id
stringlengths
9
12
query
stringlengths
665
1.43k
answer
stringlengths
3
1.21k
label
sequencelengths
1
138
token
sequencelengths
1
138
finerord700
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: On the other hand , the improving economy might be more than enough to offset the lowering of affordability levels , which as mentioned above , still remain relatively high , compared with historical levels . Answer:
On:O the:O other:O hand:O ,:O the:O improving:O economy:O might:O be:O more:O than:O enough:O to:O offset:O the:O lowering:O of:O affordability:O levels:O ,:O which:O as:O mentioned:O above:O ,:O still:O remain:O relatively:O high:O ,:O compared:O with:O historical:O levels:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "On", "the", "other", "hand", ",", "the", "improving", "economy", "might", "be", "more", "than", "enough", "to", "offset", "the", "lowering", "of", "affordability", "levels", ",", "which", "as", "mentioned", "above", ",", "still", "remain", "relatively", "high", ",", "compared", "with", "historical", "levels", "." ]
finerord701
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Trade , Job Openings , and CoreLogic Home Prices on Tap for Next Week Answer:
Trade:O ,:O Job:O Openings:O ,:O and:O CoreLogic:B-ORG Home:O Prices:O on:O Tap:O for:O Next:O Week:O
[ "O", "O", "O", "O", "O", "O", "B-ORG", "O", "O", "O", "O", "O", "O", "O" ]
[ "Trade", ",", "Job", "Openings", ",", "and", "CoreLogic", "Home", "Prices", "on", "Tap", "for", "Next", "Week" ]
finerord702
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We do n't necessarily view the trade report as terribly important as it remains in a relatively narrow channel and exports comprise just 13 % or so of GDP . Answer:
We:O do:O n't:O necessarily:O view:O the:O trade:O report:O as:O terribly:O important:O as:O it:O remains:O in:O a:O relatively:O narrow:O channel:O and:O exports:O comprise:O just:O 13:O %:O or:O so:O of:O GDP:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "We", "do", "n't", "necessarily", "view", "the", "trade", "report", "as", "terribly", "important", "as", "it", "remains", "in", "a", "relatively", "narrow", "channel", "and", "exports", "comprise", "just", "13", "%", "or", "so", "of", "GDP", "." ]
finerord703
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: That does n't prevent violent quarter - to - quarter swings . Answer:
That:O does:O n't:O prevent:O violent:O quarter:O -:O to:O -:O quarter:O swings:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "That", "does", "n't", "prevent", "violent", "quarter", "-", "to", "-", "quarter", "swings", "." ]
finerord704
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Those wild moves are the biggest impediment , with the possible exception of inventories , to forecasting quarterly GDP . Answer:
Those:O wild:O moves:O are:O the:O biggest:O impediment:O ,:O with:O the:O possible:O exception:O of:O inventories:O ,:O to:O forecasting:O quarterly:O GDP:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Those", "wild", "moves", "are", "the", "biggest", "impediment", ",", "with", "the", "possible", "exception", "of", "inventories", ",", "to", "forecasting", "quarterly", "GDP", "." ]
finerord705
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Combined imports and exports took a full 2 % off of first - quarter GDP growth . Answer:
Combined:O imports:O and:O exports:O took:O a:O full:O 2:O %:O off:O of:O first:O -:O quarter:O GDP:O growth:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Combined", "imports", "and", "exports", "took", "a", "full", "2", "%", "off", "of", "first", "-", "quarter", "GDP", "growth", "." ]
finerord706
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The best guess right now is that trade will have no impact on the second quarter . Answer:
The:O best:O guess:O right:O now:O is:O that:O trade:O will:O have:O no:O impact:O on:O the:O second:O quarter:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "best", "guess", "right", "now", "is", "that", "trade", "will", "have", "no", "impact", "on", "the", "second", "quarter", "." ]
finerord707
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: At this moment that is just a guess , presuming that the horrible slippage in the first quarter was almost entirely due to West Coast port labor actions that will reverse out in the second quarter . Answer:
At:O this:O moment:O that:O is:O just:O a:O guess:O ,:O presuming:O that:O the:O horrible:O slippage:O in:O the:O first:O quarter:O was:O almost:O entirely:O due:O to:O West:O Coast:O port:O labor:O actions:O that:O will:O reverse:O out:O in:O the:O second:O quarter:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "At", "this", "moment", "that", "is", "just", "a", "guess", ",", "presuming", "that", "the", "horrible", "slippage", "in", "the", "first", "quarter", "was", "almost", "entirely", "due", "to", "West", "Coast", "port", "labor", "actions", "that", "will", "reverse", "out", "in", "the", "second", "quarter", "." ]
finerord708
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: April trade data supported that case and now the question is , will the May data continue the trend ? Answer:
April:O trade:O data:O supported:O that:O case:O and:O now:O the:O question:O is:O ,:O will:O the:O May:O data:O continue:O the:O trend:O ?:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "April", "trade", "data", "supported", "that", "case", "and", "now", "the", "question", "is", ",", "will", "the", "May", "data", "continue", "the", "trend", "?" ]
finerord709
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: For now , the consensus is for a modest worsening of the trade deficit in May from April 's improved $ 40.9 billion . Answer:
For:O now:O ,:O the:O consensus:O is:O for:O a:O modest:O worsening:O of:O the:O trade:O deficit:O in:O May:O from:O April:O 's:O improved:O $:O 40.9:O billion:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "For", "now", ",", "the", "consensus", "is", "for", "a", "modest", "worsening", "of", "the", "trade", "deficit", "in", "May", "from", "April", "'s", "improved", "$", "40.9", "billion", "." ]
finerord710
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: If that proves to be correct , the trade deficit for the second quarter could be a slight negative for the second quarter instead of completely neutral . Answer:
If:O that:O proves:O to:O be:O correct:O ,:O the:O trade:O deficit:O for:O the:O second:O quarter:O could:O be:O a:O slight:O negative:O for:O the:O second:O quarter:O instead:O of:O completely:O neutral:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "If", "that", "proves", "to", "be", "correct", ",", "the", "trade", "deficit", "for", "the", "second", "quarter", "could", "be", "a", "slight", "negative", "for", "the", "second", "quarter", "instead", "of", "completely", "neutral", "." ]
finerord711
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: With the employment data relatively soft , the job openings report is likely to get a fair amount of attention next week . Answer:
With:O the:O employment:O data:O relatively:O soft:O ,:O the:O job:O openings:O report:O is:O likely:O to:O get:O a:O fair:O amount:O of:O attention:O next:O week:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "With", "the", "employment", "data", "relatively", "soft", ",", "the", "job", "openings", "report", "is", "likely", "to", "get", "a", "fair", "amount", "of", "attention", "next", "week", "." ]
finerord712
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The data for April showed record openings that did n't really translate into strong jobs growth . Answer:
The:O data:O for:O April:O showed:O record:O openings:O that:O did:O n't:O really:O translate:O into:O strong:O jobs:O growth:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "data", "for", "April", "showed", "record", "openings", "that", "did", "n't", "really", "translate", "into", "strong", "jobs", "growth", "." ]
