesn't don't necessarily want to , you know , unload all the notation every time we see that it 's past tense .", "C : Right .", "F : So , you know , we could think of our uh , just like X - schema \" walk \" refers to this complicated structure , past refers to , you know , a certain configuration of this thing with respect to it .", "C : I think that 's exactly right .", "F : So so we 're kind of like having our cake and eating it ", "F : you know , having it both ways , right ?", "C : Yeah .", "F : So , i i", "C : Yeah . No , I think I think that i we 'll have to see how it works out when we do the details", "F : Mm - hmm .", "C : but my intuition would be that that 's right .", "F : Mm - hmm .", "F : Yeah , OK .", "A : Do you want to do the same for space ?", "F : Wha - sorry ?", "A : Space ?", "F : Space ?", "A : Here ?", "A : Now ?", "F : Oh , oh , oh , oh , instead of just time ?", "F : Yeah , yeah , yeah .", "A : Mm - hmm .", "F : Same thing .", "F : So there are very conventionalized like deictic ones , right ?", "F : And then I think for other spaces that you introduce , you could just attach y", "F : whatever ", "F : You could build up an appropriately uh , appropriate structure according to the l the sentence .", "A : Hmm .", "C : Yeah .", "A : Hmm ,", "A : well this this basically would involve everything you can imagine to fit under your C dot something ", "E : N", "F : Yeah .", "A : you know , where where it 's contextually dependent ,", "F : Right .", "F : Mm - hmm .", "A : \" what is now ,", "A : what was past ,", "A : what is in the future ,", "A : where is this ,", "A : what is here ,", "A : what is there ,", "A : what is \"", "F : Mm - hmm .", "F : Yeah .", "F : So time and space .", "F : Um , we 'll we 'll get that on the other side a little , like very minimally .", "F : There 's a sort of there 's a slot for setting time and setting place .", "C : Good .", "F : And", "F : you know , you could imagine for both of those are absolute things you could say about the time and place ,", "F : and then there are many in more interestingly , linguistically anyway , there are relative things that , you know , you relate the event in time and space to where you are now .", "F : If there 's something a lot more complicated like , or so hypothetical or whatever , then you have to do your job ,", "E : Mm - hmm .", "F : like or somebody 's job anyway .", "E : Yeah .", "F : I 'm gonna point to at random .", "E : Yeah .", "E : I mean , I 'm I 'm s curious about how much of the mental ", "E : I mean , I 'm not sure that the formalism , sort of the grammatical side of things , is gonna have that much going on in terms of the mental space stuff .", "E : You know , um , basically all of these so - called space builders that are in the sentence are going to sort of ", "E : I think of it as , sort of giving you the coordinates of , you know ", "E : assuming that at any point in discourse there 's the possibility that we could be sort of talking about a bunch of different world scenarios , whatever , and the speaker 's supposed to be keeping track of those . The , um the construction that you actually get is just gonna sort of give you a cue as to which one of those that you 've already got going , um , you 're supposed to add structure to .", "F : Mm - hmm .", "E : So \" in France , uh , Watergate wouldn't have hurt Nixon \" or something like that .", "E : Um , well , you say , \" alright , I 'm supposed to add some structure to my model of this hypothetical past France universe \" or something like that .", "E : The information in the sentence tells you that much", "E : but it doesn't tell you like exactly what it what the point of doing so is .", "E : So for example , depending on the linguistic con uh , context it could be ", "E : like the question is for example , what does \" Watergate \" refer to there ?", "E : Does it , you know does it refer to , um ", "E : if you just hear that sentence cold , the assumption is that when you say \" Watergate \" you 're referring to \" a Watergate - like scandal as we might imagine it happening in France \" .", "E : But in a different context , \" oh , you know , if Nixon had apologized right away it wouldn't you know , Watergate wouldn't have hurt him so badly in the US", "E : and in France it wouldn't have hurt him at all \" .", "E : Now we 're s now that \" Watergate \" we 're now talking about the real one ,", "F : They 're real , right .", "E : and the \" would \" sort of it 's a sort of different dimension of hypothe - theticality , right ?", "F : I see right .", "E : We 're not saying What 's hypothetical about this world .", "E : In the first case , hypothetically we 're imagining that Watergate happened in France .", "F : Hmm .", "E : In the second case we 're imagining hypothetically that Nixon had apologized right away", "F : Mm - hmm .", "F : Right .", "E : or something . Right ?", "E : So a lot of this isn't happening at the grammatical level .", "E : Uh , um , and so ", "C : Correct .", "F : Mm - hmm .", "E : I don't know where that sits then ,", "A : Hmm .", "E : sort of the idea of sorting out what the person meant .", "F : It seems like , um , the grammatical things such as the auxiliaries that you know introduce these conditionals , whatever , give you sort of the the most basi", "E : Mm - hmm .", "F : th those we I think we can figure out what the possibilities are , right ?", "E : Mm - hmm .", "F : There are sort of a relatively limited number .", "F : And then how they interact with some extra thing like \" in France \" or \" if such - and - such \" , that 's like there are certain ways that they c they can ", "E : Yeah .", "F : You know , one is a more specific version of the general pattern that the grammat grammar gives you .", "F : I think .", "E : Yeah .", "F : But , you know , whatever ,", "F : we we 're ", "C : Yeah , in the short run all we need is a enough mechanism on the form side to get things going .", "E : Mm - hmm .", "E : Yeah .", "C : Uh , I uh , you you ", "E : But the whole point of the whole point of what Fauconnier and Turner have to say about , uh , mental spaces , and blending , and all that stuff is that you don't really get that much out of the sentence .", "E : You know , there 's not that much information contained in the sentence .", "E : It just says , \" Here .", "E : Add this structure to this space . \"", "E : and exactly what that means for the overall ongoing interpretation is quite open .", "E : An individual sentence could mean a hundred different things depending on , quote , \" what the space configuration is at the time of utterance \" .", "F : Mm - hmm .", "F : Mm - hmm .", "E : And so somebody 's gonna have to be doing a whole lot of work but not me , I think .", "C : Well ", "C : I think that 's right .", "C : Oh , I yeah , I , uh , uh I think that 's Not k I th", "C : I don't think it 's completely right .", "C : I mean , in fact a sentence examples you gave in f did constrain the meaning b", "C : the form did constrain the meaning ,", "E : Yeah .", "C : and so , um , it isn't , uh ", "E : Sure , but like what what was the point of saying that sentence about Nixon and France ?", "E : That is not there is nothing about that in the in the sentence really .", "F : That 's OK .", "F : We usually don't know the point of the sentence at all .", "E : Yeah .", "F : But we know what it 's trying to say .", "F : We we know that it 's what predication it 's setting up .", "E : Y yeah .", "C : Yeah .", "C : But but bottom line , I agree with you ,", "E : Yeah .", "F : That 's all .", "E : Yeah .", "C : that that that we 're not expecting much out of the , uh f", "F : Purely linguistic cues , right ?", "C : uh , the purely form cues , yeah .", "F : So .", "F : Mmm .", "C : And , um I mean , you 're you 're the linguist", "C : but , uh , it seems to me that th these we we you know , we 've talked about maybe a half a dozen linguistics theses in the last few minutes or something .", "E : Yeah , yeah .", "C : Yeah , I mean ", "E : Yeah .", "E : Oh , yeah .", "C : uh , I I mean , that that 's my feeling that that these are really hard uh , problems that decide exactly what what 's going on .", "E : Mm - hmm .", "E : Yeah .", "E : Yeah .", "C : OK .", "F : OK , so , um , one other thing I just want to point out is there 's a lot of confusion about the terms like \" profile , designate , focus \" , et cetera , et cetera .", "E : Mm - hmm .", "C : Uh , right , right , right .", "F : Um , for now I 'm gonna say like \" profile \" 's often used ", "F : like two uses that come to mind immediately .", "F : One is in the traditional like semantic highlight of one element with respect to everything else .", "F : So \" hypotenuse \" , you profiled this guy against the background of the right t right triangle .", "E : Mm - hmm .", "F : OK .", "F : And the second use , um , is in FrameNet. It 's", "F : slightly different .", "F : Oh , I was asking Hans about this .", "F : They use it to really mean , um , this in a frame th this is the profiles on the these are the ones that are required .", "F : So they have to be there or expressed in some way .", "F : Which which I 'm not saying one and two are mutually exclusive", "F : but they 're they 're different meanings .", "C : Right .", "E : Mm - hmm .", "F : So the closest thing ", "F : so I was thinking about how it relates to this notation .", "F : For us , um ", "F : OK , so how is it ", "F : so \" designate \" FrameNet ?", "C : Does that Is that really what they mean in in ", "F : FrameNet ?", "C : I didn't know that .", "F : Yeah , yeah .", "F : I I mean , I I was a little bit surprised about it too .", "F : I knew that ", "C : Yeah .", "F : I thought that that would be something like ", "F : there 's another term that I 've heard for that thing", "C : Right , OK .", "F : but they I mean uh , well , at least Hans says they use it that way .", "F : And ", "C : Well , I 'll check .", "F : and may maybe he 's wrong .", "F : Anyway , so", "F : I think the the \" designate \" that we have in terms of meaning is really the \" highlight this thing with respect to everything else \" . OK ?", "C : Right .", "F : So this is what what it means .", "F : But the second one seems to be useful", "F : but we might not need a notation for it ?", "F : We don't have a notation for it", "F : but we might want one .", "F : So for example we 've talked about if you 're talking about the lexical item \" walk \" ,", "F : you know it 's an action .", "F : Well , it also has this idea it carries along with it the idea of an actor", "F : or somebody 's gonna do the walking .", "F : Or if you talk about an adjective \" red \" , it carries along the idea of the thing that has the property of having color red .", "F : So we used to use the notation \" with \" for this", "F : and I think that 's closest to their second one .", "C : Right .", "F : So I d don't yet know , I have no commitment , as to whether we need it .", "F : It might be ", "F : it 's the kind of thing that w a parser might want to think about whether we require ", "F : you know , these things are like it 's semantically part of it ", "C : N no , no .", "C : Well , uh , th critically they 're not required syntactically .", "C : Often they 're pres presu presupposed and all that sort of stuff .", "F : Right .", "F : Right , right . Yeah , um , definitely .", "F : So , um ,", "F : \" in \" was a good example .", "F : If you walk \" in \" , like well , in what ?", "F : You know , like you have to have the So so it 's only semantically is it ", "C : Right , there 's ", "F : it is still required , say , by simulation time though", "F : to have something .", "C : Right .", "F : So it 's that I meant the idea of like that the semantic value is filled in by sim simulation .", "F : I don't know if that 's something we need to spa to to like say ever as part of the requirement ? ", "F : or the construction ?", "F : or not .", "F : We 'll we 'll again defer .", "C : Or I mean , or or , uh", "C : so the ", "F : Have it construed ,", "F : is that the idea ?", "C : Yeah , yeah .", "F : Just point at Robert .", "F : Whenever I 'm confused just point to him .", "C : Right .", "F : You tell me .", "C : It 's it 's his thesis , right ?", "F : OK .", "C : Anyway ,", "C : right , yeah , w this is gonna be a b", "C : you 're right , this is a bit of in a mess", "C : and we still have emphasis as well , or stress , or whatever .", "F : OK , well we 'll get , uh uh , I we have thoughts about those as well .", "C : Yeah .", "C : Great .", "F : Um ,", "F : the I w I would just", "F : s some of this is just like my you know , by fiat .", "F : I 'm going to say , this is how we use these terms .", "F : I don't - you know , there 's lots of different ways in the world that people use it .", "E : Yeah .", "C : I that 's fine .", "F : I think that , um , the other terms that are related are like focus and stress .", "F : So , s", "C : Mm - hmm .", "F : I think that the way I we would like to think , uh , I think is focus is something that comes up in , I mean , lots of ", "F : basically this is the information structure .", "F : OK , it 's like uh , it 's not it might be that there 's a syntactic , uh , device that you use to indicate focus", "C : Mm - hmm .", "F : or that there are things like , you know ,", "F : I think Keith was telling me , things toward the end of the sentence , post - verbal , tend to be the focused focused element ,", "E : Mmm .", "F : the new information .", "F : You know , if I \" I walked into the room \" , you tend to think that , whatever , \" into the room \" is sort of like the more focused kind of thing .", "E : Mm - hmm .", "E : Yeah .", "F : And when you , uh , uh , you have stress on something that might be , you know , a cue that the stressed element , or for instance , the negated element is kind of related to information structure .", "F : So that 's like the new the sort of like import or whatever of of this thing .", "F : Uh , so so I think that 's kind of nice to keep \" focus \" being an information structure term .", "F : \" Stress \" ", "F : I th and then there are different kinds of focus that you can bring to it .", "F : So , um , like \" stress \" , th stress is kind of a pun on you might have like whatever , like , um , accent kind of stress .", "E : Mm - hmm .", "F : And that 's just a ", "F : uh , w we 'll want to distinguish stress as a form device .", "F : You know , like , oh , high volume or whatever .", "E : Yeah .", "F : Um ,", "F : t uh , and distinguish that from it 's effect", "F : which is , \" Oh , the kind of focus we have is we 're emphasizing this value often as opposed to other values \" , right ?", "F : So focus carries along a scope .", "F : Like if you 're gonna focus on this thing and you wanna know it sort of evokes all the other possibilities that it wasn't .", "F : Um , so my classic my now - classic example of saying , \" Oh , he did go to the meeting ? \" ,", "E : Mm - hmm .", "E : Yeah .", "F : that was my way of saying as opposed to , you know , \" Oh , he didn't g \" or \" There was a meeting ? \"", "F : I think that was the example that was caught on by the linguists immediately .", "E : Yeah .", "E : Yeah .", "F : And so , um , the like if you said he ", "F : you know , there 's all these different things that if you put stress on a different part of it then you 're , c focusing , whatever , on , uh ", "E : Mm - hmm .", "F : \" he walked to the meeting \" as opposed to \" he ran \" ,", "F : or \" he did walk to the meeting \" as opposed to \" he didn't walk \" . You know ,", "E : Mm - hmm .", "F : so we need to have a notation for that", "F : which , um ,", "F : I think that 's still in progress .", "F : So , sort of I 'm still working it out .", "F : But it did one one implication it does f have for the other side , which we 'll get to in a minute is that I couldn't think of a good way to say \" here are the possible things that you could focus on \" ,", "F : cuz it seems like any entity in any sentence , you know , or any meaning component of anyth you know all the possible meanings you could have , any of them could be the subject of focus .", "C : Mmm .", "F : But I think one the one thing you can schematize is the kind of focus , right ?", "F : So for instance , you could say it 's the the tense on this as opposed to , um , the the action . OK .", "F : Or it 's uh , it 's an identity thing or a contrast with other things ,", "F : or stress this value as opposed to other things .", "F : So , um ,", "F : it 's it is kind of like a profile profile - background thing", "F : but I I can't think of like the limited set of possible meanings that you would that you would focu", "E : Light up with focus , yeah .", "F : light highlight", "F : as opposed to other ones .", "F : So it has some certain complications for the , uh , uh later on .", "F : Li - I mean , uh , the best thing I can come up with is that information has a list of focused elements .", "F : For instance , you Oh , one other type that I forgot to mention is like query elements", "F : and that 's probably relevant for the like \" where is \" , you know , \" the castle \" kind of thing ?", "E : Mm - hmm .", "F : Because you might want to say that , um , location or cert certain WH words bring you know , sort of automatically focus in a , you know , \" I don't know the identity of this thing \" kind of way on certain elements .", "F : So .", "F : OK . Anyway . So that 's onl there are there are many more things that are uncl that are sort of like a little bit unstable about the notation", "F : but it 's most I think it 's this is , you know , the current current form .", "F : Other things we didn't totally deal with , um ,", "F : well , we 've had a lot of other stuff that Keith and I have them working on in terms of like how you deal with like an adjective .", "E : Oh , there 's a bunch .", "E : Yeah .", "F : You know , a a nominal expression .", "F : And , um ,", "E : Yeah .", "F : I mean , we should have put an example of this", "F : and we could do that later .", "E : Yeah .", "F : But I think the not inherently like the general principles still work though ,", "F : that , um , we can have constructions that have sort of constituent structure in that there is like , you know , for instance , one ", "F : Uh , you know , they they have constituents , right ?", "F : So you can like nest things when you need to ,", "F : but they can also overlap in a sort of flatter way .", "F : So if you don't have like a lot of grammar experience , then like this this might , you know , be a little o opaque .", "F : But , you know , we have the properties of dependency grammars and some properties of constituents constituent - based grammar .", "F : So that 's I think that 's sort of the main thing we wanted to aim for", "E : Mm - hmm .", "F : and so far it 's worked out OK .", "C : Good .", "F : So .", "F : OK .", "A : I can say two things about the f", "F : Yes .", "A : Maybe you want to forget stress .", "A : This my f", "F : As a word ?", "A : No , as as Just don't don't think about it .", "F : As a ", "F : What 's that ?", "F : Sorry .", "A : If ", "A : canonically speaking you can if you look at a a curve over sentence , you can find out where a certain stress is", "A : and say , \" hey , that 's my focus exponent . \"", "F : Mm - hmm .", "E : Right .", "A : It doesn't tell you anything what the focus is .", "F : Mm - hmm .", "A : If it 's just that thing ,", "F : Or the constituent that it falls in .", "A : a little bit more or the whole phrase .", "E : Mm - hmm .", "A : Um ", "F : You mean t forget about stress , the form cue ?", "A : The form bit", "E : Yeah .", "A : because , uh , as a form cue , um , not even trained experts can always well , they can tell you where the focus exponent is sometimes .", "F : OK .", "A : And that 's also mostly true for read speech .", "A : In in real speech , um , people may put stress .", "A : It 's so d context dependent on what was there before , phrase ba breaks , um , restarts .", "F : Yeah .", "F : Mm - hmm .", "A : It 's just , um ", "A : it 's absurd .", "A : It 's complicated .", "F : OK ,", "E : Yeah , I mean , I I 'm sort of inclined to say let 's worry about specifying the information structure focus of the sentence", "A : And all ", "F : I believe you , yeah .", "F : Mm - hmm .", "E : and then ,", "F : Ways that you can get it come from th", "E : hhh , the phonology component can handle actually assigning an intonation contour to that .", "F : right .", "E : You know , I mean , later on we 'll worry about exactly how ", "A : Or or map from the contour to to what the focus exponent is .", "E : y Yeah .", "F : Mm - hmm .", "E : Exactly .", "E : But figure out how the ", "E : Yeah .", "A : But , uh , if you don't know what you 're what you 're focus is then you 're you 're hopeless - uh - ly lost anyways ,", "F : Right .", "F : That 's fine , yeah .", "F : Mm - hmm .", "A : and the only way of figuring out what that is , is , um , by sort of generating all the possible alternatives to each focused element , decide which one in that context makes sense and which one doesn't .", "F : Mm - hmm .", "A : And then you 're left with a couple three .", "F : Mm - hmm .", "A : So , you know , again , that 's something that h humans can do ,", "A : um , but far outside the scope of of any anything .", "A : So .", "A : You know .", "A : It 's ", "F : OK . Well , uh , yeah , I wouldn't have assumed that it 's an easy problem in in absence of all the oth", "A : u u", "F : you need all the other information I guess .", "A : But it 's it 's what it uh , it 's pretty easy to put it in the formalism , though .", "A : I mean , because", "F : Yeah .", "A : you can just say whatever stuff , \" i is the container being focused or the the entire whatever , both , and so forth . \"", "F : Mm - hmm , mm - hmm .", "E : Mm - hmm .", "F : Yeah . Exactly . So the sort of effect of it is something we want to be able to capture .", "C : Yeah , so b b but I think the poi", "C : I 'm not sure I understand", "C : but here 's what I th think is going on .", "C : That if we do the constructions right when a particular construction matches , it the fact that it matches , does in fact specify the focus .", "F : W uh ,", "F : I 'm not sure about that .", "F : Or it might limit it cert certainly constrains the possibilities of focus .", "C : OK .", "C : Uh k uh , at at the very least it constrai", "F : I think that 's that 's , th that 's certainly true .", "F : And depending on the construction it may or may not f specify the focus , right ?", "C : Oh , uh , for sure , yes .", "C : There are constrai yeah , it 's not every but there are constructions , uh , where you t explicitly take into account those considerations", "F : Yeah .", "F : Mm - hmm .", "C : that you need to take into account in order to decide which what is being focused .", "F : Mm - hmm .", "A : Mm - hmm .", "A : So we talked about that a little bit this morning .", "A : \" John is on the bus ,", "F : Mm - hmm .", "A : not Nancy . \"", "F : Hmm .", "A : So that 's focuses on John .", "C : Right .", "A : \" John is on the bus and not on the train . \"", "F : Mm - hmm .", "C : Right .", "A : \" John is on the bus \" versus \" John is on the train . \"", "F : Right .", "A : And \" John is on the bus \" versus \" was \" , and e", "F : Is on .", "F : \" John is on the bus \" .", "A : \" it 's the bu \" so e", "F : Yeah . Yeah .", "C : Right .", "C : Yeah , all all of those .", "C : Yeah .", "F : Right .", "A : All of these", "A : and will we have u", "A : is it all the same constructions ?", "A : Just with a different foc focus constituent ?", "F : Yeah , I would say that argument structure in terms of like the main like sort of ,", "A : Mm - hmm .", "F : I don't know ", "F : the fact that you can get it without any stress and you have some whatever is predicated anyway should be the same set of constructions .", "F : So that 's why I was talking about overlapping constructions .", "F : So , then you have a separate thing that picks out , you know , stress on something relative to everything else .", "C : Yeah .", "E : Mm - hmm .", "C : So , the question is actually ", "F : And it would ", "C : oh , I 'm sorry ,", "F : yeah ,", "C : go ahead ,", "C : finish .", "F : and it w and that would have to uh it might be ambiguous as , uh , whether it picks up that element , or the phrase , or something like that .", "F : But it 's still is limited possibility .", "A : Hmm .", "F : So that should , you know , interact with it should overlap with whatever other construction is there .", "A : Yeah .", "C : S s the question is , do we have a way on the other page , uh , when we get to the s semantic side , of saying what the stressed element was , or stressed phrase , or something .", "F : Mm - hmm .", "F : Well , so that 's why I was saying how ", "F : since I couldn't think of an easy like limited way of doing it , um , all I can say is that information structure has a focused slot", "C : Right .", "F : and I think that should be able to refer to ", "C : So that 's down at the bottom here when we get over there .", "F : Yeah ,", "C : OK .", "F : and , infer and I don't have I don't have a great way or great examples", "C : I 'll - I 'll wait .", "F : but I think that something like that is probably gonna be , uh , more more what we have to do .", "C : OK .", "A : Hmm .", "F : But , um ,", "C : OK .", "A : So", "F : OK , that was one comment .", "F : And you had another one ?", "A : Yeah , well the once you know what the focus is the everything else is background .", "A : How about \" topic - comment \"", "A : that 's the other side of information .", "F : How about what ?", "A : Topic - comment .", "F : Yeah , so that was the other thing .", "F : And so I didn't realize it before .", "F : It 's like , \" oh ! \" It was an epiphany that it you know , topic and focus are a contrast set .", "F : So topic is ", "F : Topic - focused seems to me like , um , background profile , OK ,", "F : or a landmark trajector , or some something like that .", "F : There 's there 's definitely , um , that kind of thing going on .", "A : Mmm .", "F : Now I don't know whether ", "F : I n I don't have as many great examples of like topic - indicating constructions", "F : on like focus , right ?", "F : Um , topic it seems kind of you know , I think that might be an ongoing kind of thing .", "A : Mm - hmm .", "E : Japanese has this though .", "F : Topic marker ?", "E : You know .", "E : Yeah , that 's what \" wa \" is ,", "A : Yeah .", "F : Mm - hmm .", "E : uh , just to mark which thing is the topic .", "F : Mm - hmm .", "E : It doesn't always have to be the subject .", "F : Right .", "F : So again , information structure has a topic slot .", "F : And , you know , I stuck it in thinking that we might use it .", "A : Mm - hmm .", "F : Um , I think I stuck it in .", "C : Yep , it 's there .", "F : Um , and", "F : one thing that I didn't do consistently , um , is when we get there , is like indicate what kind of thing fits into every role .", "F : I think I have an idea of what it should be", "F : but th", "F : you know , so far we 've been getting away with like either a type constraint or , um , you know , whatever .", "F : I forg it 'll be a frame .", "F : You know , it 'll be it 'll be another predication or it 'll be , um , I don't know , some value from from some something , some variable and scope or something like that ,", "F : or a slot chain based on a variable and scope .", "F : OK , so well that 's ", "F : should we flip over to the other side officially then ?", "A : Mm - hmm , hmm .", "F : I keep , uh , like , pointing forward to it .", "E : OK , side one .", "F : Yeah .", "F : Now we 'll go back to s", "F : OK , so this doesn't include something which mi mi may have some effect on on it ,", "F : which is , um , the discourse situation context record ,", "F : right ? So I didn't I I meant just like draw a line", "F : and like , you know , you also have , uh , some tracking of what was going on .", "C : Right .", "F : And sort of this is a big scale comment before I , you know , look into the details of this .", "F : But for instance you could imagine instead of having ", "F : I I changed the name of um it used to be \" entities \" .", "F : So you see it 's \" scenario \" , \" referent \" and \" discourse segment \" .", "F : And \" scenario \" is essentially what kind of what 's the basic predication ,", "F : what event happened .", "F : And actually it 's just a list of various slots from which you would draw draw in order to paint your picture , a bunch of frames , bi and bindings , right ?", "F : Um , and obviously there are other ones that are not included here ,", "F : general cultural frames and general like , uh , other action f", "F : you know , specific X - schema frames .", "E : Mm - hmm .", "F : OK , whatever .", "F : The middle thing used to be \" entities \"", "F : because you could imagine it should be like really a list", "F : where here was various information .", "F : And this is intended to be grammatically specifiable information about a referent uh , you know , about some entity that you were going to talk about .", "F : So \" Harry walked into the room \" ,", "F : \" Harry \" and \" room \" ,", "F : you know , the room ", "F : th but they would be represented in this list somehow .", "F : And it could also have for instance , it has this category slot .", "F : Um , it should be either category or in or instance .", "F : Basically , it could be a pointer to ontology .", "F : So that everything you know about this could be could be drawn in .", "F : But the important things for grammatical purposes are for things like number , gender ,", "F : um ki the ones I included here are slightly arbitrary", "F : but you could imagine that , um , you need to figure out wheth", "F : if it 's a group whether , um , some event is happening ,", "F : linear time , linear spaces ,", "F : like , you know , are are they doing something serially", "F : or is it like ,", "F : um , uh I 'm I 'm not sure .", "F : Because this partly came from , uh , Talmy 's schema", "F : and I 'm not sure we 'll need all of these actually .", "F : But ", "F : Um , and then the \" status \" I used was like , again , in some languages , you know , like for instance in child language you might distinguish between different status .", "F : So , th the the big com and and finally \" discourse segment \" is about sort of speech - act - y information structure - y ,", "F : like utterance - specific kinds of things .", "F : So the comment I was going to make about , um , changing entity the entity 's block to reference is that you can imagine your discourse", "F : like situation context ,", "F : you have a set of entities that you 're sort of referring to .", "F : And you might that might be sort of a general , I don't know , database of all the things in this discourse that you could refer to .", "F : And I changed to \" reference \"", "F : cuz I would say , for a particular utterance you have particular referring expressions in it .", "F : And those are the ones that you get information about that you stick in here .", "F : For instance , I know it 's going to be plural .", "F : I know it 's gonna be feminine or something like that .", "F : And and these could actually just point to , you know , the the ID in my other list of enti active entities , right ?", "F : So , um ,", "F : uh , th there 's there 's all this stuff about discourse status . We 've talked about .", "F : I almost listed \" discourse status \" as a slot where you could say it 's active .", "F : You know , there 's this , um , hierarchy uh there 's a schematization of , you know , things can be active or they can be , um , accessible , inaccessible .", "E : Yeah .", "F : It was the one that , you know , Keith , um , emailed to us once , to some of us ,", "F : not all of us .", "F : And", "F : the thing is that that I noticed that that , um , list was sort of discourse dependent .", "F : It was like in this particular set , s you know , instance , it has been referred to recently or it hasn't been ,", "E : Yeah .", "F : or this is something that 's like in my world knowledge but not active .", "C : This ", "C : Uh yeah , well there there seems to be context properties .", "F : So .", "F : Yeah , they 're contex and for instance , I used to have a location thing there", "C : Yeah .", "F : but actually that 's a property of the situation .", "F : And it 's again , time , you know at cert certain points things are located , you know , near or far from you", "C : Well , uh , uh , this is recursive", "F : and ", "C : cuz until we do the uh , mental space story , we 're not quite sure Th - th", "F : Yeah .", "C : which is fine .", "F : Yeah , yeah .", "C : We 'll just we 'll j", "F : So some of these are , uh ", "C : we just don't know yet .", "F : Right .", "F : So I so for now I thought , well maybe I 'll just have in this list the things that are relevant to this particular utterance , right ?", "F : Everything else here is utterance - specific .", "F : Um , and I left the slot , \" predications \" , open", "F : because you can have , um , things like \" the guy I know from school \" .", "F : Or , you know , like your referring expression might be constrained by certain like unbounded na amounts of prep you know , predications that you might make .", "E : Mm - hmm .", "F : And it 's unclear whether ", "F : I mean , you could just have in your scenario , \" here are some extra few things that are true \" , right ?", "E : Mm - hmm .", "F : And then you could just sort of not have this slot here . Right ?", "F : You 're but but it 's used for identification purposes .", "F : So it 's it 's a little bit different from just saying \" all these things are true from my utterance \" .", "E : Yeah .", "C : Right .", "E : Yeah .", "F : Um .", "E : Right , \" this guy I know from school came for dinner \" does not mean , um , \" there 's a guy , I know him from school ,", "E : and he came over for dinner \" .", "E : That 's not the same effect .", "F : Yeah , it 's a little bit it 's a little bit different .", "F : Right ? So ", "F : Or maybe that 's like a restrictive , non - restrictive ", "F : you know , it 's like it gets into that kind of thing for ", "E : Yeah .", "F : um , but maybe I 'm mixing , you know this is kind of like the final result after parsing the sentence .", "F : So you might imagine that the information you pass to , you know in identifying a particular referent would be , \" oh , some \" you know , \" it 's a guy", "E : Mm - hmm .", "F : and it 's someone I know from school \" .", "F : So maybe that would , you know , be some intermediate structure that you would pass into the disc to the , whatever , construal engine or whatever , discourse context , to find you know , either create this reference ,", "E : Yeah .", "F : in which case it 'd be created here ,", "E : Mm - hmm .", "F : and ", "F : you know , so so you could imagine that this might not ", "F : So , uh , I 'm uncommitted to a couple of these things .", "A : But to make it m precise at least in my mind , uh , it 's not precise .", "F : Um .", "A : So \" house \" is gender neuter ?", "F : Um , it could be in ", "A : In reality", "A : or in ", "C : Semantically .", "F : semantically , yeah . Yeah .", "A : semantically .", "A : So ", "F : So it uh , uh , a table .", "F : You know , a thing that c doesn't have a gender .", "F : So . Uh , it could be that I mean ,", "F : maybe you 'd maybe not all these ", "F : I mean , I wou I would say that I tried to keep slots here that were potentially relevant to most most things .", "A : No , just to make sure that we everybody that 's completely agreed that it it has nothing to do with , uh , form .", "F : Yeah .", "F : OK , that is semantic as opposed to ", "F : Yeah . Yeah . That 's right .", "F : Um .", "F : S so again ", "A : Then \" predications \" makes sense to to have it open for something like , uh , accessibility or not .", "F : Open to various things .", "A : Yeah .", "F : Right .", "F : OK , so . Let 's see .", "F : So maybe having made that big sca sort of like large scale comment , should I just go through each of these slots uh , each of these blocks , um , a little bit ?", "E : Sure .", "F : Um , mostly the top one is sort of image schematic .", "F : And just a note ,", "F : which was that , um ", "F : s so when we actually ha", "F : so for instance , um ,", "F : some of them seem more inherently static , OK ,", "F : like a container or sort of support - ish .", "F : And others are a little bit seemingly inherently dynamic", "F : like \" source , path , goal \" is often thought of that way", "F : or \" force \" , or something like that .", "F : But in actual fact , I think that they 're intended to be sort of neutral with respect to that .", "F : And different X - schemas use them in a way that 's either static or dynamic .", "F : So \" path \" , you could just be talking about the path between this and this .", "E : Mmm .", "F : And you know , \" container \" that you can go in and out .", "F : All of these things .", "F : And so , um ,", "F : I think this came up when , uh , Ben and I were working with the Spaniards , um , the other day ", "F : the \" Spaniettes \" , as we called them ", "F : um , to decide like how you want to split up , like , s image schematic contributions versus , like , X - schematic contributions .", "F : How do you link them up .", "F : And I think again , um , it 's gonna be something in the X - schema that tells you \" is this static or is this dynamic \" .", "F : So we definitely need that sort of aspectual type gives you some of that .", "F : Um , that , you know , is it , uh , a state or is it a change of state , or is it a , um , action of some kind ?", "A : Uh , i i i is there any meaning to when you have sort of parameters behind it and when you don't ?", "F : Uh .", "F : Yeah .", "A : Just means ", "F : Oh , oh ! You mean , in the slot ?", "A : Mm - hmm .", "F : Um , no , it 's like X - sc it 's it 's like I was thinking of type constraints", "F : but X - schema , well it obviously has to be an X - schema .", "F : \" Agent \" , I mean , the the performer of the X - schema ,", "F : that s depends on the X - schema .", "F : You know , and I in general it would probably be , you know ", "E : So the difference is basically whether you thought it was obvious what the possible fillers were .", "F : Yeah , basically .", "A : Mm - hmm .", "E : OK .", "F : Um , \" aspectual type \" probably isn't obvious", "F : but I should have So , I just neglected to stick something in . \" Perspective \" , \" actor \" , \" undergoer \" , \" observer \" , um ,", "B : Mmm .", "F : I think we 've often used \" agent \" , \" patient \" , obser", "E : \" Whee ! \"", "E : That 's that one , right ?", "F : Yeah , exactly . Exactly .", "F : Um , and so one nice thing that , uh , we had talked about is this example of like , if you have a passive construction then one thing it does is ch you know definitely , it is one way to for you to , you know , specifically take the perspective of the undergoing kind of object .", "F : And so then we talked about , you know , whether well , does that specify topic as well ?", "F : Well , maybe there are other things .", "F : You know , now that it 's subject is more like a topic .", "F : And now that , you know ", "F : Anyway . So .", "F : Sorry . I 'm gonna trail off on that one", "F : cuz it 's not that f important right now .", "F : Um ,", "C : N now , for the moment we just need the ability to l l write it down if if somebody figured out what the rules were .", "F : To know how ", "F : Yeah . Yeah . Exactly .", "C : Yeah .", "F : Um ,", "F : some of these other ones ,", "F : let 's see .", "F : So ,", "F : uh , one thing I 'm uncertain about is how polarity interacts .", "F : So polarity , uh , is using for like action did not take place for instance .", "C : Mm - hmm .", "F : So by default it 'll be like \" true \" , I guess ,", "F : you know , if you 're specifying events that did happen .", "F : You could imagine that you skip out this you know , leave off this polarity ,", "F : you know , not don't have it here .", "F : And then have it part of the speech - act in some way .", "C : Mm - hmm .", "F : There 's some negation .", "F : But the reason why I left it in is cuz you might have a change of state , let 's say , where some state holds and then some state doesn't hold ,", "F : and you 're just talking , you know ", "F : if you 're trying to have the nuts and bolts of simulation you need to know that , you know , whatever , the holder doesn't and ", "C : No , I th I think at this lev which is it should be where you have it .", "F : OK , it 's so it 's it 's it 's fine where it is .", "F : So ,", "C : I mean , how you get it may may in will often involve the discourse", "F : OK .", "F : May come from a few places .", "C : but but by the time you 're simulating you sh y you should know that .", "F : Right .", "F : Right .", "E : So , I 'm still just really not clear on what I 'm looking at .", "E : The \" scenario \" box ,", "E : like , what does that look like for an example ?", "F : Yeah .", "E : Like , not all of these things are gonna be here .", "F : Mm - hmm .", "C : Correct .", "E : This is just basically says", "F : It 's a grab bag of ", "E : \" part of what I 'm going to hand you is a whole bunch of s uh , schemas , image , and X - schemas .", "E : Here are some examples of the sorts of things you might have in there \" .", "F : So that 's exactly what it is .", "E : OK .", "F : And for a particular instance which I will , you know , make an example of something , is that you might have an instance of container and path , let 's say , as part of your , you know , \" into \" you know , definition .", "E : Mm - hmm .", "E : Mm - hmm .", "F : So you would eventually have instances filled in with various various values for all the different slots .", "E : Mm - hmm .", "F : And they 're bound up in , you know , their bindings and and and values .", "E : OK .", "C : W it c", "E : Do you have to say about the binding in your ", "E : is there a slot in here for that tells you how the bindings are done ?", "C : No , no , no .", "C : I let 's see , I think we 're we 're not ", "C : I don't think we have it quite right yet .", "E : OK .", "C : So , uh , what this is ,", "C : let 's suppose for the moment it 's complete .", "E : OK .", "C : OK , uh , then this says that when an analysis is finished , the whole analysis is finished , you 'll have as a result , uh , some s resulting s semspec for that utterance in context ,", "E : Mm - hmm .", "C : which is made up entirely of these things and , uh , bindings among them .", "E : Mm - hmm .", "C : And bindings to ontology items .", "E : Mm - hmm .", "C : So that that the who that this is the tool kit under whi out of which you can make a semantic specification .", "E : Mm - hmm .", "C : So that 's A .", "E : OK .", "C : But B , which is more relevant to your life , is this is also the tool kit that is used in the semantic side of constructions .", "E : Mm - hmm .", "C : So this is an that anything you have , in the party line , anything you have as the semantic side of constructions comes , from pieces of this ignoring li", "E : OK .", "C : I mean , in general , you ignore lots of it .", "E : Right .", "C : But it 's got to be pieces of this along with constraints among them .", "E : OK .", "C : Uh , so that the , you know , goal of the , uh uh , \" source , path , goal \" has to be the landmark of the conta you know , the interior of this container .", "E : Mm - hmm .", "C : Or whate whatever .", "E : Yeah .", "C : So those constraints appear in constructions", "E : Mm - hmm .", "C : but pretty much this is the full range of semantic structures available to you .", "E : OK .", "F : Except for \" cause \" , that I forgot .", "F : But anyway , there 's som some kind of causal structure for composite events .", "E : Yeah .", "C : OK , good .", "C : Let 's let 's mark that .", "C : So we need a c", "F : Uh , I mean , so it gets a little funny .", "F : These are all so far these structures , especially from \" path \" and on down , these are sort of relatively familiar , um , image schematic kind of slots .", "F : Now with \" cause \" , uh , the fillers will actually be themselves frames . Right ?", "F : So you 'll say , \" event one causes event B ", "C : Right .", "E : Mm - hmm .", "C : And and and and this this this again may ge our , um and we and and , of course , worlds .", "F : uh , event two \" ,", "F : and ", "F : Yeah . So that 's , uh these are all implicitly one within , uh within one world .", "E : Mm - hmm .", "F : Um ,", "F : even though saying that place takes place , whatever .", "F : Uh , if y if I said \" time \" is , you know , \" past \" , that would say \" set that this world \" , you know , \" somewhere , before the world that corresponds to our current speech time \" .", "E : Mm - hmm .", "E : Mm - hmm .", "E : Yeah .", "F : So . But that that that 's sort of OK .", "F : The the within the event it 's st it 's still one world .", "F : Um .", "F : Yeah , so \" cause \" and ", "F : Other frames that could come in I mean , unfortunately you could bring in say for instance , um , uh , \" desire \" or something like that ,", "E : Mm - hmm .", "F : like \" want \" .", "F : And actually there is right now under \" discourse segments \" , um , \" attitude \" ?", "E : Mm - hmm .", "F : \" Volition \" ?", "F : could fill that .", "F : So there are a couple things where I like , \" oh , I 'm not sure if I wanted to have it there", "E : Well that 's ", "F : or \" Basically there was a whole list of of possible speaker attitudes that like say Talmy listed .", "F : And , like , well , I don't you know , it was like \" hope , wish . desire \" ,", "C : Right .", "E : Uh - huh .", "F : blah - blah - blah .", "F : And it 's like , well , I feel like if I wanted to have an extra meaning ", "F : I don't know if those are grammatically marked in the first place .", "F : So They 're more lexically marked , right ?", "F : At least in English .", "E : Mmm .", "F : So if I wanted to I would stick in an extra frame in my meaning , saying ,", "F : e so th it 'd be a hierarchical frame them , right ?", "F : You know , like \" Naomi wants wants su a certain situation", "F : and that situation itself is a state of affairs \" .", "C : S right .", "C : So so , \" want \" itself can be i i i i i", "F : u", "F : Can be just another frame that 's part of your ", "C : Well , and it i basically it 's an action .", "C : In in our s in our in our ", "F : Yeah .", "F : Situation . Right , right .", "C : in in our in our s terminology , \" want \" can be an action", "C : and \" what you want \" is a world .", "F : Mm - hmm .", "B : Hmm .", "F : Mmm .", "C : So that 's I mean , it 's certainly one way to do it .", "E : Mm - hmm .", "C : Yeah , there there are other things .", "C : Causal stuff we absolutely need .", "C : Mental space we need .", "F : Mm - hmm .", "C : The context we need .", "C : Um , so anyway , Keith ", "C : So is this comfortable to you", "C : that , uh , once we have this defined , it is your tool kit for building the semantic part of constructions .", "E : Mm - hmm .", "C : And then when we combine constructions semantically , the goal is going to be to fill out more and more of the bindings needed in order to come up with the final one .", "E : Mm - hmm .", "E : Yeah .", "C : And that 's the wh and and I mean , that according to the party line , that 's the whole story .", "E : Mm - hmm .", "E : Yeah .", "E : Um . y", "E : Right . That makes sense .", "E : So I mean , there 's this stuff in the off in the scenario ,", "E : which just tells you how various what schemas you 're using", "E : and they 're how they 're bound together .", "E : And I guess that some of the discourse segment stuff ", "F : Mm - hmm .", "E : is that where you would sa", "E : I mean , that 's ", "E : OK , that 's where the information structure is", "E : which sort of is a kind of profiling on different parts of , um , of this .", "F : Right .", "F : Exactly .", "E : I mean , what 's interesting is that the information structure stuff ", "E : Hmm .", "E : There 's almost I mean , we keep coming back to how focus is like this this , uh , trajector - landmark thing .", "F : Yeah .", "E : So if I say , um , You know , \" In France it 's like this \" .", "E : You know , great ,", "E : we 've learned something about France", "E : but the fact is that utterances of that sort are generally used to help you draw a conclusion also about some implicit contrast ,", "E : like \" In France it 's like this \" .", "F : Right .", "E : And therefore you 're supposed to say , \" Boy , life sure \"", "E : You know , \" in France kids are allowed to drink at age three \" .", "E : And w you 're that 's not just a fact about France .", "F : Right , right .", "E : You also conclude something about how boring it is here in the U S . Right ?", "C : Right .", "E : And so ", "F : S so I would prefer not to worry about that for right now", "E : OK .", "F : and to think that there are , um ,", "E : That comes in", "F : discourse level constructions in a sense , topic topic - focus constructions that would say , \" oh , when you focus something \" then ", "E : and , uh ", "E : Mm - hmm .", "E : Yeah .", "F : just done the same way just actually in the same way as the lower level .", "F : If you stressed , you know , \" John went to the \" , you know , \" the bar \" whatever , you 're focusing that", "E : Mm - hmm .", "F : and a in a possible inference is \" in contrast to other things \" .", "E : Yeah .", "F : So similarly for a whole sentence , you know , \" in France such - and - such happens \" .", "E : Yeah .", "E : Yeah , yeah .", "F : So the whole thing is sort of like again implicitly as opposed to other things that are possible .", "E : Yeah .", "A : Uh , just just , uh , look read uh even sem semi formal Mats Rooth .", "F : I mean ", "F : Yeah .", "F : Uh - huh .", "A : If you haven't read it . It 's nice .", "A : And just pick any paper on alternative semantics .", "F : Uh - huh .", "E : OK .", "A : So that 's his that 's the best way of talking about focus , is I think his way .", "E : OK , what was the name ?", "A : Mats .", "A : MATS .", "E : OK .", "A : Rooth .", "A : I think two O 's ,", "A : yes , TH .", "E : OK .", "A : I never know how to pronounce his name", "A : because he 's sort of ,", "C : S Swede ?", "A : uh , he is Dutch", "C : Dutch ?", "E : Yeah .", "C : Oh , Dutch .", "A : and , um but very confused background I think .", "C : Uh - huh .", "A : So and , um ,", "E : Mats Gould .", "A : And sadly enough he also just left the IMS in Stuttgart .", "A : So he 's not there anymore .", "E : Hmm .", "A : But , um ", "A : I don't know where he is right now", "A : but alternative semantics is if you type that into an , uh , uh , browser or search engine you 'll get tons of stuff .", "E : OK .", "E : OK .", "E : OK , thanks .", "A : And what I 'm kind of confused about is is what the speaker and the hearer is is sort of doing there .", "F : So for a particular segment it 's really just a reference to some other entity again in the situation , right ?", "F : So for a particular segment the speaker might be you or might be me .", "A : Yeah .", "F : Um , hearer is a little bit harder .", "F : It could be like multiple people .", "F : I guess that that that that 's not very clear from here ", "F : I mean , that 's not allowed here .", "A : Yeah , but you Don't we ultimately want to handle that analogously to the way we handle time and place ,", "A : because \" you \" , \" me \" , \" he \" , \" they \" ,", "A : you know , \" these guys \" ,", "A : all these expressions , nuh , are in in much the same way contextually dependent as \" here , \" and \" now , \" and \" there \" ", "F : Mm - hmm .", "C : Now , this is this is assuming you 've already solved that .", "F : Ye - yeah .", "C : So it 's it 's Fred and Mary ,", "F : So th", "C : so the speaker would be Fred", "C : and the ", "A : Ah !", "F : Right , so the constructions might of course will refer , using pronouns or whatever .", "A : Mm - hmm .", "F : In which case they have to check to see , uh , who the , uh , speaker in here wa in order to resolve those .", "F : But when you actually say that \" he walked into \" , whatever , um , the \" he \" will refer to a particular ", "F : You you will already have figured who \" he \" or \" you \" , mmm , or \" I \" , maybe is a bett better example , who \" I \" refers to .", "F : Um , and then you 'd just be able to refer to Harry , you know , in wherever that person whatever role that person was playing in the event .", "A : Mmm .", "A : That 's up at the reference part .", "F : Yeah , yeah .", "A : And down there in the speaker - hearer part ?", "F : S so , that 's I think that 's just ", "F : n", "F : for instance , Speaker is known from the situation , right ?", "F : You 're when you hear something you 're told who the speaker is ", "F : I mean , you know who the speaker is .", "F : In fact , that 's kind of constraining how ", "F : in some ways you know this before you get to the you fill in all the rest of it .", "F : I think .", "C : Mmm .", "F : I mean , how else would you um ", "A : You know , uh , uh , it 's the speaker may in English is allowed to say \" I . \"", "C : Yeah .", "C : Well , here ", "F : Yeah .", "A : Uh , among the twenty - five percent most used words .", "F : Right .", "A : But wouldn't the \" I \" then set up the the s s referent ", "F : Mm - hmm .", "A : that happens to be the speaker this time", "F : Right , right .", "A : and not \" they , \" whoever they are .", "F : So ", "A : Or \" you \" ", "A : much like the \" you \" could n", "F : S so OK , so I would say ref under referent should be something that corresponds to \" I \" .", "F : And maybe each referent should probably have a list of way whatever , the way it was referred to .", "F : So that 's \" I \"", "F : but , uh , uh , should we say it it refers to , what ?", "F : Uh , if it were \" Harry \" it would refer to like some ontology thing .", "F : If it were if it 's \" I \" it would refer to the current speaker ,", "F : OK , which is given to be like , you know , whoever it is .", "A : Well , not not always .", "A : I mean , so there 's \" and then he said , I w \" Uh - huh .", "F : \" I \" within the current world .", "C : Uh ", "A : Yeah .", "C : Yeah .", "C : That 's right .", "F : Yeah , yeah , yeah , yeah .", "C : So so again , this uh , this this is gonna to get us into the mental space stuff", "C : and t because you know , \" Fred said that Mary said \" , and whatever .", "F : Mm - hmm .", "E : Mmm .", "C : And and so we 're , uh gonna have to , um , chain those as well .", "A : Mm - hmm .", "A : Twhhh - whhh .", "A : But ", "F : Mm - hmm .", "F : So this entire thing is inside a world ,", "C : Right .", "F : not just like the top part .", "C : Right .", "C : I I think , uh ", "F : That 's ", "A : Mm - hmm .", "C : Except s it 's it 's trickier than that", "C : because um , the reference for example ", "C : So he where it gets really tricky is there 's some things ,", "F : Yeah .", "C : and this is where blends and all terribl", "F : Yeah .", "C : So , some things which really are meant to be identified and some things which aren't .", "F : Right .", "C : And again , all we need for the moment is some way to say that .", "F : Right .", "F : So I thought of having like for each referent , having the list of of the things t with which it is identified .", "F : You know , which which , uh you know , you you you ", "C : You could do that .", "F : for instance , um ", "F : So , I guess ,", "F : it sort of depends on if it is a referring exp if it 's identifiable already or it 's a new thing .", "E : Mm - hmm .", "F : If it 's a new thing you 'd have to like create a structure or whatever .", "F : If it 's an old thing it could be referring to , um , usually w something in a situation , right ?", "F : Or something in ontology .", "C : uh - huh .", "F : So ,", "F : there 's a you know , whatever , it c it could point at one of these .", "C : I just had a I just had an an idea that would be very nice if it works .", "F : For what ?", "F : If it works .", "C : Uh , uh , uh , I haven't told you what it is yet .", "F : Mm - hmm . Mmm .", "C : This was my build - up .", "C : An i an idea that would be nice i", "F : Yeah .", "F : OK , we 're crossing our fingers .", "C : Right .", "C : If it worked .", "B : So we 're building a mental space , good .", "C : Yeah .", "F : OK .", "C : Right , it was a space builder .", "C : Um , we might be able to handle context in the same way that we handle mental spaces", "C : because , uh , you have somewhat the same things going on of , uh , things being accessible or not .", "F : Mm - hmm .", "C : And so , i", "F : Yep .", "C : it c it it , uh I think if we did it right we might be able to get at least a lot of the same structure .", "F : Use the same Yep .", "C : So that pulling something out of a discourse context is I think similar to other kinds of , uh , mental space phenomena .", "B : I see .", "F : Mm - hmm .", "F : And ", "C : Uh , I 've I 've I 've never seen anybody write that up", "F : And ", "C : but maybe they did .", "C : I don't know .", "C : That may be all over the literature .", "F : Yeah .", "F : So so by default ", "E : There 's things like ther you know , there 's all kinds of stuff like , um , in I think I mentioned last time in Czech if you have a a verb of saying then", "E : um , you know , you say something like or or I was thinking you can say something like , \" oh , I thought , uh , you are a republican \" or something like that .", "E : Where as in English you would say , \" I thought you were \" .", "C : Right .", "E : Um , you know , sort of the past tense being copied onto the lower verb doesn't happen there ,", "E : so you have to say something about , you know , tense is determined relative to current blah - blah - blah .", "F : Mm - hmm .", "F : Mm - hmm .", "E : Same things happens with pronouns .", "E : There 's languages where , um ,", "E : if you have a verb of saying then ,", "E : ehhh ,", "E : where ", "E : OK , so a situation like \" Bob said he was going to the movies \" , where that lower subject is the same as the person who was saying or thinking , you 're actually required to have \" I \" there .", "F : Mm - hmm .", "E : Um , and it 's sort of in an extended function ", "C : Mm - hmm .", "C : So we would have it be in quotes in English .", "E : Yeah .", "F : Right .", "B : Right .", "E : But it 's not perceived as a quotative construction .", "E : I mean , it 's been analyzed by the formalists as being a logophoric pronoun ,", "C : Yeah .", "E : um which means a pronoun which refers back to the person who is speaking or that sort of thing , right ?", "F : Oh , right . Yeah , that makes sense .", "C : OK .", "E : Um ,", "E : but uh , that happens to sound like the word for \" I \"", "E : but is actually semantically unrelated to it .", "F : Oh , no !", "C : Oh , good , I love the formali", "E : Um ,", "F : Really ?", "E : Yeah . Yeah .", "F : You 're kidding .", "E : There 's a whole book which basically operates on this assumption .", "E : Uh , Mary Dalrymple ,", "E : uh , this book , a ninety - three book on , uh on pronoun stuff .", "F : No , that 's horrible .", "F : OK .", "F : That 's horrible . OK .", "E : Well , yeah . And then the same thing for ASL", "E : where , you know , you 're signing", "E : and someone says something .", "E : And then , you know , so \" he say \" , and then you sort of do a role shift .", "E : And then you sign \" I , this , that , and the other \" .", "F : Uh - huh .", "E : And you know , \" I did this \" .", "E : That 's also been analyzed as logophoric and having nothing to do with \" I \" .", "E : And the role shift thing is completely left out and so on .", "E : So ,", "E : I mean , the point is that pronoun references , uh , you know , sort of ties in with all this mental space stuff and so on , and so forth .", "F : Uh - huh .", "E : And so , yeah , I mean ", "F : Yeah .", "C : So that that d that does sound like it 's co consistent with what we 're saying , yeah .", "E : Right .", "E : Yeah .", "F : OK , so it 's kind of like the unspecified mental spaces just are occurring in context .", "F : And then when you embed them sometimes you have to pop up to the h you know , depending on the construction or the whatever , um , you you you 're scope is m might extend out to the the base one .", "E : Mm - hmm .", "C : Mm - hmm .", "E : Yeah .", "F : It would be nice to actually use the same , um , mechanism", "F : since there are so many cases where you actually need", "F : it 'll be one or the other .", "E : Yeah .", "F : It 's like , oh , actually , it 's the same same operation .", "C : Oh , OK , so this this is worth some thought .", "F : So .", "E : It 's like it 's like what 's happening that , yeah , what what 's happening , uh , there is that you 're moving the base space or something like that , right ?", "F : Yeah , yeah .", "E : So that 's that 's how Fauconnier would talk about it .", "F : Mm - hmm .", "E : And it happens diff under different circumstances in different languages .", "E : And so ,", "F : Mm - hmm .", "E : um , things like pronoun reference and tense which we 're thinking of as being these discourse - y things actually are relative to a Bayes space", "E : which can change .", "F : Mm - hmm ,", "E : And we need all the same machinery .", "F : right .", "A : Mm - hmm .", "F : Robert .", "C : Well , but , uh , this is very good actually", "E : Schade .", "C : cuz it it it to the extent that it works , it y", "F : Ties it all into it .", "C : it it ties together several of of these things .", "F : Yeah .", "F : Yep .", "A : Mm - hmm .", "A : Mm - hmm .", "A : And I 'm sure gonna read the transcript of this one .", "A : So .", "A : But the , uh , But it 's too bad that we don't have a camera .", "A : You know , all the pointing is gonna be lost .", "F : Oh , yeah .", "E : Yeah .", "F : Yeah , that 's why I said \" point to Robert \" , when I did it .", "B : Well every time Nancy giggles it means it means that it 's your job .", "A : Uh .", "A : Yeah .", "A : Mmm , isn't I mean , I 'm I was sort of dubious why why he even introduces this sort of reality , you know , as your basic mental space", "A : and then builds up ", "E : Mm - hmm .", "A : d doesn't start with some because it 's so obvi it should be so obvious , at least it is to me , that whenever I say something I could preface that with \" I think . \"", "E : Yeah .", "A : Nuh ?", "A : So there should be no categorical difference between your base and all the others that ensue .", "E : Yeah .", "C : No , but there 's there 's a Gricean thing going on there ,", "C : that when you say \" I think \" you 're actually hedging .", "F : Mmm .", "E : Yeah , I mean ", "F : It 's like I don't totally think ", "E : Yeah .", "C : Right .", "F : I mostly think , uh ", "E : Y", "A : Yeah , it 's Absolutely .", "E : Yeah , it 's an it 's an evidential .", "E : It 's sort of semi - grammaticalized .", "E : People have talked about it this way .", "E : And you know , you can do sort of special things .", "E : You can , th", "E : put just the phrase \" I think \" as a parenthetical in the middle of a sentence and so on , and so forth .", "F : Actually one of the child language researchers who works with T Tomasello studied a bunch of these constructions", "A : Yeah .", "E : So ", "F : and it was like it 's not using any kind of interesting embedded ways", "F : just to mark , you know , uncertainty or something like that .", "E : Yeah .", "F : So .", "A : Yeah , but about linguistic hedges ,", "A : I mean , those those tend to be , um , funky anyways", "A : because they blur ", "C : So we don't have that in here either do we ?", "E : Yeah .", "F : Hedges ?", "C : Yeah , yeah .", "F : Hhh , I there used to be a slot for speaker , um ,", "F : it was something like factivity .", "F : I couldn't really remember what it meant", "F : so I took it out .", "E : Yeah .", "F : But it 's something ", "E : Um .", "E : Well we were just talking about this sort of evidentiality and stuff like that , right ?", "F : we we were talking about sarcasm too , right ?", "F : Oh , oh .", "E : I mean ,", "F : Oh , yeah , yeah , right .", "E : that 's what I think is ,", "E : um , sort of telling you what percent reality you should give this", "C : So we probably should .", "F : Yeah .", "E : or the , you know ", "A : Mm - hmm .", "C : Confidence or something like that .", "E : Yeah ,", "E : and the fact that I 'm , you know the fact maybe if I think it versus he thinks that might , you know , depending on how much you trust the two of us or whatever ,", "F : Yeah .", "E : you know ", "A : Uh great word in the English language is called \" about \" .", "A : If you study how people use that it 's also ", "F : What 's the word ?", "A : \" about . \"", "C : About .", "A : It 's about ", "C : Oh , that in that use of \" about \" , yeah .", "A : clever .", "F : Oh , oh , oh , as a hedge .", "E : Yeah .", "C : And I think And I think y if you want us to spend a pleasant six or seven hours you could get George started on that .", "E : He wrote a paper about thirty - five years ago on that one .", "B : I r I read that paper ,", "C : Yeah .", "B : the hedges paper ?", "E : Yeah .", "B : I read some of that paper actually .", "E : Would you believe that that paper lead directly to the development of anti - lock brakes ?", "C : Yeah .", "F : What ?", "C : No .", "E : Ask me about it later", "E : I 'll tell you how .", "E : When we 're not on tape .", "F : I 'd love to know .", "B : Oh , man .", "F : So , and and I think , uh , someone had raised like sarcasm as a complication at some point .", "C : There 's all that stuff .", "C : Yeah , let 's I I don't I think ", "F : And", "F : we just won't deal with sarcastic people .", "C : Yeah , I mean ", "E : I don't really know what like ", "E : We we don't have to care too much about the speaker attitude , right ?", "E : Like there 's not so many different ", "F : Certainly not as some ", "E : hhh , I don't know , m", "F : Well , they 're intonational markers I think for the most part .", "E : Yeah .", "F : I don't know too much about the like grammatical ", "E : I just mean ", "E : There 's lots of different attitudes that that the speaker could have and that we can clearly identify , and so on , and so forth .", "F : Yeah .", "E : But like what are the distinctions among those that we actually care about for our current purposes ?", "C : Right .", "C : Right , so , uh , this this raises the question of what are our current purposes .", "F : Mm - hmm .", "C : Right ?", "F : Oh , yeah , do we have any ?", "E : Oh , shoot .", "E : Here it is three - fifteen already .", "A : Mmm .", "A : Yeah .", "C : Uh , so , um ,", "C : I I don't know the answer", "C : but but , um , it does seem that ,", "C : you know , this is this is coming along .", "C : I think it 's it 's converging .", "C : It 's as far as I can tell there 's this one major thing we have to do", "C : which is the mental the whole s mental space thing .", "C : And then there 's some other minor things .", "F : Mm - hmm .", "C : Um , and we 're going to have to s sort of bound the complexity .", "C : I mean , if we get everything that anybody ever thought about you know , w we 'll go nuts .", "E : Yeah .", "C : So we had started with the idea that the actual , uh , constraint was related to this tourist domain", "C : and the kinds of interactions that might occur in the tourist domain ,", "C : assuming that people were being helpful and weren't trying to d", "C : you know , there 's all sorts of God knows , irony , and stuff like ", "E : Yeah .", "C : which you isn't probably of much use in dealing with a tourist guide .", "E : Yeah .", "C : Yeah ?", "C : Uh .", "F : M mockery .", "C : Right .", "C : Whatever .", "C : So", "C : y uh , no end of things th that that , you know , we don't deal with .", "A : But it ", "C : And ", "C : Go ahead .", "A : i isn't that part easy though", "A : because in terms of the s simspec , it would just mean you put one more set of brack brackets around it ,", "A : and then just tell it to sort of negate whatever the content of that is in terms of irony", "E : Yeah .", "C : N no .", "F : Mmm .", "A : or ", "E : Right .", "C : No .", "F : Maybe .", "F : Yeah , in model theory cuz the semantics is always like \" speaker believes not - P \" ,", "C : No .", "F : you know ?", "F : Like \" the speaker says P and believes not - P \" .", "C : Right .", "E : We have a theoretical model of sarcasm now .", "F : But ", "E : Yeah , right , I mean .", "F : Right , right , but ,", "C : Right .", "C : No , no .", "C : Anyway , so so , um , I guess", "C : uh , let me make a proposal on how to proceed on that ,", "C : which is that , um ,", "C : it was Keith 's , uh , sort of job over the summer to come up with this set of constructions .", "C : Uh , and my suggestion to Keith is that you , over the next couple weeks , n", "E : Mmm .", "C : don't try to do them in detail or formally", "C : but just try to describe which ones you think we ought to have .", "E : OK .", "C : Uh , and then when Robert gets back we 'll look at the set of them .", "E : OK .", "C : Just just sort of , you know , define your space .", "E : Yeah , OK .", "C : And , um , so th these are this is a set of things that I think we ought to deal with .", "E : Yeah .", "C : And then we 'll we 'll we 'll go back over it", "C : and w people will will give feedback on it .", "E : OK .", "C : And then then we 'll have a at least initial spec of of what we 're actually trying to do .", "E : Yeah .", "C : And that 'll also be useful for anybody who 's trying to write a parser .", "E : Mm - hmm .", "C : Knowing uh ", "E : In case there 's any around .", "F : If we knew anybody like that .", "C : Right ,", "C : \" who might want \"", "C : et cetera .", "C : So , uh ", "E : OK .", "C : So a and we get this this , uh , portals fixed", "C : and then we have an idea of the sort of initial range .", "C : And then of course Nancy you 're gonna have to , uh , do your set of ", "C : but you have to do that anyway .", "F : For the same , yeah , data . Yeah , mm - hmm .", "C : So so we 're gonna get the w we 're basically dealing with two domains ,", "C : the tourist domain and the and the child language learning .", "B : Mmm .", "C : And we 'll see what we need for those two .", "C : And then my proposal would be to , um , not totally cut off more general discussion", "C : but to focus really detailed work on the subset of things that we 've we really want to get done .", "E : Mm - hmm .", "E : Mm - hmm .", "C : And then as a kind of separate thread , think about the more general things and and all that .", "E : Mm - hmm .", "A : Well , I also think the detailed discussion will hit you know , bring us to problems that are of a general nature", "C : Uh , without doubt .", "A : and maybe even ", "F : Yeah .", "C : Yeah .", "C : But what I want to do is is is to to constrain the things that we really feel responsible for .", "A : even suggest some solutions .", "A : Yeah .", "A : Mmm .", "E : Mm - hmm .", "C : So that that we say these are the things we 're really gonna try do by the end of the summer", "C : and other things we 'll put on a list of of research problems or something ,", "C : because you can easily get to the point where nothing gets done", "C : because every time you start to do something you say , \" oh , yeah , but what about this case ? \"", "E : Mm - hmm .", "C : This is this is called being a linguist .", "A : Mmm .", "E : Yeah .", "C : And , uh ,", "E : Basically .", "F : Or me .", "C : Huh ?", "F : Or me .", "F : Anyways ", "B : There 's that quote in Jurafsky and Martin where where it goes where some guy goes , \" every time I fire a linguist the performance of the recognizer goes up . \"", "C : Right .", "F : Yeah .", "E : Exactly .", "C : Right .", "C : But anyway .", "C : So , is is that does that make sense as a , uh a general way to proceed ?", "F : Sure , yeah .", "E : Yeah , yeah , we 'll start with that ,", "E : just figuring out what needs to be done", "E : then actually the next step is to start trying to do it .", "C : Exactly right .", "E : Got it .", "A : Mmm .", "A : Mmm .", "E : OK .", "A : We have a little bit of news , uh , just minor stuff .", "A : The one big ", "B : Ooo , can I ask a ", "E : You ran out of power .", "A : Huh ?", "B : Can I ask a quick question about this side ?", "F : Yes .", "A : Yeah .", "B : Is this , uh was it intentional to leave off things like \" inherits \"", "B : and ", "F : Oops .", "F : Um ,", "E : No .", "F : not really ", "F : just on the constructions , right ?", "B : Yeah ,", "F : Um ,", "B : like constructions can inherit from other things ,", "B : am I right ?", "F : yeah .", "B : Yeah .", "F : I didn't want to think too much about that for for now .", "B : OK .", "F : So , uh , maybe it was subconsciously intentional .", "C : Yeah .", "C : Yeah , uh ", "E : Um , yeah , there should be I I wanted to s find out someday if there was gonna be some way of dealing with , uh , if this is the right term , multiple inheritance ,", "C : yeah .", "E : where one construction is inheriting from , uh from both parents ,", "C : Mm - hmm .", "F : Uh - huh .", "F : Yep .", "E : uh , or different ones ,", "E : or three or four different ones .", "C : Yeah .", "F : Yeah .", "E : Cuz the problem is that then you have to ", "C : So let me ", "F : Refer to them .", "E : which of you know , which are how they 're getting bound together .", "C : Yeah , right , right , right .", "F : Yeah , and and there are certainly cases like that .", "C : Yeah , yeah , yeah .", "F : Even with just semantic schemas we have some examples .", "C : Right .", "F : So ,", "F : and we 've been talking a little bit about that anyway .", "C : Yeah . So what I would like to do is separate that problem out .", "F : Inherits .", "C : So um ,", "E : OK .", "C : my argument is there 's nothing you can do with that that you can't do by just having more constructions .", "E : Yeah , yes .", "C : It 's uglier and it d doesn't have the deep linguistic insights and stuff .", "E : That 's right .", "C : Uh ,", "E : But whatever .", "F : Uh , those are over rated .", "E : Yeah , no , no , no no .", "C : Right .", "E : No , by all means ,", "E : right . Uh , sure .", "C : And so I what I 'd like to do is is in the short run focus on getting it right .", "C : And when we think we have it right then saying , \" aha ! ,", "E : Yeah .", "C : can we make it more elegant ? \"", "E : Yeah , that 's ", "E : Yeah .", "C : Can can we , uh What are the generalizations , and stuff ?", "E : Connect the dots .", "E : Yeah .", "C : But rather than try to guess a inheritance structure and all that sort of stuff before we know what we 're doing .", "E : Yep .", "E : Yeah .", "C : So I would say in the short run we 're not gonna b", "E : Yeah .", "C : First of all , we 're not doing them yet at all .", "C : And and it could be that half way through we say , \" aha ! ,", "C : we we now see how we want to clean it up . \"", "E : Mm - hmm .", "C : Uh , and inheritance is only one I mean , that 's one way to organize it", "C : but there are others .", "E : Yeah .", "C : And it may or may not be the best way .", "C : I 'm sorry ,", "C : you had news .", "A : Mmm .", "A : Oh , just small stuff .", "A : Um , thanks to Eva on our web site we can now ,", "A : if you want to run JavaBayes , uh , you could see get download these classes .", "A : And then it will enable you ", "A : she modified the GUI so it has now a m a m a button menu item for saving it into the embedded JavaBayes format .", "D : Mm - hmm .", "B : Mmm .", "A : So that 's wonderful .", "C : Great .", "A : And , um and she , a You tested it out .", "A : Do you want to say something about that , that it works , right ?", "D : I was just checking like , when we wanna , um , get the posterior probability of , like , variables .", "A : With the ", "D : You know how you asked whether we can , like , just observe all the variables like in the same list ?", "D : You can't .", "D : You have to make separate queries every time .", "A : Uh - huh .", "A : OK , that 's that 's a bit unfortunate", "D : So ", "D : Yeah .", "A : but", "A : for the time being it 's it 's it 's fine to do it ", "D : You just have to have a long list of , you know , all the variables .", "A : Yeah .", "D : Basically .", "A : But uh ", "F : Uh , all the things you want to query , you just have to like ask for separately .", "D : Yeah , yeah .", "A : Well that 's probably maybe in the long term that 's good news", "A : because it forces us to think a little bit more carefully how how we want to get an out output .", "A : Um , but that 's a different discussion for a different time .", "A : And , um ,", "A : I don't know .", "A : We 're really running late ,", "A : so I had , uh , an idea yesterday", "A : but , uh , I don't know whether we should even start discussing .", "C : W what ", "C : Yeah , sure , tell us what it is .", "A : Um , the construal bit that , um , has been pointed to but hasn't been , um , made precise by any means , um , may w may work as follows .", "A : I thought that we would , uh ", "A : that the following thing would be in incredibly nice", "A : and I have no clue whether it will work at all or nothing .", "A : So that 's just a tangent , a couple of mental disclaimers here .", "A : Um , imagine you you write a Bayes - net ,", "A : um ", "F : Bayes ?", "A : Bayes - net ,", "F : OK .", "A : um , completely from scratch every time you do construal .", "A : So you have nothing .", "A : Just a white piece of paper .", "C : Mmm ,", "C : right .", "A : You consult consult your ontology", "A : which will tell you a bunch of stuff , and parts , and properties , uh - uh - uh", "F : Grout out the things that that you need .", "C : Right .", "A : then y you 'd simply write , uh , these into onto your your white piece of paper .", "A : And you will get a lot of notes and stuff out of there .", "A : You won't get you won't really get any C P", "A : T 's , therefore we need everything that that configures to what the situation is ,", "A : IE , the context dependent stuff .", "A : So you get whatever comes from discourse", "A : but also filtered .", "A : Uh , so only the ontology relevant stuff from the discourse", "F : Mm - hmm .", "A : plus the situation and the user model .", "A : And that fills in your CPT 's with which you can then query , um , the the net that you just wrote", "A : and find out how thing X is construed as an utterance U .", "A : And the embedded JavaBayes works exactly like that ,", "A : that once you ", "A : we have , you know , precise format in which to write it ,", "A : so we write it down .", "A : You query it .", "A : You get the result ,", "A : and you throw it away .", "A : And the the nice thing about this idea is that you don't ever have to sit down and think about it or write about it .", "A : You may have some general rules as to how things can be can be construed as what , so that will allow you to craft the the the initial notes .", "A : But it 's in that respect it 's completely scalable .", "A : Because it doesn't have any prior , um , configuration .", "A : It 's just you need an ontology of the domain", "A : and you need the context dependent modules .", "A : And if this can be made to work at all , that 'd be kind of funky .", "C : Um , it sounds to me like you want P R", "A : P R Ms - uh , PRM", "A : I mean , since you can unfold a PRM into a straightforward Bayes - net ", "C : Beca - because it b because ", "C : No , no , you can't .", "C : See the the critical thing about the PRM is it gives these relations in general form .", "C : So once you have instantiated the PRM with the instances and ther then you can then you can unfold it .", "A : Then you can .", "A : Mm - hmm , yeah .", "A : No , I was m using it generic .", "A : So , uh , probabilistic , whatever , relational models .", "A : Whatever you write it .", "C : Well , no , but it matters a lot", "A : In ", "C : because you what you want are these generalized rules", "C : about the way things relate ,", "C : th that you then instantiate in each case .", "A : And then then instantiate them .", "C : Yeah , and that 's ", "A : That 's ma maybe the the way the only way it works .", "C : Yeah , that 's the only way it could work .", "C : I we have a our local expert on P R", "C : uh , but my guess is that they 're not currently good enough to do that .", "C : But we 'll we 'll have to see .", "C : Uh ", "A : But , uh ,", "C : Yes .", "C : This is that 's that would be a good thing to try .", "C : It 's related to the Hobbs abduction story", "C : in that you th you throw everything into a pot", "C : and you try to come up with the , uh ", "A : Except there 's no no theorem prover involved .", "F : Best explanation .", "C : No , there isn't a theorem prover", "C : but there but but the , um , The cove the the P R Ms are like rules of inference", "C : and you 're you 're coupling a bunch of them together .", "A : Mm - hmm , yeah .", "C : And then ins instead of proving you 're trying to , you know , compute the most likely .", "C : Uh ", "C : Tricky .", "C : But you yeah , it 's a good it 's a it 's a good thing to put in your thesis proposal .", "A : What 's it ?", "C : So are you gonna write something for us before you go ?", "A : Yes .", "A : Um .", "C : Oh , you have something .", "A : In the process thereof , or whatever .", "C : OK .", "C : So , what 's what when are we gonna meet again ?", "F : When are you leaving ?", "F : Thursday ,", "A : Fri - uh ,", "F : Friday ?", "D : Fri", "A : Thursday 's my last day here .", "F : OK .", "C : Yeah .", "A : So ", "A : I would suggest as soon as possible .", "A : Do you mean by we , the whole ben gang ?", "C : N no , I didn't mean y just the two of us .", "C : We obviously we can we can do this .", "C : But the question is do you want to , for example , send the little group , uh , a draft of your thesis proposal", "C : and get , uh , another session on feedback on that ?", "C : Or ", "A : We can do it Th - Thursday again .", "A : Yeah .", "E : Fine with me .", "E : Should we do the one PM time", "E : for Thursday since we were on that before", "E : or ?", "A : Sure .", "E : OK .", "C : Alright .", "D : Hmm .", "A : Thursday at one ?", "A : I can also maybe then sort of run through the , uh the talk I have to give at EML", "A : which highlights all of our work .", "C : OK .", "A : And we can make some last minute changes on that .", "C : OK .", "B : You can just give him the abstract that we wrote for the paper .", "C : That - that 'll tell him exactly what 's going on .", "C : Yeah , that ", "F : Can we do can we do one - thirty ?", "C : Alright .", "A : No .", "F : Oh , you already told me no .", "F : One ,", "A : But we can do four .", "F : OK , it 's fine .", "F : I can do one .", "F : It 's fine .", "F : It 's fine .", "A : One or four .", "A : I don't care .", "E : To me this is equal .", "E : I don't care .", "A : If it 's equal for all ?", "A : What should we do ?", "F : Yeah , it 's fine .", "F : Fine .", "A : Four ?", "F : Yeah ", "F : no , no , no , uh , I don't care .", "F : It 's fine .", "A : It 's equal to all of us ,", "A : so you can decide", "A : one or four .", "B : The pressure 's on you Nancy .", "A : Liz actually said she likes four", "A : because it forces the Meeting Recorder people to cut , you know the discussions short .", "F : OK .", "F : OK , four .", "F : OK ?", "F : OK . I am .", "E : Well , if you insist , then ." ], "abstractive summary": [ { "id": "Bed015.s.1", "text": "The discussion concerned the revised semantic specification and the construction formalism.", "type": "abstract" }, { "id": "Bed015.s.2", "text": "The different levels of the latter focus on what construction types are encountered and what bindings there are between them.", "type": "abstract" }, { "id": "Bed015.s.3", "text": "The notation maintains properties of both dependency and constituent-based grammars.", "type": "abstract" }, { "id": "Bed015.s.4", "text": "The encoding of features is still incomplete: frame profiles, focus, adjectives, nominal expressions are phenomena in the process of being integrated.", "type": "abstract" }, { "id": "Bed015.s.5", "text": "Similarly, ways to handle mental spaces will have to be added on top.", "type": "abstract" }, { "id": "Bed015.s.6", "text": "On the other hand, the semantic specification structures information in terms of \"scenario\", \"referent\" and \"discourse segment\".", "type": "abstract" }, { "id": "Bed015.s.7", "text": "Each category comprises a number of slots filled in by information derived from the utterance.", "type": "abstract" }, { "id": "Bed015.s.8", "text": "It is, essentially, a toolkit with which to create semantic constructions, as well as the bindings between them and with the ontology.", "type": "abstract" }, { "id": "Bed015.s.9", "text": "Among the issues still being defined, mental spaces and context (eg pronoun references) present similarities that can be echoed in the specification.", "type": "abstract" }, { "id": "Bed015.s.10", "text": "Work on both of these formalisms will continue with circumscription of the construction space that will be studied in more detail.", "type": "abstract" }, { "id": "Bed015.s.11", "text": "Work on construal will use Bayes-nets, which will be fed information from other modules and implement general rules to infer how utterances are construed.", "type": "abstract" }, { "id": "Bed015.s.12", "text": "The semantic specification requires some adjustments.", "type": "decisions" }, { "id": "Bed015.s.13", "text": "Amongst other things \"cause\" has to be added as another X-schema.", "type": "decisions" }, { "id": "Bed015.s.14", "text": "Linguistic hedging will also be encoded as a demarcation of evidentiality or speaker confidence.", "type": "decisions" }, { "id": "Bed015.s.15", "text": "Mental spaces can be tackled with mechanisms that can also deal with context issues (time, space etc.): creating a base space and rules of interaction with other interconnected spaces.", "type": "decisions" }, { "id": "Bed015.s.16", "text": "However, the complexity of these mechanisms has to be bound as well: it is necessary to define the range of constructions to be studied.", "type": "decisions" }, { "id": "Bed015.s.17", "text": "Given the domains currently used (tourist, child language learning), some features, like speaker attitude, are not of equal importance at this stage.", "type": "decisions" }, { "id": "Bed015.s.18", "text": "On the other hand, it was decided for the inheritance between constructions to be left out for now, as the notation can be rendered more elegant later on.", "type": "decisions" }, { "id": "Bed015.s.19", "text": "Finally, a preliminary presentation on the idea of how to use