finerord713
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: With job gains relatively sparse in May and June , openings could remain above the 5 million level . Answer:
With:O job:O gains:O relatively:O sparse:O in:O May:O and:O June:O ,:O openings:O could:O remain:O above:O the:O 5:O million:O level:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "With", "job", "gains", "relatively", "sparse", "in", "May", "and", "June", ",", "openings", "could", "remain", "above", "the", "5", "million", "level", "." ]
finerord714
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: If openings stay sky - high , it would suggest that job growth is not being limited by employers ' desire to hire but by the lack of employees with the proper skills , which is an entirely different problem from the Fed is trying to fight with their policies . Answer:
If:O openings:O stay:O sky:O -:O high:O ,:O it:O would:O suggest:O that:O job:O growth:O is:O not:O being:O limited:O by:O employers:O ':O desire:O to:O hire:O but:O by:O the:O lack:O of:O employees:O with:O the:O proper:O skills:O ,:O which:O is:O an:O entirely:O different:O problem:O from:O the:O Fed:B-ORG is:O trying:O to:O fight:O with:O their:O policies:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "If", "openings", "stay", "sky", "-", "high", ",", "it", "would", "suggest", "that", "job", "growth", "is", "not", "being", "limited", "by", "employers", "'", "desire", "to", "hire", "but", "by", "the", "lack", "of", "employees", "with", "the", "proper", "skills", ",", "which", "is", "an", "entirely", "different", "problem", "from", "the", "Fed", "is", "trying", "to", "fight", "with", "their", "policies", "." ]
finerord715
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: With affordability issues heating up , as noted above , we will be watching the CoreLogic home prices indexes with a more careful eye . Answer:
With:O affordability:O issues:O heating:O up:O ,:O as:O noted:O above:O ,:O we:O will:O be:O watching:O the:O CoreLogic:B-ORG home:O prices:O indexes:O with:O a:O more:O careful:O eye:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "With", "affordability", "issues", "heating", "up", ",", "as", "noted", "above", ",", "we", "will", "be", "watching", "the", "CoreLogic", "home", "prices", "indexes", "with", "a", "more", "careful", "eye", "." ]
finerord716
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: These reports have showed stronger price growth than either the FHFA or the Case - Shiller data , as noted earlier . Answer:
These:O reports:O have:O showed:O stronger:O price:O growth:O than:O either:O the:O FHFA:O or:O the:O Case:O -:O Shiller:O data:O ,:O as:O noted:O earlier:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "These", "reports", "have", "showed", "stronger", "price", "growth", "than", "either", "the", "FHFA", "or", "the", "Case", "-", "Shiller", "data", ",", "as", "noted", "earlier", "." ]
finerord717
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: It will be very interesting to see how these data sets converge over the next month . Answer:
It:O will:O be:O very:O interesting:O to:O see:O how:O these:O data:O sets:O converge:O over:O the:O next:O month:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "It", "will", "be", "very", "interesting", "to", "see", "how", "these", "data", "sets", "converge", "over", "the", "next", "month", "." ]
finerord718
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: It is still a little surprising to us that a country the size of Greece , which what one commentator mentioned was about the economic size of Alabama , was moving world markets this much . Answer:
It:O is:O still:O a:O little:O surprising:O to:O us:O that:O a:O country:O the:O size:O of:O Greece:B-LOC ,:O which:O what:O one:O commentator:O mentioned:O was:O about:O the:O economic:O size:O of:O Alabama:B-LOC ,:O was:O moving:O world:O markets:O this:O much:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-LOC", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-LOC", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "It", "is", "still", "a", "little", "surprising", "to", "us", "that", "a", "country", "the", "size", "of", "Greece", ",", "which", "what", "one", "commentator", "mentioned", "was", "about", "the", "economic", "size", "of", "Alabama", ",", "was", "moving", "world", "markets", "this", "much", "." ]
finerord719
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: It does , however , say mountains about Europe 's ability to cope with problems and that there are potentially serious problems with the structure of the eurozone . Answer:
It:O does:O ,:O however:O ,:O say:O mountains:O about:O Europe:B-LOC 's:O ability:O to:O cope:O with:O problems:O and:O that:O there:O are:O potentially:O serious:O problems:O with:O the:O structure:O of:O the:O eurozone:B-LOC .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "B-LOC", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-LOC", "O" ]
[ "It", "does", ",", "however", ",", "say", "mountains", "about", "Europe", "'s", "ability", "to", "cope", "with", "problems", "and", "that", "there", "are", "potentially", "serious", "problems", "with", "the", "structure", "of", "the", "eurozone", "." ]
finerord720
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The short - term drama should settle down soon , but the longer - term structure of the European Union , with disparate countries with different growth rates tied to one rigid currency , could prove to be a real problem . Answer:
The:O short:O -:O term:O drama:O should:O settle:O down:O soon:O ,:O but:O the:O longer:O -:O term:O structure:O of:O the:O European:B-ORG Union:I-ORG ,:O with:O disparate:O countries:O with:O different:O growth:O rates:O tied:O to:O one:O rigid:O currency:O ,:O could:O prove:O to:O be:O a:O real:O problem:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "short", "-", "term", "drama", "should", "settle", "down", "soon", ",", "but", "the", "longer", "-", "term", "structure", "of", "the", "European", "Union", ",", "with", "disparate", "countries", "with", "different", "growth", "rates", "tied", "to", "one", "rigid", "currency", ",", "could", "prove", "to", "be", "a", "real", "problem", "." ]
finerord721
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: This week Morningstar Investment Management senior economist Francisco Torralba and corporate bond strategist Dave Sekera offered their take on the crisis and Sunday 's referendum . Answer:
This:O week:O Morningstar:B-ORG Investment:I-ORG Management:I-ORG senior:O economist:O Francisco:B-PER Torralba:I-PER and:O corporate:O bond:O strategist:O Dave:B-PER Sekera:I-PER offered:O their:O take:O on:O the:O crisis:O and:O Sunday:O 's:O referendum:O .:O
[ "O", "O", "B-ORG", "I-ORG", "I-ORG", "O", "O", "B-PER", "I-PER", "O", "O", "O", "O", "B-PER", "I-PER", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "This", "week", "Morningstar", "Investment", "Management", "senior", "economist", "Francisco", "Torralba", "and", "corporate", "bond", "strategist", "Dave", "Sekera", "offered", "their", "take", "on", "the", "crisis", "and", "Sunday", "'s", "referendum", "." ]
finerord722
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: For an interesting editorial on the longer - term context , we also think this editorial by Brett Arends of MarketWatch is worth a read . Answer:
For:O an:O interesting:O editorial:O on:O the:O longer:O -:O term:O context:O ,:O we:O also:O think:O this:O editorial:O by:O Brett:B-PER Arends:I-PER of:O MarketWatch:B-ORG is:O worth:O a:O read:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-PER", "I-PER", "O", "B-ORG", "O", "O", "O", "O", "O" ]