Bayes-nets for construal will take place in the next meeting.", "type": "decisions" }, { "id": "Bed015.s.20", "text": "The construction formalism is not yet complete as to the semantic constraints -the terminology has also been met with objections- and does not deal with mental spaces.", "type": "problems" }, { "id": "Bed015.s.21", "text": "On their own, constructions can only give limited information regarding mental spaces: forms can provide cues to create a different mental space, but the semantic nuances are defined by context.", "type": "problems" }, { "id": "Bed015.s.22", "text": "It is not decided at this stage whether the necessary values should be coded within the construction or as part of construal.", "type": "problems" }, { "id": "Bed015.s.23", "text": "Other issues concern focus and stress: focus is seen as an information structure device, but there has been no suggestion as to how to predict its effects or break it down in possible focused elements; as to stress, it may not be useful as a form value, as it shows the focus exponent, but not what the focus is on.", "type": "problems" }, { "id": "Bed015.s.24", "text": "Moving to the semantic specification, the analysis still needs mechanisms to deal with causality, as well as mental spaces and bindings between them.", "type": "problems" }, { "id": "Bed015.s.25", "text": "Additionally, how referring expressions are linked to referents or even how mental spaces affect this linking are still to lay down in detail.", "type": "problems" }, { "id": "Bed015.s.26", "text": "The revised semantic specification and construction formalism are more stable than the previous versions.", "type": "progress" }, { "id": "Bed015.s.27", "text": "In the latter, we find both construction types and meaning types along with formal considerations like verb subcategorisation, or the ones a \"directed motion\" construction would dictate.", "type": "progress" }, { "id": "Bed015.s.28", "text": "Semantic constraints also come into play.", "type": "progress" }, { "id": "Bed015.s.29", "text": "The semantic specification, on the other hand, is split into three levels: \"scenario\" is a list of schemas and bindings between them, which describes the current event in terms of Source-Path-Goal, Container, etc.; \"referent\" is about the entities in the discourse and includes grammatical information and pointers to the ontology; \"discourse segment\" comprises utterance-specific things.", "type": "progress" }, { "id": "Bed015.s.30", "text": "Apart from the presentation, JavaBayes can now run through the modified web page of the project.", "type": "progress" } ], "extractive summary": [ "F : So basically the result of m much thinking since the last time we met , um , but not as much writing , um , is a sheet that I have a lot of , like , thoughts and justification of comments on", "F : And so one on one side is on one side is a sort of the revised sort of updated semantic specification .", "F : And the other side is , um , sort of a revised construction formalism .", "F : it 's only slightly more stable than it was before .", "F : let 's start b let 's start on number two actually on the notation ,", "F : OK , so the top block is just sort of a sort of abstract nota", "F : it 's sort of like , um , listings of the kinds of things that we can have .", "F : So they could all have a type at the beginning .", "F : Um , and then they say the key word construction", "F : OK , and then it has a block that is constituents .", "F : But in general instead of the word \" construct \" , th here you might have \" meaning \" or \" form \" as well .", "F : Um , or you could just say in the second case that you only know the meaning type .", "F : And usually it has formal um , considerations that will go along with it .", "F : OK , and then again semantic constraints here are just are just bindings .", "F : And so it was an issue whether constraints um , there were some linguists who reacted against \" constraints \" , saying , \" oh , if it 's not used for matching , then it shouldn't be called a constraint \" .", "F : so I think the middle block doesn't really give you any more information , ex than the top block .", "F : And the bottom block similarly only just illus you know , all it does is illustrate that you can drop the subscripts and and that you can drop the , um uh , that you can give dual types .", "F : Oh , one thing I should mention is about \" designates \" .", "F : So for instance , if some other construction says , you know , \" are you of type is this part of type whatever \" , um , the \" designates \" tells you which sort of part is the meaning part .", "F : Well , right now it 's a c uh contr construction type and meaning type .", "F : So I don't know what a form type is .", "F : So one option that , you know , Keith had mentioned also was like , well if you have more abstract constructions such as subject , predicate , basically things like grammatical relations ,", "F : those could intersect with these", "F : in such a way that subject , predicate , or subject , predicate , subject , verb , ob you know , verb object would require that those things that f fill a subject and object are NOM expressions .", "E : I just know that this is Like , we didn't originally have in mind that , uh that verbs would subcategorize for a particular sort of form .", "C : But they do .", "C : The whole the mental space thing is clearly not here .", "F : So there 's going to be some extra you know , definitely other notation we 'll need for that which we skip for now .", "F : so something like \" past \" which i you know , we think is a very simple uh , we 've often just stuck it in as a feature ,", "F : Um , it 's often thought of as it is also considered a mental space ,", "F : whereas past is a very conventionalized one", "A : Do you want to do the same for space ?", "F : Oh , oh , oh , oh , instead of just time ?", "F : Same thing .", "F : So there are very conventionalized like deictic ones , right ?", "E : I mean , I 'm not sure that the formalism , sort of the grammatical side of things , is gonna have that much going on in terms of the mental space stuff .", "E : You know , um , basically all of these so - called space builders that are in the sentence are going to sort of ", "E : assuming that at any point in discourse there 's the possibility that we could be sort of talking about a bunch of different world scenarios , whatever , and the speaker 's supposed to be keeping track of those . The , um the construction that you actually get is just gonna sort of give you a cue as to which one of those that you 've already got going , um , you 're supposed to add structure to .", "C : that that that we 're not expecting much out of the , uh f", "C : uh , the purely form cues , yeah .", "F : OK , so , um , one other thing I just want to point out is there 's a lot of confusion about the terms like \" profile , designate , focus \" , et cetera , et cetera .", "F : Um , for now I 'm gonna say like \" profile \" 's often used ", "F : One is in the traditional like semantic highlight of one element with respect to everything else .", "F : And the second use , um , is in FrameNet. It 's", "F : They use it to really mean , um , this in a frame th this is the profiles on the these are the ones that are required .", "F : I think the the \" designate \" that we have in terms of meaning is really the \" highlight this thing with respect to everything else \" . OK ?", "F : But the second one seems to be useful", "F : So I d don't yet know , I have no commitment , as to whether we need it .", "C : and we still have emphasis as well , or stress , or whatever .", "F : I think that , um , the other terms that are related are like focus and stress .", "F : OK , it 's like uh , it 's not it might be that there 's a syntactic , uh , device that you use to indicate focus", "F : Uh , so so I think that 's kind of nice to keep \" focus \" being an information structure term .", "F : uh , w we 'll want to distinguish stress as a form device .", "F : You know , like , oh , high volume or whatever .", "F : Like if you 're gonna focus on this thing and you wanna know it sort of evokes all the other possibilities that it wasn't .", "F : so we need to have a notation for that", "F : I think that 's still in progress .", "F : But it did one one implication it does f have for the other side , which we 'll get to in a minute is that I couldn't think of a good way to say \" here are the possible things that you could focus on \" ,", "F : but I I can't think of like the limited set of possible meanings that you would that you would focu", "E : Light up with focus , yeah .", "F : Li - I mean , uh , the best thing I can come up with is that information has a list of focused elements .", "F : Other things we didn't totally deal with , um ,", "F : well , we 've had a lot of other stuff that Keith and I have them working on in terms of like how you deal with like an adjective .", "F : You know , a a nominal expression .", "F : So you can like nest things when you need to ,", "F : but they can also overlap in a sort of flatter way .", "F : But , you know , we have the properties of dependency grammars and some properties of constituents constituent - based grammar .", "A : Maybe you want to forget stress .", "A : canonically speaking you can if you look at a a curve over sentence , you can find out where a certain stress is", "A : and say , \" hey , that 's my focus exponent . \"", "A : because , uh , as a form cue , um , not even trained experts can always well , they can tell you where the focus exponent is sometimes .", "A : It doesn't tell you anything what the focus is .", "A : and the only way of figuring out what that is , is , um , by sort of generating all the possible alternatives to each focused element , decide which one in that context makes sense and which one doesn't .", "F : the fact that you can get it without any stress and you have some whatever is predicated anyway should be the same set of constructions .", "F : So , then you have a separate thing that picks out , you know , stress on something relative to everything else .", "F : since I couldn't think of an easy like limited way of doing it , um , all I can say is that information structure has a focused slot", "A : Yeah , well the once you know what the focus is the everything else is background .", "A : How about \" topic - comment \"", "F : Um , topic it seems kind of you know , I think that might be an ongoing kind of thing .", "F : So again , information structure has a topic slot .", "F : one thing that I didn't do consistently , um , is when we get there , is like indicate what kind of thing fits into every role .", "F : You know , it 'll be it 'll be another predication or it 'll be , um , I don't know , some value from from some something , some variable and scope or something like that ,", "F : or a slot chain based on a variable and scope .", "E : OK , side one .", "F : So you see it 's \" scenario \" , \" referent \" and \" discourse segment \" .", "F : And \" scenario \" is essentially what kind of what 's the basic predication ,", "F : what event happened .", "F : And actually it 's just a list of various slots from which you would draw draw in order to paint your picture , a bunch of frames , bi and bindings , right ?", "F : The middle thing used to be \" entities \"", "F : And this is intended to be grammatically specifiable information about a referent uh , you know , about some entity that you were going to talk about .", "F : Basically , it could be a pointer to ontology .", "F : Because this partly came from , uh , Talmy 's schema", "F : and I 'm not sure we 'll need all of these actually .", "F : So , th the the big com and and finally \" discourse segment \" is about sort of speech - act - y information structure - y ,", "F : like utterance - specific kinds of things .", "F : Um , mostly the top one is sort of image schematic .", "F : So maybe having made that big sca sort of like large scale comment , should I just go through each of these slots uh , each of these blocks , um , a little bit ?", "F : some of them seem more inherently static , OK ,", "F : like a container or sort of support - ish .", "F : And others are a little bit seemingly inherently dynamic", "F : or \" force \" , or something like that .", "F : like \" source , path , goal \" is often thought of that way", "F : And I think again , um , it 's gonna be something in the X - schema that tells you \" is this static or is this dynamic \" .", "F : uh , one thing I 'm uncertain about is how polarity interacts .", "F : So polarity , uh , is using for like action did not take place for instance .", "E : The \" scenario \" box ,", "F : It 's a grab bag of ", "E : \" part of what I 'm going to hand you is a whole bunch of s uh , schemas , image , and X - schemas .", "C : OK , uh , then this says that when an analysis is finished , the whole analysis is finished , you 'll have as a result , uh , some s resulting s semspec for that utterance in context ,", "C : which is made up entirely of these things and , uh , bindings among them .", "C : And bindings to ontology items .", "C : So that that the who that this is the tool kit under whi out of which you can make a semantic specification .", "C : So this is an that anything you have , in the party line , anything you have as the semantic side of constructions comes , from pieces of this ignoring li", "C : But it 's got to be pieces of this along with constraints among them .", "C : but pretty much this is the full range of semantic structures available to you .", "F : Except for \" cause \" , that I forgot .", "C : Let 's let 's mark that .", "F : Yeah . So that 's , uh these are all implicitly one within , uh within one world .", "F : or \" Basically there was a whole list of of possible speaker attitudes that like say Talmy listed .", "F : I don't know if those are grammatically marked in the first place .", "C : in in our in our s terminology , \" want \" can be an action", "C : and \" what you want \" is a world .", "C : Causal stuff we absolutely need .", "C : Mental space we need .", "C : The context we need .", "E : So I mean , there 's this stuff in the off in the scenario ,", "E : which just tells you how various what schemas you 're using", "E : and they 're how they 're bound together .", "E : OK , that 's where the information structure is", "E : which sort of is a kind of profiling on different parts of , um , of this .", "F : If you stressed , you know , \" John went to the \" , you know , \" the bar \" whatever , you 're focusing that", "F : and a in a possible inference is \" in contrast to other things \" .", "A : Uh , just just , uh , look read uh even sem semi formal Mats Rooth .", "A : And what I 'm kind of confused about is is what the speaker and the hearer is is sort of doing there .", "F : So for a particular segment it 's really just a reference to some other entity again in the situation , right ?", "A : Yeah , but you Don't we ultimately want to handle that analogously to the way we handle time and place ,", "F : Right , so the constructions might of course will refer , using pronouns or whatever .", "F : But when you actually say that \" he walked into \" , whatever , um , the \" he \" will refer to a particular ", "A : That 's up at the reference part .", "C : So so again , this uh , this this is gonna to get us into the mental space stuff", "C : And and so we 're , uh gonna have to , um , chain those as well .", "F : it sort of depends on if it is a referring exp if it 's identifiable already or it 's a new thing .", "F : If it 's a new thing you 'd have to like create a structure or whatever .", "F : If it 's an old thing it could be referring to , um , usually w something in a situation , right ?", "C : I just had a I just had an an idea that would be very nice if it works .", "C : Um , we might be able to handle context in the same way that we handle mental spaces", "C : it c it it , uh I think if we did it right we might be able to get at least a lot of the same structure .", "C : So that pulling something out of a discourse context is I think similar to other kinds of , uh , mental space phenomena .", "E : There 's languages where , um ,", "E : OK , so a situation like \" Bob said he was going to the movies \" , where that lower subject is the same as the person who was saying or thinking , you 're actually required to have \" I \" there .", "E : But it 's not perceived as a quotative construction .", "E : I mean , it 's been analyzed by the formalists as being a logophoric pronoun ,", "E : I mean , the point is that pronoun references , uh , you know , sort of ties in with all this mental space stuff and so on , and so forth .", "E : um which means a pronoun which refers back to the person who is speaking or that sort of thing , right ?", "E : It 's like it 's like what 's happening that , yeah , what what 's happening , uh , there is that you 're moving the base space or something like that , right ?", "F : OK , so it 's kind of like the unspecified mental spaces just are occurring in context .", "F : And then when you embed them sometimes you have to pop up to the h you know , depending on the construction or the whatever , um , you you you 're scope is m might extend out to the the base one .", "A : Mmm , isn't I mean , I 'm I was sort of dubious why why he even introduces this sort of reality , you know , as your basic mental space", "A : d doesn't start with some because it 's so obvi it should be so obvious , at least it is to me , that whenever I say something I could preface that with \" I think . \"", "C : that when you say \" I think \" you 're actually hedging .", "F : Actually one of the child language researchers who works with T Tomasello studied a bunch of these constructions", "F : and it was like it 's not using any kind of interesting embedded ways", "F : just to mark , you know , uncertainty or something like that .", "C : So we don't have that in here either do we ?", "F : Hhh , I there used to be a slot for speaker , um ,", "E : Well we were just talking about this sort of evidentiality and stuff like that , right ?", "C : Confidence or something like that .", "F : So , and and I think , uh , someone had raised like sarcasm as a complication at some point .", "E : We we don't have to care too much about the speaker attitude , right ?", "E : But like what are the distinctions among those that we actually care about for our current purposes ?", "C : It 's as far as I can tell there 's this one major thing we have to do", "C : which is the mental the whole s mental space thing .", "C : Um , and we 're going to have to s sort of bound the complexity .", "C : So we had started with the idea that the actual , uh , constraint was related to this tourist domain", "C : and the kinds of interactions that might occur in the tourist domain ,", "C : Uh , and my suggestion to Keith is that you , over the next couple weeks , n", "C : it was Keith 's , uh , sort of job over the summer to come up with this set of constructions .", "C : but just try to describe which ones you think we ought to have .", "C : Just just sort of , you know , define your space .", "C : And that 'll also be useful for anybody who 's trying to write a parser .", "C : So so we 're gonna get the w we 're basically dealing with two domains ,", "C : the tourist domain and the and the child language learning .", "C : but to focus really detailed work on the subset of things that we 've we really want to get done .", "C : And then my proposal would be to , um , not totally cut off more general discussion", "B : Is this , uh was it intentional to leave off things like \" inherits \"", "F : I didn't want to think too much about that for for now .", "E : Um , yeah , there should be I I wanted to s find out someday if there was gonna be some way of dealing with , uh , if this is the right term , multiple inheritance ,", "E : where one construction is inheriting from , uh from both parents ,", "C : Yeah . So what I would like to do is separate that problem out .", "C : And so I what I 'd like to do is is in the short run focus on getting it right .", "C : And when we think we have it right then saying , \" aha ! ,", "C : can we make it more elegant ? \"", "A : Um , thanks to Eva on our web site we can now ,", "A : if you want to run JavaBayes , uh , you could see get download these classes .", "A : she modified the GUI so it has now a m a m a button menu item for saving it into the embedded JavaBayes format .", "A : Um , the construal bit that , um , has been pointed to but hasn't been , um , made precise by any means , um , may w may work as follows .", "A : that the following thing would be in incredibly nice", "A : Um , imagine you you write a Bayes - net ,", "A : um , completely from scratch every time you do construal .", "A : You consult consult your ontology", "F : Grout out the things that that you need .", "A : You won't get you won't really get any C P", "A : T 's , therefore we need everything that that configures to what the situation is ,", "A : So you get whatever comes from discourse", "A : plus the situation and the user model .", "A : And that fills in your CPT 's with which you can then query , um , the the net that you just wrote", "A : and find out how thing X is construed as an utterance U .", "A : You may have some general rules as to how things can be can be construed as what , so that will allow you to craft the the the initial notes .", "C : So once you have instantiated the PRM with the instances and ther then you can then you can unfold it .", "C : But the question is do you want to , for example , send the little group , uh , a draft of your thesis proposal", "A : We can do it Th - Thursday again .", "A : One or four .", "F : OK , four .", "E : but it doesn't tell you like exactly what it what the point of doing so is .", "E : And I guess that some of the discourse segment stuff ", "C : don't try to do them in detail or formally" ] }, "log": [] }, "ICSI--train--4": { "original dialog id": "Bmr002.json", "dialog index": 4, "original dialog info": { "dialog history": [ "B : OK .", "B : So .", "C : I think that 's recording now .", "B : OK , I 'm on I 'm talking on mike two .", "B : So ,", "B : Which channel is that ?", "C : Um , do it again .", "B : Mike number two , I 'm talking right now on ?", "C : That 's uh That 's in channel one .", "B : Mike number two ?", "B : Channel one ?", "B : Is that channel one or channel zero ?", "C : Channel one .", "C : Let me make sure .", "A : Channel one .", "C : There are no There 's only one connection element .", "B : which is zero - based or one - based ?", "C : It 's zero - based .", "B : OK , So we 're not going to do that .", "B : OK .", "C : Unfortunately the mikes are you know ", "C : This this has the disadvantage of needing to put the numbers on the mikes .", "A : off by one .", "C : Consistently not equal to the numbers numbers on the channels .", "A : Right .", "B : Wh - that 's why I wanted to write it down ,", "B : but", "B : Dan just doesn't want to do that .", "D : Well", "C : Well , you can write it down .", "C : I mean Write down whatever you like .", "B : Thank you .", "D : Write down that poem .", "D : Let us know when you 're done , though", "D : cuz we want to start the meeting .", "D : Now , how do you do the mike ?", "B : OK , now , that means if I 'm going to be standing in there looking , then someone else has to write them down .", "D : I keep getting it in my nose .", "C : I don't know .", "C : You can do whatever any way you like .", "D : Oh !", "D : Oh ! This is good .", "D : Oh ! I like this .", "B : because I won", "C : The weird thing is that um it means that it varies when you move your head around", "D : I like this .", "B : OK .", "B : Fine .", "C : but I think it 's probably the better solution .", "B : Let 's not try to do anything too easy .", "D : I like this ,", "D : this is much more comfortable .", "C : OK ?", "B : Testing .", "D : I could have it more at the midline .", "D : Boy ! It 's very adjustable now .", "A : So go ahead .", "D : I 'm on number four .", "C : Do we have the numbers ? the the sheets , that we did from last time ?", "C : so we can do them again ?", "A : Can So we can do them again ?", "B : Yep .", "B : They 're in the folder right there .", "A : Yeah .", "C : Great .", "D : May I pass the folder to Dan ?", "B : You may .", "D : OK .", "B : So , Jane , can you talk into your mike ?", "B : You said it 's mike four ?", "D : I 'm mike I 'm on mike four .", "D : and it 's a very nice microphone ,", "D : very modern .", "B : Well , hmm , I 'm glad you like it .", "D : I like it very much .", "B : Uh Eric , what mike are you ?", "A : I 'm on channel one ", "B : Mike one ", "A : Mike one ,", "A : channel zero , probably .", "B : Looks like it . S", "B : talk again ?", "A : Channel zero .", "B : Thank you .", "C : I 'm on uh mike three .", "C : This is mike three ,", "C : coming in here .", "C : It 's moving around as I move my head backwards and forwards .", "C : So that 's just the way it goes .", "C : That 's life .", "C : That 's just what it does .", "B : OK .", "D : You didn't get the s head restraint with your system ?", "C : That 's right .", "A : Head restraint system ?", "D : I did modify my my mike uh distance from my speaking apparatus .", "B : Um , can you do taps on the P D", "D : Slightly .", "A : OK .", "B : and the P Z", "C : OK .", "B : And you have to actually check what the nu", "C : P Z Ms ? or or not P Z", "B : The - Start with P Z", "C : Because this is PZM number three .", "B : Thank you .", "C : Let 's list them in order .", "A : OK .", "C : PZM number three was the furthest away .", "C : This is the next one in , which is PZM number four , apparently .", "B : I can't tell .", "C : PZ", "C : What - Can you can you see other stuff happening ?", "B : Yep .", "C : This is PZM number four .", "B : Thanks .", "D : That 's a good way .", "A : OK .", "A : This is PZM two .", "B : F Thank you .", "A : And this is PZM one .", "C : which is closest to the machine room .", "A : Which ", "D : Should I draw a map ?", "B : OK .", "D : OK .", "C : I mean I don't know .", "B : And PDA ,", "B : left , right ?", "C : It 's all hopeless .", "C : This is uh one side of PDA , I guess .", "C : We could actually give these sides names ,", "C : right ?", "C : This would be left ,", "C : wouldn't it .", "B : I don't see anything .", "C : I know ,", "A : Huh - uh .", "C : but I 'm just asking .", "D : He 's labelling he 's labelling right now .", "C : I wasn't asking you .", "A : He 's ?", "B : Ah .", "C : This is this is PDA left .", "D : Now .", "C : You getting something there ?", "B : Yep .", "C : And this is PDA right .", "B : All OK .", "C : They will vary in future .", "B : Thank you .", "D : OK .", "A : Right .", "C : Because it depends which one of these shows more consistency then .", "B : Did we just record all that ?", "B : Should we start it over ?", "C : No .", "B : OK .", "C : Well , we 're w we 're so ahead of the game now ,", "C : cuz we 've got built - in downsampling now .", "D : We have what ? What do we have ?", "C : We 've got built - in downsampling .", "C : And so it 's only recording sixteen kilohertz data ,", "C : and we 've got ", "A : Wait a second .", "C : We 're not select We 're not recording the empty channels .", "B : The ones that aren't filled out are the ones you want to hand out .", "C : What about the ones that we didn't record ?", "A : No , I was looking for the ones ", "B : I threw the ones out , that we already recorded", "B : but we didn't record .", "A : Oh . OK .", "C : OK ?", "A : Well , then , fine .", "C : Fine .", "C : Then it doesn't matter ,", "D : Well done .", "A : See if we care .", "C : does it .", "D : Well done .", "A : OK .", "D : You 're probably wondering why I brought you all here today .", "A : Right .", "A : But before w before we ", "C : Before that .", "B : Right . So the first task is to read some digits .", "A : Before that .", "D : Yeah .", "B : So I 'll go ahead and start .", "B : And This is Adam on mike number two ,", "B : So . If you could just fill out the form ,", "B : and then later on we 'll read it , with pauses between .", "D : Good .", "B : And we 're session two .", "A : Do you wanna read next ,", "D : Shall I n", "A : or d ?", "D : I 'd like to hear one ", "B : Let her fill it out and go ahead and read .", "D : Yeah , I 'd like to hear one more .", "A : uh", "A : OK .", "A : Um , this is Eric on mike one ,", "A : the wireless lapel ,", "A : channel zero .", "A : um ,", "D : Could I hear you next ?", "C : OK .", "C : So , this is Dan on mike three ,", "C : wireless headset ,", "C : I 'm wearing it around my neck .", "C : which may be different than wearing it on the head .", "D : OK .", "D : So , basically they 're kind of like sentences but with n numbers instead of words .", "D : OK .", "D : So , I 'm I 'm on uh number four with the microphone around my neck", "D : and I will read these .", "B : So you wanna pause briefly between each one", "B : so that the person transcribing can tell where one ends and one begins .", "D : OK .", "C : And ", "B : So , a short sentence between each line .", "D : It really sounds like sentences .", "D : Your intonation sounds like somewhat like sentence intonation .", "B : Well , you can intone it however you want .", "B : Imagine that you 're reading a number phone number to someone .", "D : OK .", "D : Yeah ?", "B : Yep .", "D : Um ,", "D : OK .", "B : OK . Good .", "B : Thanks .", "C : Great .", "B : So we 'll probably do another one at the end of the meeting .", "B : just to get some more digits .", "A : Right .", "B : And , uh", "A : OK .", "B : Thank you .", "D : If I were to be dictating a phone number , I would have done that more slowly ,", "D : but uh which would have been easier for the of course , for the machine .", "D : But . Was that an OK pace ?", "A : Sure .", "C : It 's kind of weird .", "A : Yeah .", "A : It 's fine .", "C : That was great , what you did .", "C : Um", "D : OK .", "D : Just as long as it 's with you know , uh within the constraints .", "B : What - what 's your mike number , Jane ?", "D : Number four .", "D : Oh , Sorry .", "B : Dan doesn't remember what sex he is .", "C : Yeah , cuz I figure no one 's going to be able to figure it out", "C : from the from my name .", "B : That 's true , it 's it 's really difficult .", "C : So ,", "C : Yeah .", "D : and the voice .", "D : Just no way .", "B : Well , don't you ", "B : I I wrote a scanner program that scans these forms in .", "C : Oh , right , right , right .", "B : Anyway . So , why did I call you all here ?", "D : Yes .", "B : Um , what I 'd like to talk about is , um the transcription .", "D : Oh , you did .", "B : So , as part of this um project , we need to transcribe these meetings .", "B : um , to do training and test and so on .", "B : and , uh , so we need word transcripts and , uh , speaker change identification .", "B : So , speaker ID .", "B : And , uh so I 'd sort of like to know who 's going to do that ,", "B : hi Jane ,", "B : and uh what tools to use ,", "B : what other resources you might need ,", "B : and also what we ", "B : things like just the data format and how we 're going to do it .", "A : So , I 'm just curious how how h how exactly you got roped into this .", "A : I mean ", "D : Oh .", "D : Um , because my name was was mentioned when I wasn't around ,", "D : but I 'm most happy .", "A : OK .", "B : Oh really ?", "A : Good .", "B : Cuz Morgan said he asked you .", "D : Oh , uh he did .", "D : and and I approved .", "B : Mm - hmm .", "D : but I think that I was uh proposed before I was asked ?", "D : And that doesn't matter", "D : cuz I think there was probably an a uh y an indication that that I was interested in advance", "A : Oh .", "A : Yeah . That would be m That would be my fault .", "D : and , in which case , I thank you .", "A : Right .", "D : Yeah .", "A : Uh right .", "B : And so , one of the first questions is , \" do you need help ? \"", "A : OK .", "C : Well , hang on .", "A : Oh ?", "B : That 's not the first question .", "C : Well , Well , I guess .", "A : OK .", "C : Do you So , do you ", "C : What are what are we talking about ?", "C : That you 're gonna you 're gonna do transcriptions ?", "C : You 're going to listen to these things and ", "D : I 'm going to provide I 'm going to be ", "D : once I 'm given the data , I will provide w word - level transcripts with uh indication of speakers ,", "C : OK .", "D : and um", "D : I guess speaker change in any case ,", "D : so speaker ID ,", "D : and words .", "A : OK . So you 're gonna do the actual transcription yourself .", "D : Yeah .", "A : OK .", "C : D I And do you have a clear I mean , is that ", "C : Do you sort of ", "C : Are you going to have to make up what you do", "C : or do you have a very clear idea of what you 're going to do ?", "C : So I mean , is it is it Is this l Is it like something you 've done before ?", "D : Well", "C : or ", "D : I 've transcribed a lot .", "C : Yeah .", "D : But um e you know , the I mean , there are always ways of doing things more efficiently or more uh in a more technical technically sophisticated way ?", "C : And ,", "C : Yeah .", "D : Do you have to speak true English sentences when you do this ?", "C : Yeah ,", "C : because as people are going to be listening to it afterwards", "C : and saying \" Oh listen to her .", "C : She couldn't even speak true English sentences ! \"", "D : \" She can't speak ! \"", "D : \" It 's terrible ! \"", "D : OK , So um", "C : Oh , you 're meant You 're meant to speak naturally .", "C : You 're meant to say \" Oh , I didn't realize we were being recorded ! \"", "B : Ah right .", "D : Yeah .", "D : \" No one told me that . \"", "D : \" This microphone , what 's this doing around my neck ? \"", "B : Uh , that 's ", "B : I I was saying before we don't really have to worry about people being surprised by it .", "C : It 's just It 's just jewelry .", "D : OK , so um typically , what I w you know I in uh i what I 've done before is I have a uh tape recorder , with a reverse button .", "C : Yeah .", "D : And you just , you know , if it 's word level e you know , you do a e a r I just hit the r re the rewind button every couple of words", "B : Play .", "D : and type what I know", "C : Yeah .", "D : and then and then at the end do a breeze through to be sure that everything 's correct .", "C : Yeah .", "D : However , they do have software these days which with a digitized uh record you can have control over replay through some sort of an interface .", "C : Yeah .", "D : Um . I 've never used one of those .", "D : I know it would be an a wonderful idea .", "D : It 's just I haven't transcribed in in awhile .", "C : The problem is The nice thing about transcribers is they have foot pedals ,", "D : I mean ", "A : Right .", "D : Foot pedals ? I 've used transcribing machines .", "D : That 's fine .", "D : um But I frankly didn't find it any more use to me than my hand - held taperecorder .", "C : Oh .", "C : I see .", "D : However , if we have a transcribing machine I would accept it .", "D : The other thing is , you know , having getting hold of a transcriber .", "D : I think that CogSci has one .", "B : Mm - hmm .", "A : Yeah .", "D : And I could probably borrow one .", "A : Well , one th", "C : I I I actually have one .", "D : You have one yourself ?", "C : Well , Sarah has one", "D : Oh , I see !", "C : because she does a lot of transcribing of interviews .", "D : Oh , super !", "C : But The crazy thing is we 'd have to convert the s we 'd have to dump the audio files onto cassette . to use it .", "A : The other thing is is that we might be ", "A : Oh .", "D : Oh , I see .", "D : OK .", "C : I mean , given that we 've already got them digitally it would be ", "C : you know , it 's quite tempting to use some kind of uh software thing .", "D : Oh , OK .", "C : But but then there 's the question of the ergonomic interface , that ", "D : I kn", "B : Oh , I actually have some foot pedals for the P C .", "A : Well , that 's what I was gonna say , is that Adam has foot pedals that we could probably try and rig something up to ", "C : OK .", "D : Oh , that 'd be super .", "C : What do they connect to ?", "B : Uh , keyboard .", "B : and they generate keystrokes .", "C : Oh , that 's totally cool .", "C : now how how do you how do you decide which keystrokes they generate ?", "B : Um There 's a little program that downloads into it .", "B : Um .", "C : Hmm .", "D : This is great .", "C : And this is Windows ?", "B : But the program probably only runs in Windows ,", "B : although I bet we could program it once", "C : Uh - huh .", "B : And and then put it on another machine and it would work .", "A : Except for the fact that we would have to unplug it and pow and change it w", "B : I haven't used them .", "A : Does it lose the memory once you 've unpowered it ?", "B : I don't know .", "B : I don't think it does .", "C : So hang on .", "C : You have You have these foot pedals but you d you 're not using them .", "B : I 'm not using them anymore .", "B : I used to use them .", "B : So they 're currently set up for Shift , Control , and Alt .", "C : Yeah .", "B : But I r I don't use them .", "B : They 're not even plugged in .", "C : Wow .", "C : That 's so cool .", "C : OK .", "D : This sounds wonderful .", "B : And so tha uh if that would help , uh I 'm certainly willing to donate them .", "C : Well ,", "D : I 'm I 'm open to ", "C : the d the well , the uh The part that 's not wonderful is someone has to write the software .", "A : Right .", "D : Yeah , that 's true .", "B : And it it is a PC connection ,", "D : Well , you know ", "B : so it would have to go on a P C somewhere .", "C : Yeah .", "D : Now , we 're at a pilot phase right now ,", "B : Mm - hmm .", "D : and what I was told is that beginning ,", "D : you know , that the the pilot data would be something like half hour of uh or f you know half hour to forty minutes of running text .", "D : And so , for the pilot thing , you know , of course this would be useful .", "A : Right , this is the pilot data .", "D : And I w I wou I 'm assu I 'm assuming that I would not be uh you know e eh uh transcribing all of it .", "B : Yeah , yeah , you are generating it .", "B : Well , that that was part of the question ,", "B : is that if , for if we 're doing this regularly ", "C : Right .", "B : Imagine we 're doing a couple of hours a week of transcripts", "D : Yeah .", "B : um , of recording ,", "B : uh , I wouldn't expect you to do it all .", "D : No .", "D : No , no .", "B : So the question is do you know of resources that we could use .", "B : So are there grad students or undergrads or just commercial services ?", "D : Yes .", "D : Mm - hmm , yeah , OK , so now I 've I 've been involved with some of that . At at CogSci ,", "D : um So , Susan E Susan Ervin - Tripp , who you know um would typically , and I 've done this too , supervise undergrads to do this kind of thing", "A : Hmm .", "D : and they c you know , you can give them like ", "D : if you give them a little bit of instruction and and make it interesting for them they can do it as a two ninety - eight or s", "D : or , not two ninety - eight , what is it , one whatever , undergraduate research credit .", "D : Because what you do is , you can you know , build it into some sort of a ", "C : Mm - hmm .", "B : Oh , really ?", "B : So we only have to give them credits ,", "B : We don't have to give them money ?", "D : Well uh , you have to be sure you give them compensation for you know . of some of some sort .", "B : Oh , you do have to pay them .", "A : Awesome .", "D : What I mean is it 's not necessarily monetary .", "D : But if they feel like they 're getting something out of it , sometimes you can do it that way .", "B : Mm - hmm .", "B : Mm - hmm .", "D : And otherwise , as part of a grant , you can you can do it .", "D : some sort of minimal wage .", "C : Yeah .", "B : Well , we we need to We need to find out how much that will cost , so I can pass that on to Morgan ,", "D : Yeah .", "B : cuz right now we have almost no money .", "D : Yeah ,", "D : OK .", "B : So so the cheaper we can do it , the better .", "C : Well , ", "D : OK . Well , you also want to get quality .", "D : I mean , someone 's going to have to pass through Either a higher level person will have to pass through , or you have to be sure you get good quality to begin with .", "C : Yeah , absolutely .", "B : Mm - hmm .", "B : Right .", "D : But um , yeah , that 's quite you know we 've That 's been done .", "D : I can ask Susan Ervin - Tripp what she would recommend on that .", "C : Yeah .", "D : She 's really She 's done Of the people that I know , she 's the one who 's supervised the most of those things locally .", "B : Mm - hmm .", "C : What 's her name ?", "D : Uh uh Doctor Susan Ervin - Tripp .", "C : OK .", "D : faculty member in Psychology , recently retired ,", "C : Uh - huh .", "C : So ", "D : and very well known in discourse analysis .", "C : Oh , OK .", "C : So , the p I mean , what I 'm hearing is that the problem of starting off with some recordings and trying to get word transcripts is kind of I mean , it 's not a it 's not an unfamiliar problem .", "C : So that ", "D : Oh , it 's very common .", "C : So that I mean that basically , there are standard solutions and we should just adopt adopt them", "D : Yeah .", "C : and it 's going to cost however much it 's going to cost ,", "C : I mean there 's We 're we 're probably not going to have a lot of flexibility in bringing down the cost", "C : cuz probably the low cost solution 's already been established .", "C : And so that 's just whatever it is .", "B : Mm - hmm .", "C : That 's how we 'll do it .", "C : See what I mean ?", "D : I think there 's always this question of the trade - off between uh paying more for the first pass , and then h and then editing less later .", "C : Mm - hmm .", "C : Uh - huh .", "D : or paying less the first pass and editing more later .", "D : So", "C : Yeah .", "C : Well , whatever they people who have had to do this ", "C : I mean the only thing is I wonder if our requirements are different from um I mean there 's than than what 's been done in the past", "C : than discourse analyzers", "C : because maybe we have different requirements in terms of time align you know , how much time aligned detail we need and uh how much detail in terms of people .", "D : Oh , But that 's a s", "D : Mm - hmm .", "D : OK .", "C : I don I don't know how much ti I don't think we do need any time aligned detail .", "C : I think we just have basically one text file which runs from beginning to end .", "C : But .", "B : Well , if in terms of transcripts , sure ,", "B : but in terms of speaker change it might be nice to get the actual time .", "C : Sure .", "C : But not if it costs more .", "B : Right .", "A : Well , OK . So what happened with Switchboard was at least , I mean , the output was that they had speaker turns", "A : and they um analyzed it in terms of , you know , \" this is where this speaker change came in \"", "A : and of course that gets really difficult when you 're talking about ", "A : I mean , this is some of the stuff that you were talking about in uh your lunch talk ,", "C : Mm - hmm .", "A : is that is that you know , you get backchannels and stuff like that , that that , you know , disrupt the um the segment you know the segmentation .", "C : Right .", "C : Yeah .", "C : Right , I mean , the it 's sort of it 's a little In some sense it 's a little bit easier with it with Switchboard", "C : because the telephone is so actually people are less flexible in what they do .", "C : In this kind of situation where we 've got nonverbal channels then there 's a lot of overlap", "C : and there 's just that it 's not a very good model .", "A : Right .", "B : Oh , in terms of data format , what are we going to do about that ?", "B : I mean , you have the S T M format , where you have you know , every phrase , every utterance is is a single speaker .", "B : I mean , is that the right way to do it ?", "B : Or ", "C : Well , every utterance is a single speaker ,", "C : right ?", "C : It 's just that utterances can overlap .", "C : I mean ,", "B : Is that the right way to do it ?", "C : I think so .", "B : OK .", "C : I think I think I mean , I think when I 'm when ", "C : Like , So , Eric made a s made a sentence , and that should be like one utterance", "C : and then I had a couple of backchannels in there", "C : and they should be just overlapping things that ", "C : I don't think that we should break Eric 's sentences for those . I think ", "B : OK . So that means that for each utterance , we 'll need the time marks ,", "A : Right .", "B : the start and end of each utterance .", "D : Yeah .", "A : So . So ", "C : Well , we 'll have to go in from somewhere", "C : but we can't necessarily I mean , we we may be able to get the transcriber to give them to us if we give them software", "C : but we may not", "C : in which case we 'll have to get them some other way ,", "C : like with forced alignment later ,", "C : I mean you know all this can be done .", "B : Mm - hmm .", "A : So we maybe we should look at the um the tools that Mississippi State has .", "C : Yeah .", "A : Because , I d I I know that they published um i annotation tools ,", "A : you know , for their ins", "B : Well , X Waves have some as well ,", "B : but they 're pretty low level .", "B : They 're designed for um for phoneme - level", "A : Yeah .", "C : Phoneme phoneme phoneme transcriptions .", "C : Yeah .", "D : I should 've ", "B : Although , they actually have a nice tool for that could be used for speaker change marking .", "C : There 's a there are there 's a whole bunch of tools .", "C : There 's a some web page , where they have a list of like ten of them or something .", "D : Yes .", "D : Are you speaking about Mississippi State per se ?", "A : OK .", "D : or are y", "C : No , no , no . There 's some I mean , I just there are there are a lot of a lot of @ @ these things .", "D : Yeah .", "D : Actually , I wanted to mention There are two projects , which are international , huge projects focused on this kind of thing , actually .", "D : One of them 's MATE ,", "D : one of them 's EAGLES .", "D : and um . And both of them have ", "C : Oh , EAGLES .", "C : Yeah , that 's the European one .", "C : Yeah .", "D : You know , I saw I know you know about the big book .", "A : Yeah .", "D : I think you got it as a prize or something .", "C : Mmm .", "A : Yeah .", "D : Got a surprise .", "D : And um and they have you know , a lot of ", "D : and MATE is a project which is being run out of Denmark", "D : and um they have a lot uh big uh , e", "D : you know , they 're trying to set up a a work bench or something or other .", "D : And they have a whole bunch of tools that are associated with them .", "C : Mm - hmm .", "D : So , I mean , there are things that in terms of like the time alignment issue", "C : Mm - hmm .", "D : I mean , you guys are more expert on that , I mean , than I am ,", "D : but I want to say that I know for a fact , I keep running across these ,", "D : and those would be my first two choices of looking at . EAGLES and MATE .", "D : And I have the documentation downstairs .", "B : OK .", "D : I didn't think to bring it .", "B : Ha - Have you looked at it already ?", "D : Oh , I have ,", "D : but I wasn't interested in that particular issue .", "B : Mm - hmm .", "B : a a is this ", "D : So , we 're talking now , once you have words you know , how do you time align it .", "D : and that 's .", "B : Oh . So , these are p particularly for time alignment ?", "D : Well , these are uh tools ", "D : So , MATE and EAGLES are both within the so - called uh Language Engineering branch ,", "D : and um so they 're intended for this kind of a comp compu s computer science application type of approach .", "C : Yeah .", "D : These are not discourse types of tools .", "C : Right .", "D : So . They ", "B : Right . are they just for time alignment or are they for transcription as well ?", "D : Can I be excused ,", "D : and I 'll go down and get my stuff ?", "D : because um I I do think ", "C : You don't ", "C : Just take the thing with you .", "B : Th - i It 's wireless .", "D : Are you serious ?", "B : I mean it won't ", "C : Well , I don't know what will happen ,", "C : but you may as well .", "D : I but but ", "D : what uh I mean , can you spare like a minute or two of a ", "C : Sure . Sure . Sure .", "B : Sure .", "A : Sure .", "D : OK .", "D : And I go get my books and I 'll be right back .", "D : I didn't think of bringing this .", "D : Do I need this ?", "D : I need my keys .", "D : OK .", "C : My uh My s", "D : Oops !", "D : I need my card key .", "A : We can hear all your conversations along the way .", "C : Yeah , so be just ", "D : I 'll try not to say , \" So , Oh dear , Boy ! That 's itching there ! \"", "B : Yeah , don't talk to yourself too much .", "C : Yeah .", "D : So ,", "C : Yeah ,", "C : that 's true .", "C : You can leave it behind as well if you want .", "B : So , For the digits For the d", "C : I I assume that these are projects where they do the whole thing ,", "C : right ?", "B : Right .", "C : So they probably will have tools to address the actual the the the transcriber support", "D : Oops .", "C : and then automated alignment of what the transcribers give you ,", "C : which is , you know , exactly the problem we have .", "B : Mm - hmm .", "A : Right , but we can do automated alignment .", "A : That 's not a problem", "D : OK", "C : Right .", "A : I think", "C : Right .", "C : But then again we don't have to do it .", "B : Well , I 'm not sure if we can do automated alignment or not .", "C : and ", "B : I mean , e especially in e with speaker change in a meeting like this , where we 're overlapping so much ,", "A : I th", "B : I don't think we 're gonna have a a prayer .", "C : Oh sure , we are .", "C : We 've got individual microphones .", "A : Well , to get to get the r to get the actual transcript , I think we 'll do OK", "A : because of because of the separate channels .", "B : Oh , that 's true .", "B : Right .", "B : We could just do it through these .", "C : Yeah .", "A : Right .", "B : I hadn't even thought about that .", "B : Sure .", "B : Um . In terms of data formats , uh what I did for the Digits stuff was just to hand - rolled my own . That looks like this form .", "B : You know , it 's just uh something that 's easy to parse in Perl .", "B : Um . Probably for the full one , we 'll want to do something like S T", "B : I mean , some format that our tools already work with .", "C : Yeah .", "C : Well ,", "C : Yes ,", "C : although that might @ @ presumably we 'll have one we 'll have several formats that you can interconvert between .", "A : Right .", "B : Yeah , presumably .", "C : When we want to do recognition we 'll do STM ,", "C : but if we 're rather than rather than designing our own format , we 'll just take whatever comes out of the tools that we can find to use , I think .", "B : Yeah , I guess that 's true .", "B : Yeah .", "C : I I I 've got no idea what that 'll look like .", "A : Right .", "C : but", "A : Well I know that I know that um the Mississippi State tools , at least th what the produce does not come out in STM format .", "C : on our speaker phones .", "C : Right .", "A : They they basically give", "A : um They subdivide into WAVE files", "A : and then they give for each , they give a one oh , you know , a WAVE file that g indicates where the start and stop times are ,", "C : Mmm - hmm .", "A : and then um and then give the uh whatever the sentence information .", "C : Yeah .", "A : But I mean Ni - i i i it as you said , it was easy to convert back and forth between um that file and STM format ,", "A : in fact what I did is I took the Mississippi State data , and just made STM files out of it .", "C : Mm - hmm .", "C : OK .", "C : Good .", "B : OK . \" MATE Mark - up Framework \" .", "B : OK ?", "C : Right .", "D : \" Specification of uh Coding workbench \" ,", "A : Mm - hmm .", "D : That 's one o thing ,", "D : and then , back in here", "B : Oh , no !", "B : It has the word \" X M L \" in it !", "C : Oh , good .", "C : It must be It must be fundable then .", "D : \" KCM file formats \" ,", "A : Oh , looks like Looks like uh work packages .", "D : This is the", "A : Must be a European project .", "C : Yeah .", "C : Oh , yeah .", "D : Yeah , it is European .", "D : Both of these are Eur - European ,", "D : but they have you know , American contributions , and stuff .", "A : Oh , so they 're using the Alembic Workbench , from Mitre .", "C : Alembic ?", "A : I guess .", "D : Yeah .", "B : Is that free ?", "A : I don't know .", "D : Well i but you see they have a whole bunch of things .", "D : What they 're doing is trying to bring together ", "D : and they rate things .", "D : I wasn't focused on that", "D : so I didn't ", "D : Let me show y", "D : Have you seen this one , Dan ?", "C : No .", "D : And I 'll try to get one for you , too .", "D : Oh , you 've got one .", "D : What are you looking at here ?", "D : He 's got the Mark - up Framework .", "D : Oh , yeah .", "D : Different coding schemes .", "B : But it doesn't have enough pictures .", "B : So , I think it will require too much reading to understand whether this is anything that we can easily use .", "D : OK .", "A : Oh ,", "A : this is ,", "A : I see .", "A : They 're reviewing , and ", "D : What they 're doing is they 're comp comparing all these existing things that have been submitted", "D : and And it includes , this one includes the TEI", "A : I see .", "D : and uh", "D : so it 's it 's um pretty broad - based .", "B : \" TEI \" is ?", "D : These things do seem to be ", "A : Mitre ,", "D : That 's the Text Encoding Initiative .", "B : Oh .", "D : That 's a way of marking mark - up .", "D : so , you know , but focus on tools to do this stuff", "D : um , I do have a file also in addition to those things ,", "D : but", "D : it 'll take me just a second here", "B : Well , I guess XML makes some amount of sense .", "B : I 'm not actually familiar with it", "B : but it seems to be the sort of thing that this would be good for .", "B : But ", "D : Well , yeah ,", "D : I mean , if you start with a a basic transcription and a good filter . You can ", "B : Mm - hmm .", "D : I mean , that 's sort of my feeling about it ,", "D : that you should be you should be able to translate between an XML and non - XML version", "D : cuz most at least , you know , with a focus on discourse , mostly .", "B : Right .", "D : Oh ! I wanted to", "D : This this approach here This is Uli Heid from Stuttgart ", "D : is um a query language for research in phonetics where he wants to build in ToBI tags", "D : and so you can determine", "D : um I wanted to mention this to you .", "D : I was going to come by and ask you about this .", "C : Mm - hmm .", "D : Do you know about the Stuttgart um people ?", "D : Because he 's saying if you represent ", "C : I don't think so .", "D : He says that uh in German \" selbst \" is used in two different ways", "D : so you have It either means uh \" himself \" or it means \" even \"", "D : and you can tell the difference in terms of uh intonation .", "C : Mm - hmm .", "D : So , they have the recordings such that they can locate things uh with by searching for part of speech and also intonational tags .", "C : Hmm .", "C : Mm - hmm .", "D : Anyway , that 's that .", "C : Because the transcriber 's put in intonational tags ?", "D : This is true .", "D : This is true .", "D : Which of course is another issue .", "C : Or maybe not .", "D : Exactly so .", "D : and then I have one more thing I need to find .", "C : Well .", "A : But that kind of thing I think we would probably want to extract automatically .", "D : And then I 'll be ", "A : I mean , e this is , the this is a conversation that we should have with Liz .", "D : Yeah .", "D : Yes .", "D : Oh . OK .", "A : Right ,", "A : I think ,", "A : because she 's going to be She 's going to be the one who 's the most concerned about this kind of thing .", "D : OK .", "A : And so she may have some ideas . on on that particular issue .", "C : Mm - hmm .", "B : Right , but it 's I think for the initial task , just getting the word transcripts and speaker change is highest on my list .", "D : Yeah .", "D : I w I ", "D : Yeah .", "D : Well that 's what Morgan mentioned to me , too .", "B : So .", "D : I mean , if in terms of the t the type ", "A : Yeah .", "D : But you asked me about uh tools and things", "D : and um There 's this one", "C : Yeah .", "D : um This is called \" Transcriber \" .", "D : um \" binary distribution for for uh Linux and Solaris and SGI and Windows \" .", "B : Mm - hmm .", "C : Mm - hmm .", "D : \" Transcriber is a tool for assisting creation of speech corpora . \"", "C : Who 's it from ?", "C : Right .", "D : \" It allows to manually segment , label and transcribe speech signals \"", "D : And ,", "C : Is it the French one ?", "D : uh , let 's see ,", "D : where did I get this ?", "D : I got this from U Penn .", "C : U Penn .", "C : I think that I 've heard of this one .", "D : But ", "C : It was the one that Steve Renals mentioned .", "D : It is possible .", "D : It could be French .", "D : I can't really tell .", "B : Well I think that uh ", "A : Can I see that ?", "B : It sounds like you 're at least a little familiar with these tools , Jane ,", "B : so Are y Do you feel comfortable just picking one ?", "B : Um , because ", "D : I 've never used any of them .", "B : OK .", "D : And so um I think it would be something that we should work uh together with , in terms of uh seeing how this ", "C : Sure .", "A : Right .", "D : Here 's another thing I got .", "A : Oh , so this 'd this w is probably what the L D C uses .", "A : I mean they do a lot of transcription at the L D", "D : OK .", "A : I could ask my contacts at the LDC what what it is they actually use .", "D : Oh !", "A : I mean , I know , I mean I 've got I 've got contacts up