[ "For", "an", "interesting", "editorial", "on", "the", "longer", "-", "term", "context", ",", "we", "also", "think", "this", "editorial", "by", "Brett", "Arends", "of", "MarketWatch", "is", "worth", "a", "read", "." ]
finerord723
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: There was a lot of economic news for the week , but none of the reports were really game - changers . Answer:
There:O was:O a:O lot:O of:O economic:O news:O for:O the:O week:O ,:O but:O none:O of:O the:O reports:O were:O really:O game:O -:O changers:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "There", "was", "a", "lot", "of", "economic", "news", "for", "the", "week", ",", "but", "none", "of", "the", "reports", "were", "really", "game", "-", "changers", "." ]
finerord724
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Most of the reports showed a U.S . economy that continued to trudge along with no real acceleration or deceleration in growth . Answer:
Most:O of:O the:O reports:O showed:O a:O U.S:B-LOC .:I-LOC economy:O that:O continued:O to:O trudge:O along:O with:O no:O real:O acceleration:O or:O deceleration:O in:O growth:O .:O
[ "O", "O", "O", "O", "O", "O", "B-LOC", "I-LOC", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Most", "of", "the", "reports", "showed", "a", "U.S", ".", "economy", "that", "continued", "to", "trudge", "along", "with", "no", "real", "acceleration", "or", "deceleration", "in", "growth", "." ]
finerord725
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Everything continues to point to 2.0 % - 2.5 % growth . Answer:
Everything:O continues:O to:O point:O to:O 2.0:O %:O -:O 2.5:O %:O growth:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Everything", "continues", "to", "point", "to", "2.0", "%", "-", "2.5", "%", "growth", "." ]
finerord726
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Another reason for the big yawn for this week 's economic data is most economists were very close to the mark on a lot of indicators , which almost never happens . Answer:
Another:O reason:O for:O the:O big:O yawn:O for:O this:O week:O 's:O economic:O data:O is:O most:O economists:O were:O very:O close:O to:O the:O mark:O on:O a:O lot:O of:O indicators:O ,:O which:O almost:O never:O happens:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Another", "reason", "for", "the", "big", "yawn", "for", "this", "week", "'s", "economic", "data", "is", "most", "economists", "were", "very", "close", "to", "the", "mark", "on", "a", "lot", "of", "indicators", ",", "which", "almost", "never", "happens", "." ]
finerord727
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Auto sales hit 17.2 million units , exactly as analyst forecasts ( which were down month to month and up year to year ) . Answer:
Auto:O sales:O hit:O 17.2:O million:O units:O ,:O exactly:O as:O analyst:O forecasts:O (:O which:O were:O down:O month:O to:O month:O and:O up:O year:O to:O year:O ):O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Auto", "sales", "hit", "17.2", "million", "units", ",", "exactly", "as", "analyst", "forecasts", "(", "which", "were", "down", "month", "to", "month", "and", "up", "year", "to", "year", ")", "." ]
finerord728
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Auto industry growth , while nice to have , was n't nearly as strong as it was at the beginning of this recovery . Answer:
Auto:O industry:O growth:O ,:O while:O nice:O to:O have:O ,:O was:O n't:O nearly:O as:O strong:O as:O it:O was:O at:O the:O beginning:O of:O this:O recovery:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Auto", "industry", "growth", ",", "while", "nice", "to", "have", ",", "was", "n't", "nearly", "as", "strong", "as", "it", "was", "at", "the", "beginning", "of", "this", "recovery", "." ]
finerord729
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Employment growth for June also matched the consensus forecast with 223,000 jobs added . Answer:
Employment:O growth:O for:O June:O also:O matched:O the:O consensus:O forecast:O with:O 223,000:O jobs:O added:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Employment", "growth", "for", "June", "also", "matched", "the", "consensus", "forecast", "with", "223,000", "jobs", "added", "." ]
finerord730
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Some previous months of employment data were also restated lower . Answer:
Some:O previous:O months:O of:O employment:O data:O were:O also:O restated:O lower:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Some", "previous", "months", "of", "employment", "data", "were", "also", "restated", "lower", "." ]
finerord731
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: All of this brings employment growth and GDP growth into closer alignment . Answer:
All:O of:O this:O brings:O employment:O growth:O and:O GDP:O growth:O into:O closer:O alignment:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "All", "of", "this", "brings", "employment", "growth", "and", "GDP", "growth", "into", "closer", "alignment", "." ]
finerord732
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We knew the relative growth rates ( high employment growth , lower GDP growth ) did n't make any sense at the beginning of the year . Answer:
We:O knew:O the:O relative:O growth:O rates:O (:O high:O employment:O growth:O ,:O lower:O GDP:O growth:O ):O did:O n't:O make:O any:O sense:O at:O the:O beginning:O of:O the:O year:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "We", "knew", "the", "relative", "growth", "rates", "(", "high", "employment", "growth", ",", "lower", "GDP", "growth", ")", "did", "n't", "make", "any", "sense", "at", "the", "beginning", "of", "the", "year", "." ]
finerord733
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We expected a big blowout event to bring the numbers back in line . Answer:
We:O expected:O a:O big:O blowout:O event:O to:O bring:O the:O numbers:O back:O in:O line:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "We", "expected", "a", "big", "blowout", "event", "to", "bring", "the", "numbers", "back", "in", "line", "." ]
finerord734
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Instead , we got a slow , irregular drift back down combined with some minor revisions to the employment data . Answer:
Instead:O ,:O we:O got:O a:O slow:O ,:O irregular:O drift:O back:O down:O combined:O with:O some:O minor:O revisions:O to:O the:O employment:O data:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Instead", ",", "we", "got", "a", "slow", ",", "irregular", "drift", "back", "down", "combined", "with", "some", "minor", "revisions", "to", "the", "employment", "data", "." ]
finerord735
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Monthly job growth will need to accelerate sharply in the back half of the year if there is to be any hope of adding as many jobs in 2015 as we did in 2014 . Answer:
Monthly:O job:O growth:O will:O need:O to:O accelerate:O sharply:O in:O the:O back:O half:O of:O the:O year:O if:O there:O is:O to:O be:O any:O hope:O of:O adding:O as:O many:O jobs:O in:O 2015:O as:O we:O did:O in:O 2014:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Monthly", "job", "growth", "will", "need", "to", "accelerate", "sharply", "in", "the", "back", "half", "of", "the", "year", "if", "there", "is", "to", "be", "any", "hope", "of", "adding", "as", "many", "jobs", "in", "2015", "as", "we", "did", "in", "2014", "." ]
finerord736
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Job growth would need to accelerate to something like 280,000 jobs per month , on average to get to the 2.9 million or so jobs added in 2014 . Answer:
Job:O growth:O would:O need:O to:O accelerate:O to:O something:O like:O 280,000:O jobs:O per:O month:O ,:O on:O average:O to:O get:O to:O the:O 2.9:O million:O or:O so:O jobs:O added:O in:O 2014:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Job", "growth", "would", "need", "to", "accelerate", "to", "something", "like", "280,000", "jobs", "per", "month", ",", "on", "average", "to", "get", "to", "the", "2.9", "million", "or", "so", "jobs", "added", "in", "2014", "." ]
finerord737
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: At the moment , that looks like a bit of a stretch . Answer:
At:O the:O moment:O ,:O that:O looks:O like:O a:O bit:O of:O a:O stretch:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "At", "the", "moment", ",", "that", "looks", "like", "a", "bit", "of", "a", "stretch", "." ]
finerord738
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: In other news , purchasing manager reports continued to support our thesis that the manufacturing sector is near a bottom . Answer:
In:O other:O news:O ,:O purchasing:O manager:O reports:O continued:O to:O support:O our:O thesis:O that:O the:O manufacturing:O sector:O is:O near:O a:O bottom:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "In", "other", "news", ",", "purchasing", "manager", "reports", "continued", "to", "support", "our", "thesis", "that", "the", "manufacturing", "sector", "is", "near", "a", "bottom", "." ]
finerord739
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Additionally , the existing - home industry continued to pick up steam as indicated by yet another strong pending home sales report . Answer:
Additionally:O ,:O the:O existing:O -:O home:O industry:O continued:O to:O pick:O up:O steam:O as:O indicated:O by:O yet:O another:O strong:O pending:O home:O sales:O report:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Additionally", ",", "the", "existing", "-", "home", "industry", "continued", "to", "pick", "up", "steam", "as", "indicated", "by", "yet", "another", "strong", "pending", "home", "sales", "report", "." ]
finerord740
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Job Growth Rates Return to Earth Headline job growth of 223,000 total jobs added was near consensus and our forecast but was viewed as a disappointment to many who had hoped for a much better number . Answer:
Job:O Growth:O Rates:O Return:O to:O Earth:O Headline:O job:O growth:O of:O 223,000:O total:O jobs:O added:O was:O near:O consensus:O and:O our:O forecast:O but:O was:O viewed:O as:O a:O disappointment:O to:O many:O who:O had:O hoped:O for:O a:O much:O better:O number:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Job", "Growth", "Rates", "Return", "to", "Earth", "Headline", "job", "growth", "of", "223,000", "total", "jobs", "added", "was", "near", "consensus", "and", "our", "forecast", "but", "was", "viewed", "as", "a", "disappointment", "to", "many", "who", "had", "hoped", "for", "a", "much", "better", "number", "." ]
finerord741
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The bulls had hung their hats on strong job openings , low initial unemployment claims , and the employment metrics of various manufacturing reports . Answer:
The:O bulls:O had:O hung:O their:O hats:O on:O strong:O job:O openings:O ,:O low:O initial:O unemployment:O claims:O ,:O and:O the:O employment:O metrics:O of:O various:O manufacturing:O reports:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "bulls", "had", "hung", "their", "hats", "on", "strong", "job", "openings", ",", "low", "initial", "unemployment", "claims", ",", "and", "the", "employment", "metrics", "of", "various", "manufacturing", "reports", "." ]
finerord742
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The bears cited a slowing and low rate of GDP growth compared with unusually strong employment growth , especially at the end of 2014 , as a reason for their pessimism . Answer:
The:O bears:O cited:O a:O slowing:O and:O low:O rate:O of:O GDP:O growth:O compared:O with:O unusually:O strong:O employment:O growth:O ,:O especially:O at:O the:O end:O of:O 2014:O ,:O as:O a:O reason:O for:O their:O pessimism:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "bears", "cited", "a", "slowing", "and", "low", "rate", "of", "GDP", "growth", "compared", "with", "unusually", "strong", "employment", "growth", ",", "especially", "at", "the", "end", "of", "2014", ",", "as", "a", "reason", "for", "their", "pessimism", "." ]
finerord743
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: As it turns out , the final number came in between the two extremes . Answer:
As:O it:O turns:O out:O ,:O the:O final:O number:O came:O in:O between:O the:O two:O extremes:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "As", "it", "turns", "out", ",", "the", "final", "number", "came", "in", "between", "the", "two", "extremes", "." ]
finerord744
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: At 223,000 the June number came in below the 12 - month average of 244,000 jobs added but above the six - month average of 208,000 jobs added . Answer:
At:O 223,000:O the:O June:O number:O came:O in:O below:O the:O 12:O -:O month:O average:O of:O 244,000:O jobs:O added:O but:O above:O the:O six:O -:O month:O average:O of:O 208,000:O jobs:O added:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "At", "223,000", "the", "June", "number", "came", "in", "below", "the", "12", "-", "month", "average", "of", "244,000", "jobs", "added", "but", "above", "the", "six", "-", "month", "average", "of", "208,000", "jobs", "added", "." ]
finerord745
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The 429,000 jobs added last November and the 329,000 jobs added in December continue to stick out like a pair of sore thumbs . Answer:
The:O 429,000:O jobs:O added:O last:O November:O and:O the:O 329,000:O jobs:O added:O in:O December:O continue:O to:O stick:O out:O like:O a:O pair:O of:O sore:O thumbs:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "429,000", "jobs", "added", "last", "November", "and", "the", "329,000", "jobs", "added", "in", "December", "continue", "to", "stick", "out", "like", "a", "pair", "of", "sore", "thumbs", "." ]
finerord746
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Those really high numbers severely complicate analysis of a lot of the data . Answer:
Those:O really:O high:O numbers:O severely:O complicate:O analysis:O of:O a:O lot:O of:O the:O data:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Those", "really", "high", "numbers", "severely", "complicate", "analysis", "of", "a", "lot", "of", "the", "data", "." ]
finerord747
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: It made the end of 2014 look boom - like and has been followed by months of payback . Answer:
It:O made:O the:O end:O of:O 2014:O look:O boom:O -:O like:O and:O has:O been:O followed:O by:O months:O of:O payback:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "It", "made", "the", "end", "of", "2014", "look", "boom", "-", "like", "and", "has", "been", "followed", "by", "months", "of", "payback", "." ]
finerord748
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Also , it is not entirely clear if the great numbers for those two months were due to faulty seasonal adjustment factors or if there just happened to be a perfect storm of good news events during those two months . Answer:
Also:O ,:O it:O is:O not:O entirely:O clear:O if:O the:O great:O numbers:O for:O those:O two:O months:O were:O due:O to:O faulty:O seasonal:O adjustment:O factors:O or:O if:O there:O just:O happened:O to:O be:O a:O perfect:O storm:O of:O good:O news:O events:O during:O those:O two:O months:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Also", ",", "it", "is", "not", "entirely", "clear", "if", "the", "great", "numbers", "for", "those", "two", "months", "were", "due", "to", "faulty", "seasonal", "adjustment", "factors", "or", "if", "there", "just", "happened", "to", "be", "a", "perfect", "storm", "of", "good", "news", "events", "during", "those", "two", "months", "." ]