there ,", "D : Good idea .", "D : Great idea .", "B : Signed language is interesting .", "A : so .", "D : Excellent idea .", "A : So .", "D : \" Automatic annotation of prosody in Verbmobil . \"", "C : Can I see ?", "A : Yeah .", "D : This is the other think that um This is part of EAGLES ,", "D : but um ,", "D : but here again , it 's focusing on representation rather than the tools to do this .", "D : Why do I feel like I 'm whispering ?", "B : Well , so , how should we proceed on picking a tool ?", "B : Jane , do you want to pick a few", "B : and we 'll download them and try them out ?", "B : Is that the right thing to do ?", "D : Oh !", "D : I think that would be lovely .", "D : Now , so , are we talking about the stage of uh simply the word stage ?", "D : or are you talking about tools that would allow the time alignment as well ?", "B : No , we 're not going to do time alignment .", "A : No .", "C : Well Well , We are going to do time alignment .", "D : OK .", "C : but we 'll probably do it ourselves ,", "C : right ?", "C : So we 're not we 're not concerned about getting getting about buying and or bringing in tools .", "A : Right .", "A : We 're very good at time alignments .", "C : We 're not very good", "D : OK .", "C : but we 're proud of ourselves .", "A : Right .", "A : OK .", "C : We 're very proud of our time alignment .", "D : Absolutely .", "D : Yeah ,", "D : OK .", "B : Right , and so but , in terms of what the transcriber needs to do , they 're not going to have to do time alignment .", "D : Good .", "D : Oh good , OK .", "B : Except maybe for word speaker change ,", "D : Fine .", "B : but ", "C : Right .", "C : I mean , it 's not a bad ", "C : Oh look !", "C : It uses Snack !", "C : Um .", "A : Uses what ?", "C : It uses Snack .", "C : That 's a TCL audio thing that I had some involvement with .", "B : OK , then we 'll have to use it .", "D : Oh , cool .", "D : Very cool .", "B : No no question .", "C : Yeah .", "C : If I have to use it", "C : I 'm sure I 'm sure It 's a TCL - based thing .", "C : I 'm sure it 's just wonderful .", "D : Cool .", "D : Cool .", "C : Yes , T \" Transcriber was developed with the scripting language TCL TK and C extensions .", "D : Yeah .", "C : It relies on the Snack sound extension , which allows support for most common audio formats . \"", "D : Cool .", "B : Boy , the recognition on what you just mumbled is going to be really horrible .", "C : Yeah .", "C : It 's the French one .", "D : Oh , cool !", "B : Yeah .", "D : OK .", "D : There .", "D : There you go .", "C : I mean , no , I I ", "D : I I love it when they", "C : But I 've never downloaded it .", "D : c", "C : Someone mentioned it to me .", "C : And that 's probably why they mentioned it to me ,", "C : because they thought I would be excited because it 's TCL - based .", "D : Did you see this one ?", "D : The - , Did you see this one here ?", "C : Oh \" Deliverable D - three point one \" ,", "C : Oh yeah .", "B : OK , So , so , it 's called \" Transcriber \" you said ?", "D : OK .", "B : gacks", "C : Transcriber", "D : Whoa !", "A : So it 's very interest I mean , it 's interesting that one of the coordinators on that is Mark Liberman from the LDC", "B : Mm - hmm .", "D : Yeah .", "A : Which means that I don't know what \" coordinator \" means in in terms of that .", "A : I mean", "B : Well , I think ", "A : What what is what is what is this web page from ?", "D : Well , I apparently , ", "D : you know , I roam around on on topics like this", "A : Sure .", "D : And I think I got this off the U Penn", "D : and it does in fact cite a French site ,", "D : so um they have some sort of an arrangement .", "B : Well , I just wrote down the French site", "A : I see .", "B : and I will check it and try to download it", "A : Right .", "B : And uh see what it looks like .", "D : Good .", "D : You - you can access it through um through the , uh , whatever ", "D : I can't remember which aspect of U of U Penn I was following ,", "A : Right .", "D : but It must have been the LDC .", "D : I I I bet you anything .", "A : Yeah .", "B : Yeah , they have a ", "B : I vaguely recall that they have a pages uh tools page somewhere ,", "D : Mm - hmm .", "D : OK .", "B : with like lists and lists and lists of uh various and assorted tools .", "A : OK .", "D : Yes .", "D : Well there 's They have a special project , as I 'm remembering ,", "D : and um , usually it doesn't deal with much that has to do with discourse ,", "B : Mm - hmm .", "D : but they uh they show up in resources .", "B : So , Um , we 'll start with \" Transcriber \"", "B : and if it looks like that 's not gonna do what we need , we 'll move on to the Mississippi State", "B : and if it doesn't look like that 's going to work , we 'll look at something else .", "D : I like that very much ,", "D : and I and I like the idea of also doing one that 's tied in with these larger corp corpus projects .", "B : OK .", "D : So that , you know , the fact that , yeah , it 's U Penn and all that .", "B : Mm - hmm .", "C : Mmm .", "A : Yeah .", "D : Yeah .", "A : Right .", "B : Also , I wanted to mention just in passing that uh another thing that we 're going to want to start collecting is queries for the meetings .", "B : So if over the next couple of days you find yourself wondering what we talked about in the meeting , write down the sort of query you 'd like to make .", "D : Oh , good idea .", "B : Do you understand what I mean ?", "D : It 's like the PDA thing .", "C : Mm - hmm .", "B : So , imagine that we had Meeting Recorder all done , and you had one ,", "B : what might you do with it ?", "B : So ,", "B : if they occur to you .", "D : So , like \" did the French system use Tcl ? \"", "B : Right .", "B : That 's a good one .", "A : Actually what I would have come up with was w w \" What was the web address of that thing \" ,", "B : Send m So , if you could just email me them .", "A : which we didn't ", "B : Yeah , except we didn't record .", "A : we didn't record that .", "B : So .", "A : Right .", "B : It 's WWW dot ETCA dot FR .", "C : That 's a fine question .", "D : if anyone wants to know .", "C : Wow .", "D : But you could get it through the U Penn ", "A : slash slash capital CTA slash , l", "B : s l slash lots of other stuff , yeah .", "A : Yeah .", "A : Yeah .", "C : So this thing , this uh deliverable D - three point one of the MATE project is a review of a lot of different tools .", "A : Right .", "D : Yes !", "D : Yeah , there you go .", "D : Exactly .", "C : They 're actually , they seem to be more", "C : the tool the main focus of the tools is you 've already got a transcription but you 're trying to tag the different turns transi transcriptions with uh discourse roles", "C : and things like that .", "A : I see .", "C : But I mean a lot them are tools that will also do transcription as well .", "C : And I could be wrong .", "D : Well , my my impression", "D : I I I really thought of these resources in the context of you you asking about time aligning the ", "D : you know , What would you need to indicate", "C : Right .", "D : and were there tools for that", "D : because um I suspect that would be uh pretty time intensive .", "D : I also wanted to say , in terms of like disc you know , if there are discussions about intonation contours I do have a certain amount of background in that", "D : and I really would be I 'd like to be involved in that .", "A : Right .", "A : No , right , I didn't mean to imply that that we that that we shouldn't discuss this now ,", "D : Yeah .", "A : but I 'm I 'm just saying that ", "D : Oh , not right now ,", "D : but I mean in the future .", "A : Right .", "D : So at this meeting with Liz I you know , I mean ", "D : I I do I 'd like to I like that stuff .", "A : Sure , sure .", "B : So , when is she showing up ?", "A : Well , I mean , they 're coming in April .", "B : April .", "B : OK .", "A : Right .", "A : But , um", "A : You know , we we may or may not talk before then on that .", "B : Mm - hmm .", "A : Um . I I think we won't probably talk until we have more data ,", "A : I would think .", "A : Right ?", "B : Right , since we have almost none right now , let 's", "A : Yeah .", "B : We need to move forward from there .", "B : But uh , yeah , I guess it 's O K .", "B : And then , uh there 's also just the general question of will it handle multiple channels or should we just ", "B : Dan and I spoke about this briefly ,", "B : do we want to try to combine channels somehow to make the transcriber 's job easier ?", "B : I was listening to some of it", "B : and uh", "B : like , if you just listen to the far P Z M the person sitting at this end of the table you can hardly hear at all .", "B : on on the transcript ,", "B : so there may be ", "B : it may be worthwhile to at least try a little bit to combine the channels .", "A : Right .", "C : I Uh , the stuff I did , I combined pairs of these into stereo pairs ,", "C : listened to them in stereo .", "B : Huh .", "C : And I was very impressed by how well you could hear separate speakers ,", "B : Neat !", "C : but I haven't I haven't done very extensive tests .", "B : Oh , so , that would be interesting as well to do that .", "A : Oh !", "A : Yeah , cuz you 'd get you get ", "C : you get a certain amount of spatial information .", "A : Right !", "C : It 's not It 's not consistent", "C : but it 's Gives you some noise ", "A : Wow .", "B : Well , I bet if you used this one the two s most separated P Z", "B : Well , maybe , maybe not .", "C : Yeah , yeah ,", "C : you 'd get get wide coverage .", "A : Well , you would you would sound like you 'd have like in this e big big head .", "B : Big ears .", "C : Right .", "C : Right .", "C : But I mean But you would you would You know , you 'd get good You 'd get good pick up at both ends of the table .", "A : Right .", "C : So .", "B : Right .", "A : Right .", "B : OK ,", "B : do we have anything else to talk about with transcribes ?", "C : Well , what is what is the goal ?", "C : again ?", "C : We 're going to end up with some I mean , ultimately , we 're going to end up with transcriptions of these meetings .", "B : And we 'll run a train a neural net", "B : and create a recognizer", "B : and do information retrieval", "C : What ,", "C : and then Then Then we 'll be done ,", "B : and ", "C : huh ?", "B : Well , that 's my interest .", "B : I mean , other people have different interests ,", "C : Um .", "B : so .", "B : That 's what I want .", "C : Of course , there are different kinds of recos", "C : so we 're going to actually train a recognizer based on these uh , far - field mikes", "B : Right .", "C : cuz what we that 's the problem we 're trying to solve .", "B : Right .", "C : OK .", "C : We need a lot of I mean ", "B : Well , what the so , the reason we 're doing near field as well is if that doesn't work we wanna back we want some something that will work .", "C : Well and also the near field will allow us to do the time alignment accurately .", "B : Right .", "B : Right .", "C : Um .", "D : So these are near near field ?", "B : Mm - hmm .", "C : Yeah .", "D : there", "D : OK .", "C : Yeah .", "C : Close mikes .", "C : I mean I ", "C : yeah , I feel like that What we 're actually going to do with these data is very unclear to me", "B : Why ?", "C : Well , because I don't think that 's a That 's not going to work .", "B : Y", "C : Right ?", "C : OK . We 're going to get horrible performance on that .", "B : The question is , \" is it will it work well enough to do information retrieval ? \"", "C : Uh - huh .", "B : So , w will the application the Meeting Recorder application , work ?", "C : Yeah .", "B : So , Is Part of the research is um \" is fifty percent accuracy enough to do an useful information retrieval ? \"", "A : Well , part", "B : It might be .", "A : I", "C : Well , we know the answer to that , but unfortunately , will we get fifty percent p accuracy ?", "B : That 's right ,", "B : Well , but you know we have to try .", "C : Yeah .", "B : We we we could simply not do the project .", "B : But I think that 's a bad solution .", "A : W We also need to d do this in terms of SmartKom ,", "A : I mean , we 're collecting this as to build general English acoustic models .", "A : that can be that they can say \" OK here 's the vocabulary", "A : and here 's the you know , here are the pronunciations . And go to it . \"", "C : And are they specifically far field mikes ?", "A : Yeah .", "A : Yeah .", "B : Oh , I didn't know that .", "C : OK .", "D : You know , if if we were to add in addition , and I 'm offering this , um to the word level , also the uh , so , stressed words , within utterances", "C : Yeah .", "D : I do think that that would help the information retrieval .", "C : Well , I mean , that would be tremendously useful information", "D : OK .", "C : assuming that we actually pursue it ,", "C : but um we 've got a huge problem ,", "C : because if we ge if we want train a recognizer we need tens of hours of labelled data .", "D : Oh , good point .", "C : And um it 's just eh it 's very frightening .", "B : Yep .", "D : Do you think ", "D : So , are are you familiar with this the MARS the MARSEC project ?", "C : uh - uh", "D : Um , So , I 'll put this over here ,", "D : cuz he seems to be the chairman .", "D : Um .", "C : Who , Eric ?", "D : Oh , he 's not ?", "A : No .", "B : He has he has the open piece of paper .", "D : Well", "D : Uh - huh .", "B : So , open notebook .", "D : He ", "D : Yeah .", "D : Well , he can y y", "A : Well , Dan 's got an open notebook , too .", "D : discuss discuss among yourselves .", "C : Outrageous !", "C : Laying out his folders .", "B : I have one", "B : but I 've carefully covered it so that it doesn't ", "A : Right .", "D : So , this is a project that um has time - aligned spoken data", "D : which is um I see it as sort of in - between um these fields", "D : because a lot of the time when people do rich um intonational and and stress - oriented transcription they don't have the digital record at the same time", "D : and the don't have it time - aligned .", "C : Mm - hmm .", "D : I don't know how large this is , actually .", "D : But I just was sort of thinking that um I 'm I 'm ", "D : and it 's British", "C : Mm - hmm .", "D : And you know", "B : Well , I think to some extent with things like the stress and the other prosodic things", "D : So .", "D : Uh - huh .", "B : an and tighter time alignments that we can do", "B : I think we should probably do those", "B : but only once we have a consumer of that data", "D : Mm - hmm .", "D : OK .", "B : ' You know , so right now we have a consumer of the word transcripts and the speaker change .", "D : Mm - hmm .", "B : Me .", "B : I want those things to work .", "B : And so I would like to get that done as quickly as possible .", "D : Mm - hmm .", "B : And so , if if Liz or someone else wants to do some work on uh prosody and on stress and so on then we can go back to the data and generate the uh the transcripts that we need .", "D : I 'm d", "D : Mm - hmm .", "D : OK .", "D : One thing about that is that is that um in the process of of doing the word level ? it really is just as easy to put the stress in at that point .", "B : Even for an undergrad ?", "D : simply the stressed ", "D : Oh , oh ,", "D : well , see in my pilot stuff ,", "D : so in the half hour or so that I 'm doing . I 'd just as soon put it in ,", "D : because I do think that you 'd find that those would be words ", "B : Mm - hmm .", "C : OK .", "D : I mean You know of course , Rosaria 's pursuing this this area", "B : Mm - hmm .", "D : but I really do think from everything that I 've run across that um you you should get gain in terms of the uh uh what you were saying ,", "D : the information retrieval aspects from that .", "B : Right .", "D : And I wouldn't be too surprised ", "D : I mean , it 's it 's like uh", "D : I mean I don't know how how reliable Lea is ,", "D : but if this idea of islands of of um for di whatever they are .", "D : r r islands of reliability or whatever it is .", "B : Accuracy ?", "D : The stressed w stressed words .", "B : Certainty ,", "B : islands of certainty .", "D : Is that it ?", "D : OK .", "B : Yeah .", "D : I mean", "A : Are you certain ?", "C : Lea ?", "B : No .", "D : The claim that uh words that get stressed are easier to recognize .", "C : I see .", "D : So , I mean , I I think it c there 's be sort of a benefit with very minimal input .", "C : Yeah .", "D : Cuz when you listen to a sentence , you can tell If it 's contrastive stress", "D : you say \" Oh ! that word was stressed \"", "D : and why the heck not put it in at the word level ?", "D : So so , I mean , the part that I 'm doing , I 'm really happy to supply that without any ", "D : cuz it it 's just it 's there .", "B : Mm - hmm .", "C : OK . If you think it won't slow you down at all .", "D : It 's not it doesn't take any thinking .", "C : Yeah .", "D : I 'm not going to talk about , you know uh , trying grade them in any kind of careful way ,", "D : but just indicate if th if there 's a prominent word , to indicate it .", "B : Sure .", "C : Yeah , that 's great .", "B : All caps .", "C : No , well , presumably there 's a an annotation standard for prefix characters or something like that .", "D : Whatever .", "D : Whatever .", "D : People ", "C : That 's what they 're using here .", "D : Yeah , people do different different ways ,", "C : OK .", "D : a and so just as long as it 's systematic in my view , you can always filter it into some other format .", "C : Yeah , exactly .", "C : Exactly .", "C : Exactly .", "B : Anything else ?", "C : Well , so , um", "C : What is the plan ?", "C : We 're gonna ", "C : I mean , How many How many How many meetings ?", "B : I 'm gonna go download \" Transcriber \"", "B : and Jane and I will look at it", "C : What about r our recording schedule ?", "B : Oh", "C : How many meetings are we going to record ?", "C : Um .", "C : Which ones are we going to transcribe ?", "C : Um .", "C : Do we ", "B : Well I think the answer is we need to record as many as we can .", "C : OK .", "B : Right ? I don't see any reason not to try to record a meeting or two a week .", "C : Absolutely .", "B : Um .", "B : I mean , more if we can .", "B : And uh so I wanna um I wanna start by in the next couple weeks just doing the Speech group", "B : until we get to the point where we 're fairly comfortable with the hardware and the software and everything works .", "C : Yeah .", "B : and then I think I 'll ", "B : also uh a Jerry volunteered to have his group do a few .", "B : So , I 'll uh try to get those as well .", "D : How many of these do you have ,", "C : OK .", "D : these near mikes ?", "C : Five .", "B : We have five of the wireless .", "D : Oh , OK .", "D : OK .", "B : And then we were going to be getting a bunch of wired ones as well .", "B : I mean notice that n", "C : Well .", "B : your you you stood up and walked around .", "B : None of the rest of us have .", "B : And so we could have wired mikes and I wouldn't be a big deal .", "D : Oh .", "D : Oh , I see .", "B : for bigger meetings .", "D : OK .", "D : I see .", "B : So .", "D : Interesting .", "D : Very interesting .", "D : Wow ,", "D : I 'm just really impressed by this ,", "D : um , yeah .", "B : It really makes you feel like you 're doing real work", "B : when you get all this hardware .", "D : Well , this is , I mean , compared to the way people normally do , like , discourse stuff ?", "D : you know , it 's like you got your ", "D : I mean they 've gotten away from w reel - to - reel tapes ,", "C : Uh - huh .", "D : but ", "D : and they do have DAT recorders ,", "D : I mean you do have ", "C : Mm - hmm .", "D : but I mean this is just wonderful , the uh added level of yeah , compactness in the channels .", "C : If we can do anything with it , yeah .", "B : Yep .", "D : I do find myself whispering .", "D : I mean , I suspect My normal s .", "B : Why ?", "D : I don't know why .", "B : I had noticed that you were talking quietly ,", "D : It 's bizarre .", "B : and I was actually going to ask you .", "D : yeah , this is not really my normal pattern .", "B : I was going to ask you to speak up .", "B : But I didn't want you to feel self - conscious .", "D : I think the problem is that I 'm afraid that I 'm talking into someone 's ear .", "B : Oh .", "D : I really do .", "D : So it 's so it 's like ", "D : And I realize now that you have volume uh control .", "D : But uh", "D : Yeah .", "C : I don't feel like you 're speaking unnaturally quietly .", "D : Oh good .", "C : I mean I feel like it 's all quite normal .", "D : OK .", "D : Good .", "A : Well , I hear what you 're saying .", "D : I 'm glad .", "C : Yeah .", "D : Good .", "C : Eric 's really got it easy cuz he 's got the lapel mike , which is so unobtrusive .", "D : That 's great .", "A : I asked if somebody else wanted the lapel mike .", "A : Somebody else can take it next time .", "B : I 've been trying to rotate it through .", "B : So , Dan will get it next time .", "D : Are these more expensive or less expensive than this kind ?", "C : They 're much more expensive .", "D : OK .", "C : I I Which I don't understand", "C : but there it is .", "D : It would be like the television . effect or something , maybe .", "C : Right , it 's because they make people feel like television stars .", "B : Do you feel like a television star ?", "C : So you can charge more for it .", "D : That 's probably why I thought he was the chairman .", "C : Yeah , that 's it .", "C : That must be it .", "B : That that makes sense ,", "A : That 's it !", "B : right ?", "B : Cuz he 's not heavily encumbered with this ", "C : Yeah .", "A : Right .", "B : N You should have run it under your sweater , though .", "A : Right .", "A : Yes .", "C : OK .", "C : Well , I don't mind .", "C : Oh , we should read some more numbers , then .", "C : Are we are we are we there ?", "D : Oh good .", "A : Yeah .", "C : Are we at that point ?", "B : I think we 're there .", "D : Good .", "A : How much space do we have left ?", "D : Good .", "C : More space than you can ", "B : uh", "C : We 've got two hours of space now .", "B : He he he changed the downsampling .", "A : Oh really .", "C : Yeah .", "B : OK , so this is Adam , on uh", "A : Oh !", "B : mike number two , wireless headset", "C : In theory .", "C : Assuming it 's recording anything .", "A : OK , this is Eric on microphone number one , uh channel zero", "D : Here , here !", "D : OK , this is Jane .", "D : OK .", "B : OK ?", "C : OK .", "B : Thanks very much .", "B : Just control - C , Dan ?", "C : Yep .", "D : Alrighty ." ] }, "log": [] }, "ICSI--train--5": { "original dialog id": "Bro018.json", "dialog index": 5, "original dialog info": { "dialog history": [ "C : OK .", "C : So uh , he 's not here ,", "D : So .", "C : so you get to ", "D : Yeah , I will try to explain the thing that I did this this week during this week .", "C : Yeah .", "D : Well eh you know that I work I begin to work with a new feature to detect voice - unvoice .", "E : Mm - hmm .", "D : What I trying two MLP to to the with this new feature and the fifteen feature uh from the eh bus base system", "E : The the mel cepstrum ?", "D : No , satly the mes the Mel Cepstrum ,", "D : the new base system the new base system .", "E : Oh the ", "E : OK ,", "D : Yeah , we ", "E : the Aurora system .", "D : yeah the Aurora system with the new filter , VAD or something like that .", "E : OK .", "D : And I 'm trying two MLP , one one that only have t three output ,", "D : voice , unvoice , and silence ,", "C : Mm - hmm .", "D : and other one that have fifty - six output .", "D : The probabilities of the allophone .", "D : And I tried to do some experiment of recognition with that", "D : and only have result with with the MLP with the three output .", "D : And I put together the fifteen features and the three MLP output .", "D : And , well , the result are li a little bit better , but more or less similar .", "C : Uh , I I 'm I 'm slightly confused .", "E : Hmm .", "C : What what feeds the uh the three - output net ?", "D : Voice , unvoice , and si", "C : No no , what feeds it ?", "C : What features does it see ?", "D : The feature the input ?", "D : The inputs are the fifteen the fifteen uh bases feature .", "C : Uh - huh .", "D : the with the new code .", "D : And the other three features are R , the variance of the difference between the two spectrum ,", "C : Uh - huh .", "D : the variance of the auto - correlation function , except the the first point , because half the height value is R - zero", "C : Mm - hmm .", "C : Mm - hmm .", "C : Mm - hmm .", "C : Mm - hmm .", "D : and also R - zero ,", "D : the first coefficient of the auto - correlation function .", "D : That is like the energy with these three feature ,", "C : Right .", "D : also these three feature .", "C : You wouldn't do like R - one over R - zero or something like that ?", "C : I mean usually for voiced - unvoiced you 'd do yeah , you 'd do something you 'd do energy", "D : Yeah .", "C : but then you have something like spectral slope , which is you get like R - one ov over R - zero or something like that .", "D : Uh yeah .", "E : What are the R 's ?", "E : I 'm sorry I missed it .", "C : R correlations .", "D : No ,", "E : Oh .", "D : R c No .", "D : Auto - correlation ?", "D : Yes , yes , the variance of the auto - correlation function that uses that", "C : Ye - Well that 's the variance ,", "C : but if you just say \" what is \"", "C : I mean , to first order , um", "C : yeah one of the differences between voiced , unvoiced and silence is energy .", "D : Yeah ,", "C : Another one is but the other one is the spectral shape .", "D : I I 'll ", "D : The spectral shape ,", "D : yeah .", "C : Yeah ,", "C : and so R - one over R - zero is what you typically use for that .", "D : No ,", "D : I don't use that ", "D : I can't use ", "C : No ,", "C : I 'm saying that 's what people us typically use .", "D : Mmm .", "C : See , because it because this is this is just like a single number to tell you um \" does the spectrum look like that or does it look like that \" .", "D : Mm - hmm .", "A : Oh . R ", "A : R R - zero .", "C : Right ?", "D : Mm - hmm .", "C : So if it 's if it 's um if it 's low energy uh but the but the spectrum looks like that or like that , it 's probably silence .", "D : Mm - hmm .", "C : Uh but if it 's low energy and the spectrum looks like that , it 's probably unvoiced .", "D : Yeah .", "C : So if you just if you just had to pick two features to determine voiced - unvoiced , you 'd pick something about the spectrum like uh R - one over R - zero , um and R - zero", "D : Mm - hmm ,", "D : OK .", "C : or i i you know you 'd have some other energy measure", "C : and", "C : like in the old days people did like uh zero crossing counts .", "D : Yeah , yeah .", "C : Right . S", "C : S", "D : Well , I can also th use this .", "C : Yeah .", "C : Um ,", "D : Bec - because the result are a little bit better", "D : but we have in a point that everything is more or less the similar more or less similar .", "C : Yeah .", "C : But um", "D : It 's not quite better .", "C : Right , but it seemed to me that what you were what you were getting at before was that there is something about the difference between the original signal or the original FFT and with the filter which is what ", "C : and the variance was one take uh on it .", "D : Yeah ,", "D : I used this too .", "C : Right .", "C : But it it could be something else .", "C : Suppose you didn't have anything like that .", "C : Then in that case , if you have two nets ,", "C : Alright , and this one has three outputs , and this one has f", "D : Mm - hmm .", "C : whatever ,", "C : fifty - six , or something ,", "D : Mm - hmm .", "C : if you were to sum up the probabilities for the voiced and for the unvoiced and for the silence here , we 've found in the past you 'll do better at voiced - unvoiced - silence than you do with this one .", "C : So just having the three output thing doesn't doesn't really buy you anything .", "D : Yeah .", "C : The issue is what you feed it .", "D : Yeah ,", "D : I have ", "D : yeah .", "C : So uh", "E : So you 're saying take the features that go into the voiced - unvoiced - silence net and feed those into the other one , as additional inputs , rather than having a separate ", "D : No ", "C : w W well that 's another way .", "D : Yeah .", "C : That wasn't what I was saying", "C : but yeah that 's certainly another thing to do .", "C : No I was just trying to say if you b if you bring this into the picture over this , what more does it buy you ?", "E : Mmm .", "C : And what I was saying is that the only thing I think that it buys you is um based on whether you feed it something different .", "C : And something different in some fundamental way .", "C : And so the kind of thing that that she was talking about before , was looking at something uh ab um something uh about the difference between the the uh um log FFT uh log power uh and the log magnitude uh F F - spectrum uh and the um uh filter bank .", "D : Yeah .", "C : And so the filter bank is chosen in fact to sort of integrate out the effects of pitch", "C : and she 's saying", "C : you know trying ", "C : So the particular measure that she chose was the variance of this m of this difference ,", "D : Mm - hmm .", "C : but that might not be the right number .", "D : Maybe .", "C : Right ?", "C : I mean maybe there 's something about the variance that 's that 's not enough", "C : or maybe there 's something else that that one could use ,", "C : but I think that , for me , the thing that that struck me was that uh you wanna get something back here ,", "C : so here 's here 's an idea .", "C : uh What about it you skip all the all the really clever things , and just fed the log magnitude spectrum into this ?", "D : Ah I 'm sorry .", "C : This is f", "C : You have the log magnitude spectrum , and you were looking at that and the difference between the filter bank and and c c computing the variance .", "D : Yeah .", "D : Mm - hmm .", "D : Mm - hmm .", "C : That 's a clever thing to do .", "C : What if you stopped being clever ?", "D : Mm - hmm .", "C : And you just took this thing in here because it 's a neural net and neural nets are wonderful", "C : and figure out what they can what they most need from things , and I mean that 's what they 're good at .", "D : Yeah .", "C : So I mean you 're you 're you 're trying to be clever and say what 's the statistic that should we should get about this difference", "C : but uh in fact , you know maybe just feeding this in or or feeding both of them in", "E : Hmm .", "C : you know , another way , saying let it figure out what 's the what is the interaction ,", "C : especially if you do this over multiple frames ?", "D : Mm - hmm .", "C : Then you have this over time , and and both kinds of measures", "C : and uh you might get uh something better .", "D : Mm - hmm .", "C : Um .", "E : So so don't uh don't do the division ,", "E : but let the net have everything .", "C : That 's another thing you could do", "C : yeah .", "C : Yeah .", "D : Yeah .", "C : Um .", "C : I mean , it seems to me , if you have exactly the right thing then it 's better to do it without the net", "C : because otherwise you 're asking the net to learn this ", "C : you know , say if you wanted to learn how to do multiplication .", "E : Mm - hmm .", "C : I mean you could feed it a bunch of s you could feed two numbers that you wanted to multiply into a net", "C : and have a bunch of nonlinearities in the middle", "C : and train it to get the product of the output and it would work .", "C : But , it 's kind of crazy ,", "C : cuz we know how to multiply", "C : and you you 'd be you know much lower error usually if you just multiplied it out .", "C : But suppose you don't really know what the right thing is .", "C : And that 's what these sort of dumb machine learning methods are good at .", "C : So .", "C : Um .", "C : Anyway .", "C : It 's just a thought .", "E : How long does it take , Carmen , to train up one of these nets ?", "D : Oh , not too much .", "E : Yeah .", "D : Mmm , one day or less .", "E : Hmm .", "C : Yeah ,", "C : it 's probably worth it .", "A : What are what are your f uh frame error rates for for this ?", "D : Eh fifty - f six", "D : uh no , the frame error rate ?", "D : Fifty - six I think .", "A : O", "C : Is that maybe that 's accuracy ?", "D : Percent .", "D : The accuracy .", "A : Fif - fifty - six percent accurate for v voice - unvoice", "D : Mm - hmm .", "D : No for ,", "D : yes f", "D : I don't remember for voice - unvoice ,", "A : Oh , OK .", "D : maybe for the other one .", "C : Yeah , voiced - unvoiced hopefully would be a lot better .", "A : OK .", "D : for voiced .", "D : I don't reme", "A : Should be in nineties somewhere .", "D : Better .", "D : Maybe for voice - unvoice .", "A : Right .", "D : This is for the other one .", "D : I should I can't show that .", "A : OK .", "D : But I think that fifty - five was for the when the output are the fifty - six phone .", "A : Mm - hmm .", "D : That I look in the with the other nnn the other MLP that we have are more or less the same number .", "D : Silence will be better but more or less the same .", "C : I think at the frame level for fifty - six that was the kind of number we were getting for for uh um reduced band width uh stuff .", "D : I think that I I I think that for the other one , for the three output , is sixty sixty - two , sixty three more or less .", "A : Mm - hmm .", "C : That 's all ?", "D : It 's ", "D : Yeah .", "C : That 's pretty bad .", "D : Yeah , because it 's noise also .", "C : Aha !", "A : Oh yeah .", "D : And we have", "C : Aha !", "C : Yeah .", "D : I know .", "C : Yeah .", "C : OK .", "C : But even i in ", "C : Oh yeah , in training .", "C : Still ,", "C : Uh .", "C : Well actually , so this is a test that you should do then .", "C : Um , if you 're getting fifty - six percent over here ,", "C : uh that 's in noise also ,", "C : right ?", "D : Yeah , yeah , yeah .", "C : Oh OK .", "C : If you 're getting fifty - six here , try adding together the probabilities of all of the voiced phones here and all of the unvoiced phones", "D : will be ", "C : and see what you get then .", "D : Yeah .", "C : I bet you get better than sixty - three .", "D : Well I don't know ,", "D : but I th I I think that we I have the result more or less . Maybe .", "D : I don't know .", "D : I don't I 'm not sure but I remember @ @ that I can't show that .", "C : OK , but that 's a That is a a good check point ,", "C : you should do that anyway ,", "D : Yeah .", "C : OK ?", "C : Given this this uh regular old net that 's just for choosing for other purposes , uh add up the probabilities of the different subclasses and see see how well you do .", "C : Uh and that you know anything that you do over here should be at least as good as that .", "D : Mm - hmm .", "C : OK .", "D : I will do that .", "E : The targets for the neural net , uh , they come from forced alignments ?", "D : But ", "D : Uh , no .", "A : TIMIT canonical ma mappings .", "D : TIMIT .", "E : Ah !", "C : Oh . So , this is trained on TIMIT .", "E : OK .", "D : Yeah .", "A : Yeah , noisy TIMIT .", "C : OK .", "D : Yeah this for TIMIT .", "C : But noisy TIMIT ?", "D : Noisy TIMIT .", "A : Right .", "D : We have noisy TIMIT with the noise of the the TI - digits .", "D : And now we have another noisy TIMIT also with the noise of uh Italian database .", "C : I see .", "C : Yeah .", "C : Well there 's gonna be it looks like there 's gonna be a noisy uh some large vocabulary noisy stuff too .", "C : Somebody 's preparing .", "E : Really ?", "C : Yeah .", "C : I forget what it 'll be ,", "C : resource management ,", "C : Wall Street Journal ,", "C : something .", "C : Some some read task actually , that they 're preparing .", "A : Hmm !", "E : For what ", "E : For Aurora ?", "C : Yeah .", "E : Oh !", "C : Yeah , so the uh ", "C : Uh , the issue is whether people make a decision now based on what they 've already seen , or they make it later .", "C : And one of the arguments for making it later is let 's make sure that whatever techniques that we 're using work for something more than than connected digits .", "E : Hmm .", "C : So .", "E : When are they planning When would they do that ?", "C : Mmm , I think late uh I think in the summer sometime .", "E : Hmm .", "C : So .", "C : OK , thanks .", "D : This is the work that I did during this date", "C : Uh - huh .", "D : and also mmm I H Hynek last week say that if I have time I can to begin to to study", "D : well seriously the France Telecom proposal", "C : Mm - hmm .", "D : to look at the code", "D : and something like that", "D : to know exactly what they are doing", "D : because maybe that we can have some ideas", "C : Mm - hmm .", "D : but not only to read the proposal .", "D : Look insi look i carefully what they are doing with the program @ @ and", "D : I begin to to work also in that .", "D : But the first thing that I don't understand is that they are using R - the uh log energy that this quite ", "D : I don't know why they have some constant in the expression of the lower energy .", "D : I don't know what that means .", "E : They have a constant in there , you said ?", "D : Yeah .", "C : Oh , at the front it says uh \" log energy is equal to the rounded version of sixteen over the log of two \"", "D : This ", "D : Yeah .", "C : Uh .", "D : Then maybe I can understand .", "C : uh times the ", "C : Well , this is natural log ,", "C : and maybe it has something to do with the fact that this is ", "E : Is that some kind of base conversion ,", "C : I I have no idea .", "E : or ?", "C : Yeah , that 's what I was thinking ,", "C : but but um , then there 's the sixty - four ,", "C : Uh , I don't know .", "D : Because maybe they 're the threshold that they are using on the basis of this value ", "E : Experimental results .", "A : Mc - McDonald 's constant .", "D : I don't know exactly ,", "D : because well th I thought maybe they have a meaning .", "D : But I don't know what is the meaning of take exactly this value .", "C : Yeah , it 's pretty funny looking .", "E : So they 're taking the number inside the log and raising it to sixteen over log base two .", "C : I don't know .", "C : Yeah , I ", "C : um Right .", "C : Sixteen over two .", "E : Does it have to do with those sixty - fours ,", "E : or ?", "C : Um .", "C : If we ignore the sixteen , the natural log of t one over the natural log of two times the natu", "C : I don't know .", "C : Well , maybe somebody 'll think of something ,", "C : but this is uh ", "C : It may just be that they they want to have for very small energies ,", "C : they want to have some kind of a ", "D : Yeah , the e The effect I don't @ @ I can understand the effect of this ,", "D : no ?", "D : because it 's to to do something like that .", "D : No ?", "C : Well , it says , since you 're taking a natural log , it says that when when you get down to essentially zero energy , this is gonna be the natural log of one , which is zero .", "D : Mm - hmm .", "C : So it 'll go down to uh to the natural log being ", "C : So the lowest value for this would be zero .", "C : So y you 're restricted to being positive .", "C : And this sort of smooths it for very small energies .", "C : Uh , why they chose sixty - four and something else , that was probably just experimental .", "D : Yeah .", "C : And the the the constant in front of it , I have no idea .", "C : um", "D : Well . I I will look to try if I move this parameter in their code what happens ,", "D : maybe everything is Maybe they tres hole are on basis of this .", "C : uh I mean it they they probably have some fi particular s fixed point arithmetic that they 're using ,", "D : I don't know .", "E : Yeah , I was just gonna say maybe it has something to do with hardware ,", "C : and then it just ", "E : something they were doing .", "C : Yeah .", "C : Yeah ,", "C : I mean that they 're s probably working with fixed point or integer or something .", "C : I think you 're supposed to on this stuff anyway ,", "C : and and so maybe that puts it in the right realm somewhere .", "E : Well it just , yeah , puts it in the right range ,", "E : or ", "C : Yeah .", "C : I think , given at the level you 're doing things in floating point on the computer , I don't think it matters , would be my guess ,", "D : Mm - hmm .", "C : but .", "D : I this more or less anything", "C : Yeah .", "C : OK , and wh when did Stephane take off ?", "C : He took off ", "D : I think that Stephane will arrive today or tomorrow .", "C : Oh , he was gone these first few days ,", "C : and then he 's here for a couple days before he goes to Salt Lake City .", "D : Mm - hmm .", "C : OK .", "D : He 's I think that he is in Las Vegas or something like that .", "C : Yeah .", "C : Yeah .", "C : So he 's he 's going to ICASSP which is good .", "C : I I don't know if there are many people who are going to ICASSP", "D : Yeah .", "C : so so I thought , make sure somebody go .", "D : Yeah .", "E : Do have Have people sort of stopped going to ICASSP in recent years ?", "C : Um , people are less consistent about going to ICASSP", "C : and I think it 's still it 's still a reasonable forum for students to to present things .", "C : Uh , it 's I think for engineering students of any kind , I think it 's it 's if you haven't been there much , it 's good to go to ,", "C : uh to get a feel for things , a range of things ,", "C : not just speech .", "C : Uh .", "C : But I think for for sort of dyed - in - the - wool speech people , um I think that ICSLP and Eurospeech are much more targeted .", "E : Mm - hmm .", "C : Uh .", "C : And then there 's these other meetings , like HLT and and uh ASRU ", "C : so there 's there 's actually plenty of meetings that are really relevant to to uh computational uh speech processing of one sort or another .", "E : Mm - hmm .", "C : Um . So .", "C : I mean , I mostly just ignored it because I was too busy and didn't get to it .", "C : So uh", "C : Wanna talk a little bit about what we were talking about this morning ?", "A : Oh !", "C : Just briefly , or Or anything else ?", "A : um uh Yeah .", "A : So . I I guess some of the progress , I I 've been getting a getting my committee members for the quals .", "A : And um so far I have Morgan and Hynek , Mike Jordan ,", "A : and I asked John Ohala and he agreed .", "E : Cool .", "A : Yeah . Yeah .", "A : So I 'm I I just need to ask um Malek .", "A : One more .", "A : Um . Tsk .", "A : Then uh I talked a little bit about um continuing with these dynamic ev um acoustic events ,", "A : and um we 're we 're we 're thinking about a way to test the completeness of a a set of um dynamic uh events .", "A : Uh , completeness in the in the sense that um if we if we pick these X number of acoustic events , do they provide sufficient coverage for the phones that we 're trying to recognize or or the f the words that we 're gonna try to recognize later on .", "A : And so Morgan and I were uh discussing um s uh s a form of a cheating experiment where we get um we have uh um a chosen set of features , or acoustic events ,", "A : and we train up a hybrid um system to do phone recognition on TIMIT .", "A : So i i the idea is if we get good phone recognition results , using um these set of acoustic events , then um that that says that these acoustic events are g sufficient to cover a set of phones ,", "A : at least found in TIMIT .", "A : Um so i it would be a a measure of \" are we on the right track with with the the choices of our acoustic events \" .", "A : Um , So that 's going on .", "A : And also , just uh working on my uh final project for Jordan 's class ,", "A : uh which is ", "C : Actually , let me ", "C : Hold that thought .", "A : Yeah .", "C : Let me back up while we 're still on it .", "A : OK , sure .", "C : The the other thing I was suggesting , though , is that given that you 're talking about binary features , uh , maybe the first thing to do is just to count", "C : and uh count co - occurrences and get probabilities for a discrete HMM", "C : cuz that 'd be pretty simple", "C : because it 's just Say , if you had ten ten events , uh that you were counting , uh each frame would only have a thousand possible values for these ten bits ,", "C : and uh so you could make a table that would say , if you had thirty - nine phone categories , that would be a thousand by thirty - nine ,", "C : and just count the co - occurrences and divide them by the the uh uh uh occ uh count the co - occurrences between the event and the phone and divide them by the number of occurrences of the phone ,", "C : and that would give you the likelihood of the of the event given the phone .", "C : And um then just use that in a very simple HMM", "C : and uh you could uh do phone recognition then and uh wouldn't have any of the issues of the uh training of the net or ", "C : I mean , it 'd be on the simple side ,", "C : but", "E : Mm - hmm .", "C : uh um", "C : you know , if uh uh the example I was giving was that if if you had um onset of voicing and and end of voicing as being two kinds of events , then if you had those a all marked correctly , and you counted co - occurrences , you should get it completely right .", "E : Mm - hmm .", "C : So .", "C : um But you 'd get all the other distinctions , you know , randomly wrong .", "C : I mean there 'd be nothing to tell you that .", "C : So um uh", "C : If you just do this by counting , then you should be able to find out in a pretty straightforward way whether you have a sufficient uh set of events to to do the kind of level of of uh classification of phones that you 'd like .", "C : So that was that was the idea .", "C : And then the other thing that we were discussing was was um OK , how do you get the your training data .", "E : Mm - hmm .", "C : Cuz uh the Switchboard transcription project uh uh you know was half a dozen people , or so working off and on over a couple years ,", "C : and", "C : uh similar similar amount of data to what you 're talking about with TIMIT training .", "C : So , it seems to me that the only reasonable starting point is uh to automatically translate the uh current TIMIT markings into the markings you want .", "C : And uh it won't have the kind of characteristic that you 'd like , of catching funny kind of things that maybe aren't there from these automatic markings ,", "E : Mm - hmm .", "C : but but", "C : uh it 's uh ", "E : It 's probably a good place to start .", "C : Yeah .", "E : Yeah .", "C : Yeah and a short short amount of time ,", "C : just to again , just to see if that information is sufficient to uh determine the phones .", "E : Mm - hmm .", "E : Hmm .", "C : So .", "E : Yeah , you could even then to to get an idea about how different it is , you could maybe take some subset and you know , go through a few sentences , mark them by hand", "E : and then see how different it is from you know , the canonical ones ,", "C : Right .", "E : just to get an idea a rough idea of h if it really even makes a difference .", "C : You can get a little feeling for it that way ,", "E : Yeah .", "C : yeah that is probably right .", "C : I mean uh my my guess would be that this is since TIMIT 's read speech that this would be less of a big deal ,", "E : Mm - hmm .", "C : if you went and looked at spontaneous speech it 'd be more more of one .", "E : Right .", "E : Right .", "C : And the other thing would be , say , if you had these ten events , you 'd wanna see , well what if you took two events or four events or ten events or t", "C : and you know , and ", "C : and hopefully there should be some point at which having more information doesn't tell you really all that much more about what the phones are .", "E : Mm - hmm .", "E : You could define other events as being sequences of these events too .", "C : Uh , you could ,", "C : but the thing is , what he 's talking about here is a uh a translation to a per - frame feature vector ,", "C : so there 's no sequence in that ,", "C : I think .", "C : I think it 's just a ", "E : Unless you did like a second pass over it or something after you 've got your ", "C : Yeah , but we 're just talking about something simple here , yeah , to see if ", "E : Yeah .", "E : Yeah , yeah .", "E : Yeah .", "E : I 'm adding complexity .", "C : Yeah .", "C : Just ", "C : You know . The idea is with a with a very simple statistical structure , could you could you uh at least verify that you 've chosen features that are sufficient .", "E : Yeah .", "C : OK , and you were saying something starting to say something else about your your class project , or ?", "A : Oh .", "A : Yeah th Um .", "C : Yeah .", "A : So for my class project I 'm um I 'm tinkering with uh support vector machines ?", "A : something that we learned in class ,", "A : and uh um basically just another method for doing classification .", "A : And so I 'm gonna apply that to um compare it with the results by um King and Taylor who did um these", "A : um", "A : using recurrent neural nets , they recognized um a set of phonological features", "A : um", "A : and made a mapping from the MFCC 's to these phonological features ,", "A : so I 'm gonna do a similar thing with with support vector machines", "A : and see if ", "E : So what 's the advantage of support vector machines ?", "E : What ", "A : Um . So , support vector machines are are good with dealing with a less amount of data", "E : Hmm .", "A : and um so if you if you give it less data it still does a reasonable job in learning the the patterns .", "E : Hmm .", "A : Um and um", "C : I guess it ", "C : yeah ,", "C : they 're sort of succinct ,", "C : and and they uh", "A : Yeah .", "E : Does there some kind of a distance metric that they use", "E : or how do they for cla what do they do for classification ?", "A : Um .", "A : Right .", "A : So , the the simple idea behind a support vector machine is um , you have you have this feature space ,", "A : right ?", "E : Mm - hmm .", "A : and then it finds the optimal separating plane , um between these two different um classes ,", "E : Mm - hmm .", "E : Mm - hmm .", "A : and um and so um ,", "A : what it i at the end of the day , what it actually does is it picks those examples of the features that are closest to the separating boundary ,", "A : and remembers those", "E : Mm - hmm .", "A : and and uses them to recreate the boundary for the test set .", "A : So , given these um these features , or or these these examples , um , critical examples , which they call support f support vectors , then um given a new example , if the new example falls um away from the boundary in one direction then it 's classified as being a part of this particular class", "E : Oh .", "A : and otherwise it 's the other class .", "E : So why save the examples ?", "E : Why not just save what the boundary itself is ?", "A : Mm - hmm .", "A : Um . Hmm .", "A : Let 's see .", "A : Uh .", "A : Yeah , that 's a good question .", "A : I ", "C : That 's another way of doing it .", "A : yeah .", "E : Mmm .", "C : Right ? So so it I mean I I guess it 's ", "E : Sort of an equivalent .", "C : You know , it it goes back to nearest - neighbor sort of thing ,", "E : Mm - hmm .", "C : right ?", "C : Um , i i if is it eh w", "C : When is nearest - neighbor good ?", "C : Well , nearest - neighbor good is good if you have lots and lots of examples .", "C : Um but of course if you have lots and lots of examples , then it can take a while to to use nearest - neighbor .", "C : There 's lots of look ups .", "C : So a long time ago people talked about things where you would have uh a condensed nearest - neighbor ,", "C : where you would you would you would pick out uh some representative examples which would uh be sufficient to represent to to correctly classify everything that came in .", "E : Oh .", "E : Mm - hmm .", "C : I I think s I think support vector stuff sort of goes back to to that kind of thing .", "E : I see .", "C : Um .", "E : So rather than doing nearest neighbor where you compare to every single one , you just pick a few critical ones ,", "E : and ", "C : Yeah .", "E : Hmm .", "C : And th the", "C : You know , um neural net approach uh or Gaussian mixtures for that matter are sort of fairly brute force kinds of things , where you sort of you predefine that there is this big bunch of parameters", "C : and then you you place them as you best can to define the boundaries ,", "C : and in fact , as you know , these things do take a lot of parameters", "C : and and uh if you have uh only a modest amount of data , you have trouble uh learning them .", "C : Um , so I I guess the idea to this is that it it is reputed to uh be somewhat better in that regard .", "E : Mm - hmm .", "A : Right .", "A : I it can be a a reduced um parameterization of of the the model by just keeping certain selected examples .", "E : Hmm .", "A : Yeah .", "A : So .", "C : But I don't know if people have done sort of careful comparisons of this on large tasks or anything .", "C : Maybe maybe they have .", "C : I don't know .", "A : Yeah , I don't know either .", "C : Yeah .", "B : S do you get some kind of number between zero and one at the output ?", "A : Actually you don't get a you don't get a nice number between zero and one .", "A : You get you get either a zero or a one .", "A : Um , uh", "A : there are there are pap", "A : Well , basically , it 's it 's um you you get a distance measure at the end of the day ,", "A : and then that distance measure is is um is translated to a zero or one .", "A : Um .", "C : But that 's looking at it for for classification for binary classification ,", "A : That 's for classification ,", "E : And you get that for each class , you get a zero or a one .", "C : right ?", "A : right .", "A : Right .", "C : But you have the distances to work with .", "A : You have the distances to work with ,", "C : Cuz actually Mississippi State people did use support vector machines for uh uh speech recognition and they were using it to estimate probabilities .", "A : yeah .", "A : Yeah .", "A : Yeah , they they had a had a way to translate the distances into into probabilities with the with the simple um uh sigmoidal function .", "C : Yeah ,", "C : and d did they use sigmoid or a softmax type thing ?", "C : And didn't they like exponentiate or something", "A : Um Yeah ,", "C : and then divide by the sum of them ,", "A : there 's some there 's like one over one plus the exponential or something like that .", "C : or ?", "C : Oh it i", "C : Oh , so it is a sigmoidal .", "A : Yeah .", "C : OK .", "C : Alright .", "E : Did the did they get good results with that ?", "C : I mean , they 're OK ,", "C : I I don't I don't think they were earth earth shattering ,", "C : but I think that uh this was a couple years ago ,", "E : Hmm .", "C : I remember them doing it at some meeting ,", "C : and and um I don't think people were very critical because it was interesting just to to try this", "C : and you know , it was the first time they tried it ,", "C : so so the you know , the numbers were not incredibly good", "E : Hmm .", "C : but there 's you know , it was th reasonable .", "E : Mm - hmm .", "C : I I don't remember anymore .", "C : I don't even remember what the task was ,", "C : it was Broadcast News , or something .", "C : I don't know .", "E : Hmm .", "A : Right .", "B : Uh s So Barry , if you just have zero and ones , how are you doing the speech recognition ?", "A : Oh", "A : I 'm not do I 'm not planning on doing speech recognition with it .", "A : I 'm just doing detection of phonological features .", "B : Oh . OK .", "A : So uh for example , this this uh feature set called the uh sound patterns of English um is just a bunch of um binary valued features .", "A : Let 's say , is this voicing , or is this not voicing ,", "A : is this sonorants , not sonorants ,", "A : and stuff like that .", "B : OK .", "A : So .", "E : Did you find any more mistakes in their tables ?", "A : Oh !", "A : Uh I haven't gone through the entire table , yet .", "A : Yeah ,", "A : yesterday I brought Chuck the table", "A : and I was like , \" wait , this is Is the mapping from N to to this phonological feature called um \" coronal \" ,", "A : is is should it be shouldn't it be a one ?", "A : or should it should it be you know coronal instead of not coronal as it was labelled in the paper ? \"", "A : So I ha haven't hunted down all the all the mistakes yet ,", "C : Uh - huh .", "A : but ", "C : But a as I was saying , people do get probabilities from these things ,", "B : OK .", "C : and and uh we were just trying to remember how they do ,", "C : but people have used it for speech recognition , and they have gotten probabilities .", "C : So they have some conversion from these distances to probabilities .", "B : OK .", "A : Right , yeah .", "C : There 's you have you have the paper ,", "C : right ?", "C : The Mississippi State paper ?", "A : Mm - hmm .", "A : Mm - hmm .", "C : Yeah , if you 're interested y you could look ,", "B : And ", "B : OK . OK .", "A : Yeah ,", "C : yeah .", "A : I can I can show you I ", "A : yeah ,", "E : So in your in in the thing that you 're doing , uh you have a vector of ones and zeros for each phone ?", "A : our ", "A : Mm - hmm .", "A : Uh , is this the class project ,", "A : or ?", "E : Yeah .", "A : OK .", "A : um", "E : Is that what you 're ", "A : Right , Right , right", "A : f so for every phone there is there is a um a vector of ones and zeros f uh corresponding to whether it exhibits a particular phonological feature or not .", "E : Mm - hmm .", "E : Mm - hmm .", "E : And so when you do your", "E : wh I 'm what is the task for the class project ?", "E : To come up with the phones ?", "A : Um", "E : or to come up with these vectors to see how closely they match the phones ,", "A : Oh .", "A : Right ,", "A : um to come up with a mapping from um MFCC 's or s some feature set , um to uh w to whether there 's existence of a particular phonological feature .", "E : or ?", "E : Mm - hmm .", "A : And um", "A : yeah ,", "A : basically it 's to learn a mapping from from the MFCC 's to