finerord749
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The good news is that with this report GDP growth and employment growth have resumed their more normal relationship . Answer:
The:O good:O news:O is:O that:O with:O this:O report:O GDP:O growth:O and:O employment:O growth:O have:O resumed:O their:O more:O normal:O relationship:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "good", "news", "is", "that", "with", "this", "report", "GDP", "growth", "and", "employment", "growth", "have", "resumed", "their", "more", "normal", "relationship", "." ]
finerord750
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We had suspected that either a couple of really horrific months or a restatement would have brought these two points back together . Answer:
We:O had:O suspected:O that:O either:O a:O couple:O of:O really:O horrific:O months:O or:O a:O restatement:O would:O have:O brought:O these:O two:O points:O back:O together:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "We", "had", "suspected", "that", "either", "a", "couple", "of", "really", "horrific", "months", "or", "a", "restatement", "would", "have", "brought", "these", "two", "points", "back", "together", "." ]
finerord751
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Instead , we have witnessed a slow , almost indiscernible , inching downward trend with good months and bad months . Answer:
Instead:O ,:O we:O have:O witnessed:O a:O slow:O ,:O almost:O indiscernible:O ,:O inching:O downward:O trend:O with:O good:O months:O and:O bad:O months:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Instead", ",", "we", "have", "witnessed", "a", "slow", ",", "almost", "indiscernible", ",", "inching", "downward", "trend", "with", "good", "months", "and", "bad", "months", "." ]
finerord752
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: As of the second quarter , year - over - year nonfarm payroll growth is averaging around 2.1 % and GDP growth over the same period is running around 2.5 % . Answer:
As:O of:O the:O second:O quarter:O ,:O year:O -:O over:O -:O year:O nonfarm:O payroll:O growth:O is:O averaging:O around:O 2.1:O %:O and:O GDP:O growth:O over:O the:O same:O period:O is:O running:O around:O 2.5:O %:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "As", "of", "the", "second", "quarter", ",", "year", "-", "over", "-", "year", "nonfarm", "payroll", "growth", "is", "averaging", "around", "2.1", "%", "and", "GDP", "growth", "over", "the", "same", "period", "is", "running", "around", "2.5", "%", "." ]
finerord753
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Not Much Hourly Wage Growth or Hours Worked Growth Month to Month Answer:
Not:O Much:O Hourly:O Wage:O Growth:O or:O Hours:O Worked:O Growth:O Month:O to:O Month:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Not", "Much", "Hourly", "Wage", "Growth", "or", "Hours", "Worked", "Growth", "Month", "to", "Month" ]
finerord754
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: With no hourly wage growth month to month and a modest slowing in the single - month year - over - year wage growth rate , a lot of economists were disappointed with the June data . Answer:
With:O no:O hourly:O wage:O growth:O month:O to:O month:O and:O a:O modest:O slowing:O in:O the:O single:O -:O month:O year:O -:O over:O -:O year:O wage:O growth:O rate:O ,:O a:O lot:O of:O economists:O were:O disappointed:O with:O the:O June:O data:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "With", "no", "hourly", "wage", "growth", "month", "to", "month", "and", "a", "modest", "slowing", "in", "the", "single", "-", "month", "year", "-", "over", "-", "year", "wage", "growth", "rate", ",", "a", "lot", "of", "economists", "were", "disappointed", "with", "the", "June", "data", "." ]
finerord755
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We think the average measure has grown a bit confusing as it muddles together many industries and is strongly affected by mix issues and undue influence of some industries . Answer:
We:O think:O the:O average:O measure:O has:O grown:O a:O bit:O confusing:O as:O it:O muddles:O together:O many:O industries:O and:O is:O strongly:O affected:O by:O mix:O issues:O and:O undue:O influence:O of:O some:O industries:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "We", "think", "the", "average", "measure", "has", "grown", "a", "bit", "confusing", "as", "it", "muddles", "together", "many", "industries", "and", "is", "strongly", "affected", "by", "mix", "issues", "and", "undue", "influence", "of", "some", "industries", "." ]
finerord756
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The number is also prone to lumpiness with good months often followed by a series of months with no change . Answer:
The:O number:O is:O also:O prone:O to:O lumpiness:O with:O good:O months:O often:O followed:O by:O a:O series:O of:O months:O with:O no:O change:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "number", "is", "also", "prone", "to", "lumpiness", "with", "good", "months", "often", "followed", "by", "a", "series", "of", "months", "with", "no", "change", "." ]
finerord757
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Being economists , we would all love to see five months of 0.2 % monthly growth than one month of 1 % growth and four months of no growth . Answer:
Being:O economists:O ,:O we:O would:O all:O love:O to:O see:O five:O months:O of:O 0.2:O %:O monthly:O growth:O than:O one:O month:O of:O 1:O %:O growth:O and:O four:O months:O of:O no:O growth:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Being", "economists", ",", "we", "would", "all", "love", "to", "see", "five", "months", "of", "0.2", "%", "monthly", "growth", "than", "one", "month", "of", "1", "%", "growth", "and", "four", "months", "of", "no", "growth", "." ]
finerord758
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Both ways get us to the same end , but it creates a lot of panic when we hit the slow spots . Answer:
Both:O ways:O get:O us:O to:O the:O same:O end:O ,:O but:O it:O creates:O a:O lot:O of:O panic:O when:O we:O hit:O the:O slow:O spots:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Both", "ways", "get", "us", "to", "the", "same", "end", ",", "but", "it", "creates", "a", "lot", "of", "panic", "when", "we", "hit", "the", "slow", "spots", "." ]
finerord759
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Also unusual in the year - over - year data are a lot of industries that are doing quite well and some that are not . Answer:
Also:O unusual:O in:O the:O year:O -:O over:O -:O year:O data:O are:O a:O lot:O of:O industries:O that:O are:O doing:O quite:O well:O and:O some:O that:O are:O not:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Also", "unusual", "in", "the", "year", "-", "over", "-", "year", "data", "are", "a", "lot", "of", "industries", "that", "are", "doing", "quite", "well", "and", "some", "that", "are", "not", "." ]
finerord760
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Surprisingly , there appears to be some convergence going on here , with some of the higher - paying sectors -- manufacturing in particular -- and mining ( includes oil extraction ) doing poorly and low - paying sectors such as leisure and hospitality doing better . Answer:
Surprisingly:O ,:O there:O appears:O to:O be:O some:O convergence:O going:O on:O here:O ,:O with:O some:O of:O the:O higher:O -:O paying:O sectors:O --:O manufacturing:O in:O particular:O --:O and:O mining:O (:O includes:O oil:O extraction:O ):O doing:O poorly:O and:O low:O -:O paying:O sectors:O such:O as:O leisure:O and:O hospitality:O doing:O better:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Surprisingly", ",", "there", "appears", "to", "be", "some", "convergence", "going", "on", "here", ",", "with", "some", "of", "the", "higher", "-", "paying", "sectors", "--", "manufacturing", "in", "particular", "--", "and", "mining", "(", "includes", "oil", "extraction", ")", "doing", "poorly", "and", "low", "-", "paying", "sectors", "such", "as", "leisure", "and", "hospitality", "doing", "better", "." ]
finerord761
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Overall Total Year - Over - Year Wage Growth Is Slowing Slightly Rolling together employment growth , hours worked growth , and the growth in the hourly wage rate , things still look pretty good but not stunning . Answer:
Overall:O Total:O Year:O -:O Over:O -:O Year:O Wage:O Growth:O Is:O Slowing:O Slightly:O Rolling:O together:O employment:O growth:O ,:O hours:O worked:O growth:O ,:O and:O the:O growth:O in:O the:O hourly:O wage:O rate:O ,:O things:O still:O look:O pretty:O good:O but:O not:O stunning:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Overall", "Total", "Year", "-", "Over", "-", "Year", "Wage", "Growth", "Is", "Slowing", "Slightly", "Rolling", "together", "employment", "growth", ",", "hours", "worked", "growth", ",", "and", "the", "growth", "in", "the", "hourly", "wage", "rate", ",", "things", "still", "look", "pretty", "good", "but", "not", "stunning", "." ]
finerord762
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Private sector employment growth is at its 12 - month average and the hourly wage growth rate is just one tick above the average . Answer:
Private:O sector:O employment:O growth:O is:O at:O its:O 12:O -:O month:O average:O and:O the:O hourly:O wage:O growth:O rate:O is:O just:O one:O tick:O above:O the:O average:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Private", "sector", "employment", "growth", "is", "at", "its", "12", "-", "month", "average", "and", "the", "hourly", "wage", "growth", "rate", "is", "just", "one", "tick", "above", "the", "average", "." ]
finerord763
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Meanwhile , hours worked growth has seemingly come to an end , as it often does at this part of a recovery . Answer:
Meanwhile:O ,:O hours:O worked:O growth:O has:O seemingly:O come:O to:O an:O end:O ,:O as:O it:O often:O does:O at:O this:O part:O of:O a:O recovery:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Meanwhile", ",", "hours", "worked", "growth", "has", "seemingly", "come", "to", "an", "end", ",", "as", "it", "often", "does", "at", "this", "part", "of", "a", "recovery", "." ]
finerord764
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Employers are opting for more employees and not working their current ones more hours . Answer:
Employers:O are:O opting:O for:O more:O employees:O and:O not:O working:O their:O current:O ones:O more:O hours:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Employers", "are", "opting", "for", "more", "employees", "and", "not", "working", "their", "current", "ones", "more", "hours", "." ]
finerord765
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Putting it all altogether , total wage growth has been slowing since February and is down to 4.7 % from a high of 5.4 % and its 12 - month average of 4.9 % . Answer:
Putting:O it:O all:O altogether:O ,:O total:O wage:O growth:O has:O been:O slowing:O since:O February:O and:O is:O down:O to:O 4.7:O %:O from:O a:O high:O of:O 5.4:O %:O and:O its:O 12:O -:O month:O average:O of:O 4.9:O %:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Putting", "it", "all", "altogether", ",", "total", "wage", "growth", "has", "been", "slowing", "since", "February", "and", "is", "down", "to", "4.7", "%", "from", "a", "high", "of", "5.4", "%", "and", "its", "12", "-", "month", "average", "of", "4.9", "%", "." ]
finerord766
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Inflation - Adjusted Data Trending Lower , Collapse Answer:
Inflation:O -:O Adjusted:O Data:O Trending:O Lower:O ,:O Collapse:O
[ "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Inflation", "-", "Adjusted", "Data", "Trending", "Lower", ",", "Collapse" ]
finerord767
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Is on the Way Low year - over - year inflation rates , almost entirely due to low gasoline prices , have meant those dollars theoretically grow much faster . Answer:
Is:O on:O the:O Way:O Low:O year:O -:O over:O -:O year:O inflation:O rates:O ,:O almost:O entirely:O due:O to:O low:O gasoline:O prices:O ,:O have:O meant:O those:O dollars:O theoretically:O grow:O much:O faster:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Is", "on", "the", "Way", "Low", "year", "-", "over", "-", "year", "inflation", "rates", ",", "almost", "entirely", "due", "to", "low", "gasoline", "prices", ",", "have", "meant", "those", "dollars", "theoretically", "grow", "much", "faster", "." ]
finerord768
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: And the table below shows just that . Answer:
And:O the:O table:O below:O shows:O just:O that:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "And", "the", "table", "below", "shows", "just", "that", "." ]
finerord769
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Consumers have spent some of the extra inflation - related cash , but certainly not all of it , as savings rates have been increasing . Answer:
Consumers:O have:O spent:O some:O of:O the:O extra:O inflation:O -:O related:O cash:O ,:O but:O certainly:O not:O all:O of:O it:O ,:O as:O savings:O rates:O have:O been:O increasing:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Consumers", "have", "spent", "some", "of", "the", "extra", "inflation", "-", "related", "cash", ",", "but", "certainly", "not", "all", "of", "it", ",", "as", "savings", "rates", "have", "been", "increasing", "." ]
finerord770
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: That is probably a really good thing as employers have been doling out 2 % raises on average , no matter what the gyrations are in the inflation rate . Answer:
That:O is:O probably:O a:O really:O good:O thing:O as:O employers:O have:O been:O doling:O out:O 2:O %:O raises:O on:O average:O ,:O no:O matter:O what:O the:O gyrations:O are:O in:O the:O inflation:O rate:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "That", "is", "probably", "a", "really", "good", "thing", "as", "employers", "have", "been", "doling", "out", "2", "%", "raises", "on", "average", ",", "no", "matter", "what", "the", "gyrations", "are", "in", "the", "inflation", "rate", "." ]
finerord771
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Both employers and employees have been acting rationally , responding only to what they believe are changes in long - term inflation rates and not temporary blips in commodity prices . Answer:
Both:O employers:O and:O employees:O have:O been:O acting:O rationally:O ,:O responding:O only:O to:O what:O they:O believe:O are:O changes:O in:O long:O -:O term:O inflation:O rates:O and:O not:O temporary:O blips:O in:O commodity:O prices:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Both", "employers", "and", "employees", "have", "been", "acting", "rationally", ",", "responding", "only", "to", "what", "they", "believe", "are", "changes", "in", "long", "-", "term", "inflation", "rates", "and", "not", "temporary", "blips", "in", "commodity", "prices", "." ]
finerord772
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: While total wages are up around 5 % year over year during the past six months , consumption growth has been stuck at around 3 % . Answer:
While:O total:O wages:O are:O up:O around:O 5:O %:O year:O over:O year:O during:O the:O past:O six:O months:O ,:O consumption:O growth:O has:O been:O stuck:O at:O around:O 3:O %:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "While", "total", "wages", "are", "up", "around", "5", "%", "year", "over", "year", "during", "the", "past", "six", "months", ",", "consumption", "growth", "has", "been", "stuck", "at", "around", "3", "%", "." ]