uh phonological features .", "A : Is it did that answer your question ?", "E : I think so .", "A : OK .", "E : I guess I mean , uh I 'm not sure what you what you 're what you get out of your system .", "A : C", "E : Do you get out a uh a vector of these ones and zeros and then try to find the closest matching phoneme to that vector ,", "A : Mm - hmm .", "E : or ?", "A : Oh .", "A : No , no .", "A : I 'm not I 'm not planning to do any any phoneme mapping yet .", "A : Just it 's it 's basically it 's it 's really simple , basically a detection of phonological features .", "E : Uh - huh .", "E : I see .", "A : Yeah ,", "A : and um cuz the uh ", "A : So King and and Taylor um did this with uh recurrent neural nets ,", "E : Yeah .", "E : Mm - hmm .", "A : and this i their their idea was to first find a mapping from MFCC 's to uh phonological features", "A : and then later on , once you have these phonological features , then uh map that to phones .", "E : Mm - hmm .", "A : So I 'm I 'm sort of reproducing phase one of their stuff .", "E : Mmm .", "E : So they had one recurrent net for each particular feature ?", "A : Right .", "A : Right . Right . Right .", "E : I see .", "E : I wo did they compare that I mean , what if you just did phone recognition and did the reverse lookup .", "A : Uh .", "E : So you recognize a phone and which ever phone was recognized , you spit out it 's vector of ones and zeros .", "A : Mm - hmm .", "C : I expect you could do that .", "E : I mean uh ", "A : Uh .", "C : That 's probably not what he 's going to do on his class project .", "E : Yeah .", "C : Yeah .", "E : No .", "C : So um have you had a chance to do this um thing we talked about yet with the uh ", "A : Yeah .", "C : um", "E : Insertion penalty ?", "C : Uh . No actually I was going a different ", "C : That 's a good question , too ,", "C : but I was gonna ask about the the um changes to the data in comparing PLP and mel cepstrum for the SRI system .", "E : Uh .", "E : Well what I 've been ", "E : \" Changes to the data \" ,", "E : I 'm not sure I ", "C : Right .", "C : So we talked on the phone about this , that that there was still a difference of a of a few percent", "E : Yeah .", "E : Right .", "C : and you told me that there was a difference in how the normalization was done .", "C : And I was asking if you were going to do redo it uh for PLP with the normalization done as it had been done for the mel cepstrum .", "E : Mm - hmm .", "E : Uh right ,", "E : no I haven't had a chance to do that .", "E : What I 've been doing is uh trying to figure out ", "C : OK .", "E : it just seems to me like there 's a um ", "E : well it seems like there 's a bug ,", "E : because the difference in performance is it 's not gigantic", "E : but it 's big enough that it it seems wrong .", "C : Yeah ,", "C : I agree ,", "E : and ", "C : but I thought that the normalization difference was one of the possibilities ,", "E : Yeah , but I don't I 'm not ", "C : right ?", "E : Yeah ,", "E : I guess I don't think that the normalization difference is gonna account for everything .", "E : So what I was working on is um just going through and checking the headers of the wavefiles ,", "C : OK .", "E : to see if maybe there was a um a certain type of compression or something that was done that my script wasn't catching .", "E : So that for some subset of the training data , uh the the the features I was computing were junk .", "C : OK .", "E : Which would you know cause it to perform OK ,", "E : but uh , you know , the the models would be all messed up .", "E : So I was going through and just double - checking that kind of think first , to see if there was just some kind of obvious bug in the way that I was computing the features .", "C : Mm - hmm .", "C : I see .", "C : OK .", "E : Looking at all the sampling rates to make sure all the sampling rates were what eight K , what I was assuming they were ,", "C : Yeah .", "E : um ", "C : Yeah , that makes sense , to check all that .", "E : Yeah .", "E : So I was doing that first , before I did these other things , just to make sure there wasn't something ", "C : Although really , uh uh , a couple three percent uh difference in word error rate uh could easily come from some difference in normalization , I would think .", "C : But", "E : Yeah ,", "E : and I think , hhh I 'm trying to remember but I think I recall that Andreas was saying that he was gonna run sort of the reverse experiment .", "E : Uh which is to try to emulate the normalization that we did", "E : but with the mel cepstral features .", "E : Sort of , you know , back up from the system that he had .", "E : I thought he said he was gonna ", "E : I have to look back through my my email from him .", "C : Yeah , he 's probably off at at uh his meeting now ,", "E : Yeah ,", "E : he 's gone now .", "E : Um .", "C : yeah .", "C : Yeah .", "C : But", "C : yeah", "C : the I sh think they should be roughly equivalent ,", "E : But ", "C : um", "C : I mean again the Cambridge folk found the PLP actually to be a little better .", "E : Right .", "C : Uh So it 's um", "C : I mean the other thing I wonder about was whether there was something just in the the bootstrapping of their system which was based on ", "C : but maybe not , since they ", "E : Yeah see one thing that 's a little bit um ", "E : I was looking I 've been studying and going through the logs for the system that um Andreas created .", "E : And um his uh the way that the S R I system looks like it works is that it reads the wavefiles directly ,", "E : uh and does all of the cepstral computation stuff on the fly .", "C : Right .", "C : Right .", "E : And , so there 's no place where these where the cepstral files are stored , anywhere that I can go look at and compare to the PLP ones ,", "E : so whereas with our features , he 's actually storing the cepstrum on disk , and he reads those in .", "C : Right .", "E : But it looked like he had to give it uh", "E : even though the cepstrum is already computed , he has to give it uh a front - end parameter file .", "E : Which talks about the kind of uh com computation that his mel cepstrum thing does ,", "C : Uh - huh .", "E : so i I I don't know if that it probably doesn't mess it up ,", "E : it probably just ignores it if it determines that it 's already in the right format or something", "E : but the the the two processes that happen are a little different .", "E : So .", "C : Yeah .", "C : So anyway , there 's stuff there to sort out .", "E : Yeah .", "E : Yeah .", "C : So ,", "C : OK .", "C : Let 's go back to what you thought I was asking you .", "E : Yeah", "E : no and I didn't have a chance to do that .", "C : Ha !", "C : Oh !", "C : You had the sa same answer anyway .", "E : Yeah .", "E : Yeah . I 've been um , I 've been working with um Jeremy on his project", "E : and then I 've been trying to track down this bug in uh the ICSI front - end features .", "C : Uh - huh .", "E : So one thing that I did notice , yesterday I was studying the um the uh RASTA code", "C : Uh - huh .", "E : and it looks like we don't have any way to um control the frequency range that we use in our analysis .", "E : We basically it looks to me like we do the FFT , um and then we just take all the bins", "E : and we use everything .", "E : We don't have any set of parameters where we can say you know , \" only process from you know a hundred and ten hertz to thirty - seven - fifty \" .", "C : Um ", "E : At least I couldn't see any kind of control for that .", "C : Yeah , I don't think it 's in there ,", "C : I think it 's in the uh uh uh the filters .", "C : So , the F F T is on everything ,", "C : but the filters", "C : um ,", "C : for instance , ignore the the lowest bins and the highest bins .", "C : And what it does is it it copies", "E : The the filters ?", "E : Which filters ?", "C : um", "C : The filter bank which is created by integrating over F F T bins .", "E : Mm - hmm .", "C : um", "E : When you get the mel When you go to the mel scale .", "C : Right .", "C : Yeah ,", "C : it 's bark scale ,", "C : and it 's it it um it actually copies the uh um the second filters over to the first .", "C : So the first filters are always ", "C : and you can s you can specify a different number of uh features different number of filters ,", "C : I think ,", "C : as I recall .", "C : So you can specify a different number of filters , and whatever um uh you specify , the last ones are gonna be ignored .", "C : So that that 's a way that you sort of change what the what the bandwidth is .", "C : Y you can't do it without I think changing the number of filters ,", "E : I saw something about uh that looked like it was doing something like that , but I didn't quite understand it .", "C : but ", "E : So maybe ", "C : Yeah , so the idea is that the very lowest frequencies and and typically the veriest highest frequencies are kind of junk .", "E : Uh - huh .", "C : And so um", "C : you just for continuity you just approximate them by by the second to highest and second to lowest .", "E : Mm - hmm .", "C : It 's just a simple thing we put in .", "C : And and so if you h", "E : But so the but that 's a fixed uh thing ?", "E : There 's nothing that lets you ", "C : Yeah , I think that 's a fixed thing .", "C : But see see my point ?", "C : If you had If you had ten filters , then you would be throwing away a lot at the two ends .", "E : Mm - hmm .", "C : And if you had if you had fifty filters , you 'd be throwing away hardly anything .", "E : Mm - hmm .", "C : Um , I don't remember there being an independent way of saying \" we 're just gonna make them from here to here \" .", "E : Use this analysis bandwidth or something .", "C : But I I I don't know ,", "C : it 's actually been awhile since I 've looked at it .", "E : Yeah , I went through the Feacalc code and then looked at you know just calling the RASTA libs and thing like that .", "E : And I didn't I couldn't see any wh place where that kind of thing was done .", "E : But um I didn't quite understand everything that I saw ,", "C : Yeah , see I don't know Feacalc at all .", "E : so ", "E : Mm - hmm .", "C : But it calls RASTA with some options ,", "C : and um", "E : Right .", "C : But I I think in ", "C : I don't know .", "C : I guess for some particular database you might find that you could tune that and tweak that to get that a little better ,", "C : but I think that in general it 's not that critical .", "C : I mean there 's ", "E : Yeah .", "C : You can You can throw away stuff below a hundred hertz or so", "C : and it 's just not going to affect phonetic classification at all .", "E : Another thing I was thinking about was um is there a ", "E : I was wondering if there 's maybe um certain settings of the parameters when you compute PLP which would basically cause it to output mel cepstrum .", "E : So that , in effect , what I could do is use our code", "E : but produce mel cepstrum", "E : and compare that directly to ", "C : Well , it 's not precisely .", "C : Yeah .", "E : Hmm .", "C : I mean ,", "C : um , um", "C : what you can do is um you can definitely change the the filter bank from being uh a uh trapezoidal integration to a a a triangular one ,", "E : Mm - hmm .", "C : which is what the typical mel mel cepstral uh filter bank does .", "E : Mm - hmm .", "C : And some people have claimed that they got some better performance doing that ,", "C : so you certainly could do that easily .", "C : But the fundamental difference ,", "C : I mean , there 's other small differences ", "E : There 's a cubic root that happens ,", "E : right ?", "C : Yeah ,", "C : but , you know , as opposed to the log in the other case .", "C : I mean the fundamental d d difference that we 've seen any kind of difference from before , which is actually an advantage for the P L P i uh , I think , is that the the smoothing at the end is auto - regressive instead of being cepstral uh , from cepstral truncation .", "C : So um it 's a little more noise robust .", "E : Hmm .", "C : Um , and that 's that 's why when people started getting databases that had a little more noise in it , like like uh um Broadcast News and so on ,", "C : that 's why c Cambridge switched to PLP I think .", "E : Mm - hmm .", "C : So um", "C : That 's a difference that I don't think we put any way to get around ,", "C : since it was an advantage .", "E : Mm - hmm .", "C : um uh", "C : but we did eh we did hear this comment from people at some point , that um it uh they got some better results with the triangular filters rather than the trapezoidal .", "C : So that is an option in RASTA .", "E : Hmm .", "C : Uh and you can certainly play with that .", "C : But I think you 're probably doing the right thing to look for bugs first .", "E : Yeah just it just seems like this kind of behavior could be caused by you know s some of the training data being messed up .", "C : I don't know .", "C : Could be .", "E : You know , you 're sort of getting most of the way there , but there 's a ", "E : So I started going through and looking ", "E : One of the things that I did notice was that the um log likelihoods coming out of the log recognizer from the PLP data were much lower , much smaller ,", "E : than for the mel cepstral stuff , and that the average amount of pruning that was happening was therefore a little bit higher for the PLP features .", "C : Oh - huh !", "E : So , since he used the same exact pruning thresholds for both , I was wondering if it could be that we 're getting more pruning .", "C : Oh !", "C : He he He used the identical pruning thresholds even though the s the range of p of the likeli", "E : Yeah .", "C : Oh well that 's That 's a pretty good point right there .", "E : Right .", "E : Right .", "C : Yeah .", "E : Yeah ,", "E : so ", "C : I would think that you might wanna do something like uh you know , look at a few points to see where you are starting to get significant search errors .", "E : That 's ", "E : Right .", "E : Well , what I was gonna do is I was gonna take um a couple of the utterances that he had run through ,", "E : then run them through again", "E : but modify the pruning threshold and see if it you know , affects the score .", "C : Yeah .", "C : Yeah .", "E : So .", "C : But I mean you could uh if if if that looks promising you could , you know , r uh run the overall test set with a with a few different uh pruning thresholds for both ,", "E : Mm - hmm .", "E : Right .", "C : and presumably he 's running at some pruning threshold that 's that 's uh , you know gets very few search errors", "C : but is is relatively fast", "E : Mm - hmm .", "E : Right . I mean , yeah , generally in these things you you turn back pruning really far ,", "C : and ", "E : so I I didn't think it would be that big a deal", "E : because I was figuring well you have it turned back so far that you know it ", "C : But you may be in the wrong range for the P L P features for some reason .", "E : Yeah .", "E : Yeah .", "E : Yeah .", "E : And the uh the the run time of the recognizer on the PLP features is longer", "E : which sort of implies that the networks are bushier ,", "E : you know , there 's more things it 's considering", "E : which goes along with the fact that the matches aren't as good .", "E : So uh , you know , it could be that we 're just pruning too much .", "E : So .", "C : Yeah .", "C : Yeah , maybe just be different kind of distributions", "C : and and", "E : Mm - hmm .", "C : yeah", "C : so that 's another possible thing .", "E : Mm - hmm .", "C : They they should really shouldn't ", "C : There 's no particular reason why they would be exactly behave exactly the same .", "E : Mm - hmm . Right .", "E : Right .", "C : So .", "E : So .", "E : There 's lots of little differences .", "E : So .", "C : Yeah .", "E : Uh .", "C : Yeah .", "E : Trying to track it down .", "C : Yeah .", "C : I guess this was a little bit off topic , I guess ,", "E : Yeah", "C : because I was I was thinking in terms of th this as being a a a a core item that once we once we had it going we would use for a number of the front - end things also .", "E : Mm - hmm .", "C : So .", "C : um", "C : Wanna ", "B : That 's as far as my stuff goes ,", "C : What 's what 's on ", "B : yeah ,", "C : Yeah .", "B : well I tried this mean subtraction method .", "B : Um .", "B : Due to Avendano , I 'm taking s um six seconds of speech ,", "B : um I 'm using two second FFT analysis frames , stepped by a half second", "B : so it 's a quarter length step", "B : and I I take that frame and four f the four I take ", "B : Sorry ,", "B : I take the current frame and the four past frames and the four future frames", "B : and that adds up to six seconds of speech .", "B : And I calculate um the spectral mean , of the log magnitude spectrum over that N .", "B : I use that to normalize the s the current center frame by mean subtraction .", "B : And I then then I move to the next frame and I I do it again .", "B : Well , actually I calculate all the means first", "B : and then I do the subtraction .", "B : And um the I tried that with HDK ,", "B : the Aurora setup of HDK training on clean TI - digits ,", "B : and um it it helped", "B : um in a phony reverberation case", "B : um where I just used the simulated impulse response um the error rate went from something like eighty it was from something like eighteen percent to um four percent .", "B : And on meeting rec recorder far mike digits , mike on channel F , it went from um forty - one percent error to eight percent error .", "E : On on the real data , not with artificial reverb ?", "B : Right .", "E : Uh - huh .", "B : And that that was um trained on clean speech only ,", "B : which I 'm guessing is the reason why the baseline was so bad .", "B : And ", "C : That 's ac actually a little side point is I think that 's the first results that we have uh uh uh of any sort on the far field uh on on the far field data uh for recorded in in meetings .", "B : Oh um actually um Adam ran the SRI recognizer .", "C : Did he ?", "C : On the near field , on the ne", "B : On the far field also .", "B : He did one PZM channel and one PDA channel .", "C : Oh did he ?", "C : Oh !", "C : I didn't recall that .", "C : What kind of numbers was he getting with that ?", "B : I I 'm not sure ,", "B : I think it was about five percent error for the PZM channel .", "C : Five .", "B : f I think .", "B : Yeah .", "C : So why were you getting forty - one here ?", "C : Is this ", "B : Um .", "B : I I 'm g I 'm guessing it was the the training data .", "B : Uh , clean TI - digits is , like , pretty pristine training data ,", "B : and if they trained the SRI system on this TV broadcast type stuff ,", "B : I think it 's a much wider range of channels", "B : and it ", "C : No ,", "C : but wait a minute .", "C : I I I th I think he ", "C : What am I saying here ?", "C : Yeah ,", "C : so that was the SRI system .", "C : Maybe you 're right .", "C : Yeah .", "C : Cuz it was getting like one percent So it 's still this kind of ratio .", "C : It was it was getting one percent or something on the near field .", "C : Wasn't it ?", "E : Mm - hmm ,", "E : or it wa a it was around one .", "C : Yeah .", "C : Yeah . I think it was getting around one percent for the near for the n for the close mike .", "E : Yeah .", "B : Huh ?", "C : So it was like one to five ", "B : OK .", "C : So it 's still this kind of ratio .", "C : It 's just ", "C : yeah ,", "C : it 's a lot more training data .", "C : So", "C : So probably it should be something we should try then is to is to see if is at some point just to take i to transform the data", "C : and then and then uh use th use it for the SRI system .", "B : b You me you mean um ta", "C : So you 're so you have a system which for one reason or another is relatively poor ,", "B : Yeah .", "C : and and uh you have something like forty - one percent error", "C : uh and then you transform it to eight by doing doing this this work .", "C : Um .", "C : So here 's this other system , which is a lot better ,", "C : but there 's still this kind of ratio .", "C : It 's something like five percent error with the the distant mike ,", "C : and one percent with the close mike .", "B : OK .", "C : So the question is how close to that one can you get if you transform the data using that system .", "B : r Right , so so I guess this SRI system is trained on a lot of s Broadcast News or Switchboard data .", "C : Yeah .", "B : Is that right ?", "B : Do you know which one it is ?", "E : It 's trained on a lot of different things .", "E : Um .", "E : It 's trained on uh a lot of Switchboard , Call Home ,", "B : Uh - huh .", "E : um a bunch of different sources ,", "E : some digits ,", "E : there 's some digits training in there .", "B : OK .", "B : O one thing I 'm wondering about is what this mean subtraction method um will do if it 's faced with additive noise .", "A : Hmm .", "B : Cuz I I it 's cuz I don't know what log magnitude spectral subtraction is gonna do to additive noise .", "C : Yeah ,", "C : well , it 's it 's not exactly the right thing", "B : That 's that 's the ", "B : Uh - huh .", "C : but uh but you 've already seen that cuz there is added noise here .", "B : That 's that 's Yeah , that 's true .", "C : Yeah .", "B : That 's a good point .", "C : So um ", "B : OK , so it 's then then it 's it 's it 's reasonable to expect it would be helpful if we used it with the SRI system", "B : and", "C : Yeah ,", "C : I mean , as helpful ", "C : I mean ,", "C : so that 's the question .", "C : Yeah , w we 're often asked this when we work with a system that that isn't isn't sort of industry industry standard great ,", "B : Uh - huh .", "C : uh and we see some reduction in error using some clever method ,", "C : then , you know , will it work on a on a on a good system .", "C : So uh", "C : you know , this other one 's it was a pretty good system .", "C : I think , you know , one one percent word error rate on digits is uh digit strings is not uh you know stellar ,", "C : but but given that this is real digits , as opposed to uh sort of laboratory ", "B : Mm - hmm .", "C : Well .", "E : And it wasn't trained on this task either .", "C : And it wasn't trained on this task .", "C : Actually one percent is sort of you know , sort of in a reasonable range .", "C : People would say \" yeah , I could I can imagine getting that \" .", "B : Mm - hmm .", "C : And uh so the the four or five percent or something is is is quite poor .", "B : Mm - hmm .", "C : Uh , you know , if you 're doing a uh a sixteen digit uh credit card number you 'll basically get it wrong almost all the time .", "C : So . So . Uh , um", "B : Hmm .", "C : a significant reduction in the error for that would be great .", "B : Huh , OK .", "C : And and then , uh", "C : Yeah .", "C : So .", "C : Yeah .", "C : Cool .", "B : Sounds good .", "C : Yeah .", "C : Alright , um , I actually have to run .", "C : So I don't think I can do the digits ,", "C : but um , I guess I 'll leave my microphone on ?", "E : Uh , yeah .", "C : Yeah .", "C : Thank you .", "E : Yep .", "E : Yeah .", "E : That 'll work .", "C : I can be out of here quickly . That 's", "C : I just have to run for another appointment . OK ,", "C : I t Yeah . I left it on .", "C : OK ." ], "abstractive summary": [ { "id": "Bro018.s.1", "text": "The ICSI Meeting Recorder Group met once more to discuss their recent progress in various projects, as well as discuss some of the issues that have arisen in the last week.", "type": "abstract" }, { "id": "Bro018.s.2", "text": "There has been further work on voiced/unvoiced detection, along with spectral subtraction.", "type": "abstract" }, { "id": "Bro018.s.3", "text": "The group discussed one members attendance at a conference, and another groups code, which is proving hard to follow.", "type": "abstract" }, { "id": "Bro018.s.4", "text": "In trying to understand France Telecom's code, there does not appear to be a reason for a particular constant value.", "type": "problems" }, { "id": "Bro018.s.5", "text": "Speaker fn002 reported on her work with mn007, who was absent attending a conference.", "type": "progress" }, { "id": "Bro018.s.6", "text": "They have been working on voiced/unvoiced detection, and have tried two different configuration of MLP, although results have not been improved greatly.", "type": "progress" }, { "id": "Bro018.s.7", "text": "She has also begun looking at France Telecom's recogniser code.", "type": "progress" }, { "id": "Bro018.s.8", "text": "Speaker me006 has been coming up with tests for his work using acoustic events.", "type": "progress" }, { "id": "Bro018.s.9", "text": "Speaker me018 is investigating differences in another system, and feels it maybe down to bugs.", "type": "progress" }, { "id": "Bro018.s.10", "text": "Speaker me026 has made real progress on his spectral subtraction work, with an implementation and early results.", "type": "progress" } ], "extractive summary": [ "C : So uh , he 's not here ,", "D : Yeah , I will try to explain the thing that I did this this week during this week .", "D : Well eh you know that I work I begin to work with a new feature to detect voice - unvoice .", "D : What I trying two MLP to to the with this new feature and the fifteen feature uh from the eh bus base system", "D : And I tried to do some experiment of recognition with that", "D : And , well , the result are li a little bit better , but more or less similar .", "D : And I 'm trying two MLP , one one that only have t three output ,", "D : voice , unvoice , and silence ,", "D : and other one that have fifty - six output .", "D : The probabilities of the allophone .", "D : and only have result with with the MLP with the three output .", "D : and also mmm I H Hynek last week say that if I have time I can to begin to to study", "D : well seriously the France Telecom proposal", "D : to look at the code", "D : I begin to to work also in that .", "D : But the first thing that I don't understand is that they are using R - the uh log energy that this quite ", "D : I don't know why they have some constant in the expression of the lower energy .", "C : OK , and wh when did Stephane take off ?", "D : I think that Stephane will arrive today or tomorrow .", "C : So he 's he 's going to ICASSP which is good .", "C : Wanna talk a little bit about what we were talking about this morning ?", "A : Then uh I talked a little bit about um continuing with these dynamic ev um acoustic events ,", "A : and um we 're we 're we 're thinking about a way to test the completeness of a a set of um dynamic uh events .", "A : Uh , completeness in the in the sense that um if we if we pick these X number of acoustic events , do they provide sufficient coverage for the phones that we 're trying to recognize or or the f the words that we 're gonna try to recognize later on .", "A : And so Morgan and I were uh discussing um s uh s a form of a cheating experiment where we get um we have uh um a chosen set of features , or acoustic events ,", "A : and we train up a hybrid um system to do phone recognition on TIMIT .", "C : So um have you had a chance to do this um thing we talked about yet with the uh ", "C : but I was gonna ask about the the um changes to the data in comparing PLP and mel cepstrum for the SRI system .", "C : So we talked on the phone about this , that that there was still a difference of a of a few percent", "C : and you told me that there was a difference in how the normalization was done .", "C : And I was asking if you were going to do redo it uh for PLP with the normalization done as it had been done for the mel cepstrum .", "E : no I haven't had a chance to do that .", "E : What I 've been doing is uh trying to figure out ", "E : it just seems to me like there 's a um ", "E : well it seems like there 's a bug ,", "E : because the difference in performance is it 's not gigantic", "E : but it 's big enough that it it seems wrong .", "C : but I thought that the normalization difference was one of the possibilities ,", "E : I guess I don't think that the normalization difference is gonna account for everything .", "E : So what I was working on is um just going through and checking the headers of the wavefiles ,", "B : That 's as far as my stuff goes ,", "B : well I tried this mean subtraction method .", "B : Due to Avendano , I 'm taking s um six seconds of speech ,", "B : um I 'm using two second FFT analysis frames , stepped by a half second", "B : And I calculate um the spectral mean , of the log magnitude spectrum over that N .", "B : I use that to normalize the s the current center frame by mean subtraction .", "B : And um the I tried that with HDK ,", "B : the Aurora setup of HDK training on clean TI - digits ,", "B : and um it it helped", "B : um in a phony reverberation case", "B : um where I just used the simulated impulse response um the error rate went from something like eighty it was from something like eighteen percent to um four percent .", "B : And on meeting rec recorder far mike digits , mike on channel F , it went from um forty - one percent error to eight percent error ." ] }, "log": [] } }