finerord773
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: That 's a very good thing because without a total collapse in gasoline prices in the near term , year - over - year inflation could begin to approach 2 % by December . Answer:
That:O 's:O a:O very:O good:O thing:O because:O without:O a:O total:O collapse:O in:O gasoline:O prices:O in:O the:O near:O term:O ,:O year:O -:O over:O -:O year:O inflation:O could:O begin:O to:O approach:O 2:O %:O by:O December:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "That", "'s", "a", "very", "good", "thing", "because", "without", "a", "total", "collapse", "in", "gasoline", "prices", "in", "the", "near", "term", ",", "year", "-", "over", "-", "year", "inflation", "could", "begin", "to", "approach", "2", "%", "by", "December", "." ]
finerord774
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: If hourly wage growth does n't accelerate ( which could be given our labor shortage thesis ) , there wo n't be any wage income growth per individual and total wages will only increase to the extent of the percentage growth of the number of people employed . Answer:
If:O hourly:O wage:O growth:O does:O n't:O accelerate:O (:O which:O could:O be:O given:O our:O labor:O shortage:O thesis:O ):O ,:O there:O wo:O n't:O be:O any:O wage:O income:O growth:O per:O individual:O and:O total:O wages:O will:O only:O increase:O to:O the:O extent:O of:O the:O percentage:O growth:O of:O the:O number:O of:O people:O employed:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "If", "hourly", "wage", "growth", "does", "n't", "accelerate", "(", "which", "could", "be", "given", "our", "labor", "shortage", "thesis", ")", ",", "there", "wo", "n't", "be", "any", "wage", "income", "growth", "per", "individual", "and", "total", "wages", "will", "only", "increase", "to", "the", "extent", "of", "the", "percentage", "growth", "of", "the", "number", "of", "people", "employed", "." ]
finerord775
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Because consumers have been cautious about spending their income gains , the impact on the economy might not be large . Answer:
Because:O consumers:O have:O been:O cautious:O about:O spending:O their:O income:O gains:O ,:O the:O impact:O on:O the:O economy:O might:O not:O be:O large:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Because", "consumers", "have", "been", "cautious", "about", "spending", "their", "income", "gains", ",", "the", "impact", "on", "the", "economy", "might", "not", "be", "large", "." ]
finerord776
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Still , those forecasting a sharply accelerating economy and reaching some type of escape velocity are likely to be disappointed yet again . Answer:
Still:O ,:O those:O forecasting:O a:O sharply:O accelerating:O economy:O and:O reaching:O some:O type:O of:O escape:O velocity:O are:O likely:O to:O be:O disappointed:O yet:O again:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Still", ",", "those", "forecasting", "a", "sharply", "accelerating", "economy", "and", "reaching", "some", "type", "of", "escape", "velocity", "are", "likely", "to", "be", "disappointed", "yet", "again", "." ]
finerord777
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Employment Sector Data Shows Answer:
Employment:O Sector:O Data:O Shows:O
[ "O", "O", "O", "O" ]
[ "Employment", "Sector", "Data", "Shows" ]
finerord778
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: It May Be Hard to Accelerate Job Growth Answer:
It:O May:O Be:O Hard:O to:O Accelerate:O Job:O Growth:O
[ "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "It", "May", "Be", "Hard", "to", "Accelerate", "Job", "Growth" ]
finerord779
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Unfortunately , the number of job sectors that are doing really well are quite limited and some of the poorer - performing sectors do n't have a lot of hope for doing much better . Answer:
Unfortunately:O ,:O the:O number:O of:O job:O sectors:O that:O are:O doing:O really:O well:O are:O quite:O limited:O and:O some:O of:O the:O poorer:O -:O performing:O sectors:O do:O n't:O have:O a:O lot:O of:O hope:O for:O doing:O much:O better:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Unfortunately", ",", "the", "number", "of", "job", "sectors", "that", "are", "doing", "really", "well", "are", "quite", "limited", "and", "some", "of", "the", "poorer", "-", "performing", "sectors", "do", "n't", "have", "a", "lot", "of", "hope", "for", "doing", "much", "better", "." ]
finerord780
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Too , some of the better performing sectors have slowed up , especially those related to manufacturing . Answer:
Too:O ,:O some:O of:O the:O better:O performing:O sectors:O have:O slowed:O up:O ,:O especially:O those:O related:O to:O manufacturing:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Too", ",", "some", "of", "the", "better", "performing", "sectors", "have", "slowed", "up", ",", "especially", "those", "related", "to", "manufacturing", "." ]
finerord781
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The table below shows particularly good and bad performers over the past 12 months and their job additions in June alone . Answer:
The:O table:O below:O shows:O particularly:O good:O and:O bad:O performers:O over:O the:O past:O 12:O months:O and:O their:O job:O additions:O in:O June:O alone:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "table", "below", "shows", "particularly", "good", "and", "bad", "performers", "over", "the", "past", "12", "months", "and", "their", "job", "additions", "in", "June", "alone", "." ]
finerord782
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Construction , a great performer year over year , had a rough June and we are hoping for some help from the sector , especially the residential market , by the end of the year . Answer:
Construction:O ,:O a:O great:O performer:O year:O over:O year:O ,:O had:O a:O rough:O June:O and:O we:O are:O hoping:O for:O some:O help:O from:O the:O sector:O ,:O especially:O the:O residential:O market:O ,:O by:O the:O end:O of:O the:O year:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Construction", ",", "a", "great", "performer", "year", "over", "year", ",", "had", "a", "rough", "June", "and", "we", "are", "hoping", "for", "some", "help", "from", "the", "sector", ",", "especially", "the", "residential", "market", ",", "by", "the", "end", "of", "the", "year", "." ]
finerord783
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Business and professional , a high - paying , long - hours sector in general , did great on both a year - over - year and a month - to - month basis . Answer:
Business:O and:O professional:O ,:O a:O high:O -:O paying:O ,:O long:O -:O hours:O sector:O in:O general:O ,:O did:O great:O on:O both:O a:O year:O -:O over:O -:O year:O and:O a:O month:O -:O to:O -:O month:O basis:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Business", "and", "professional", ",", "a", "high", "-", "paying", ",", "long", "-", "hours", "sector", "in", "general", ",", "did", "great", "on", "both", "a", "year", "-", "over", "-", "year", "and", "a", "month", "-", "to", "-", "month", "basis", "." ]
finerord784
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The same can be said for health care , which has been gaining some momentum . Answer:
The:O same:O can:O be:O said:O for:O health:O care:O ,:O which:O has:O been:O gaining:O some:O momentum:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "same", "can", "be", "said", "for", "health", "care", ",", "which", "has", "been", "gaining", "some", "momentum", "." ]
finerord785
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The leisure category , led by restaurants , has had a good year but has been losing some momentum . Answer:
The:O leisure:O category:O ,:O led:O by:O restaurants:O ,:O has:O had:O a:O good:O year:O but:O has:O been:O losing:O some:O momentum:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "leisure", "category", ",", "led", "by", "restaurants", ",", "has", "had", "a", "good", "year", "but", "has", "been", "losing", "some", "momentum", "." ]
finerord786
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: MCM FX Joins the GWAZY League as a Competition Sponsor MCM FX Joins the GWAZY League as a Competition Sponsor Answer:
MCM:B-ORG FX:I-ORG Joins:O the:O GWAZY:B-ORG League:I-ORG as:O a:O Competition:O Sponsor:O MCM:B-ORG FX:I-ORG Joins:O the:O GWAZY:B-ORG League:I-ORG as:O a:O Competition:O Sponsor:O
[ "B-ORG", "I-ORG", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O" ]
[ "MCM", "FX", "Joins", "the", "GWAZY", "League", "as", "a", "Competition", "Sponsor", "MCM", "FX", "Joins", "the", "GWAZY", "League", "as", "a", "Competition", "Sponsor" ]
finerord787
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The offering was launched earlier this year , as a user - friendly interface for its brands to have a unique trading competition experience . Answer:
The:O offering:O was:O launched:O earlier:O this:O year:O ,:O as:O a:O user:O -:O friendly:O interface:O for:O its:O brands:O to:O have:O a:O unique:O trading:O competition:O experience:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "offering", "was", "launched", "earlier", "this", "year", ",", "as", "a", "user", "-", "friendly", "interface", "for", "its", "brands", "to", "have", "a", "unique", "trading", "competition", "experience", "." ]
finerord788
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Advertising Print This Post The UK - based broker MCM FX has recently started offering binary options trading based on the technology of the provider GWAZY . Answer:
Advertising:O Print:O This:O Post:O The:O UK:B-LOC -:O based:O broker:O MCM:B-ORG FX:I-ORG has:O recently:O started:O offering:O binary:O options:O trading:O based:O on:O the:O technology:O of:O the:O provider:O GWAZY:B-ORG .:O
[ "O", "O", "O", "O", "O", "B-LOC", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "O" ]
[ "Advertising", "Print", "This", "Post", "The", "UK", "-", "based", "broker", "MCM", "FX", "has", "recently", "started", "offering", "binary", "options", "trading", "based", "on", "the", "technology", "of", "the", "provider", "GWAZY", "." ]
finerord789
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: As part of its promotional activities , MCM FX is now also sponsoring a competition in the GWAZY League . Answer:
As:O part:O of:O its:O promotional:O activities:O ,:O MCM:B-ORG FX:I-ORG is:O now:O also:O sponsoring:O a:O competition:O in:O the:O GWAZY:B-ORG League:I-ORG .:O
[ "O", "O", "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O" ]
[ "As", "part", "of", "its", "promotional", "activities", ",", "MCM", "FX", "is", "now", "also", "sponsoring", "a", "competition", "in", "the", "GWAZY", "League", "." ]
finerord790
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The offering was launched earlier this year , as a user - friendly interface for a unique trading competition experience . Answer:
The:O offering:O was:O launched:O earlier:O this:O year:O ,:O as:O a:O user:O -:O friendly:O interface:O for:O a:O unique:O trading:O competition:O experience:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "offering", "was", "launched", "earlier", "this", "year", ",", "as", "a", "user", "-", "friendly", "interface", "for", "a", "unique", "trading", "competition", "experience", "." ]
finerord791
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: Anyone over the age of eighteen can join the GWAZY League and participate in free trading competitions and challenges where they can win gadgets , prizes or live accounts , depending on what is offered by the sponsors . Answer:
Anyone:O over:O the:O age:O of:O eighteen:O can:O join:O the:O GWAZY:B-ORG League:I-ORG and:O participate:O in:O free:O trading:O competitions:O and:O challenges:O where:O they:O can:O win:O gadgets:O ,:O prizes:O or:O live:O accounts:O ,:O depending:O on:O what:O is:O offered:O by:O the:O sponsors:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "Anyone", "over", "the", "age", "of", "eighteen", "can", "join", "the", "GWAZY", "League", "and", "participate", "in", "free", "trading", "competitions", "and", "challenges", "where", "they", "can", "win", "gadgets", ",", "prizes", "or", "live", "accounts", ",", "depending", "on", "what", "is", "offered", "by", "the", "sponsors", "." ]
finerord792
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: All brokers who currently sponsor competitions and promotions also offer live trading . Answer:
All:O brokers:O who:O currently:O sponsor:O competitions:O and:O promotions:O also:O offer:O live:O trading:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "All", "brokers", "who", "currently", "sponsor", "competitions", "and", "promotions", "also", "offer", "live", "trading", "." ]
finerord793
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: β€œ The exciting gaming environment offered by the GWAZY platform opens new doors to brokers , giving them the opportunity to capture different market segments as most trading platforms that are currently available in the market reflect more β€˜ traditional ’ trading methods . Answer:
β€œ:O The:O exciting:O gaming:O environment:O offered:O by:O the:O GWAZY:B-ORG platform:O opens:O new:O doors:O to:O brokers:O ,:O giving:O them:O the:O opportunity:O to:O capture:O different:O market:O segments:O as:O most:O trading:O platforms:O that:O are:O currently:O available:O in:O the:O market:O reflect:O more:O β€˜:O traditional:O ’:O trading:O methods:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "B-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "β€œ", "The", "exciting", "gaming", "environment", "offered", "by", "the", "GWAZY", "platform", "opens", "new", "doors", "to", "brokers", ",", "giving", "them", "the", "opportunity", "to", "capture", "different", "market", "segments", "as", "most", "trading", "platforms", "that", "are", "currently", "available", "in", "the", "market", "reflect", "more", "β€˜", "traditional", "’", "trading", "methods", "." ]
finerord794
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: We would like to welcome MCM FX on board and look forward to more FX and Binary brokers joining us ” said Martin , Head of Development at GWAZY Ltd . Answer:
We:O would:O like:O to:O welcome:O MCM:B-ORG FX:I-ORG on:O board:O and:O look:O forward:O to:O more:O FX:O and:O Binary:O brokers:O joining:O us:O ”:O said:O Martin:B-PER ,:O Head:O of:O Development:O at:O GWAZY:B-ORG Ltd:I-ORG .:O
[ "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-PER", "O", "O", "O", "O", "O", "B-ORG", "I-ORG", "O" ]
[ "We", "would", "like", "to", "welcome", "MCM", "FX", "on", "board", "and", "look", "forward", "to", "more", "FX", "and", "Binary", "brokers", "joining", "us", "”", "said", "Martin", ",", "Head", "of", "Development", "at", "GWAZY", "Ltd", "." ]
finerord795
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The GWAZY platform was launched as an alternative method to FX trading , designed for beginners and people who are interested in trading but in a stress free environment . Answer:
The:O GWAZY:B-ORG platform:O was:O launched:O as:O an:O alternative:O method:O to:O FX:O trading:O ,:O designed:O for:O beginners:O and:O people:O who:O are:O interested:O in:O trading:O but:O in:O a:O stress:O free:O environment:O .:O
[ "O", "B-ORG", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "GWAZY", "platform", "was", "launched", "as", "an", "alternative", "method", "to", "FX", "trading", ",", "designed", "for", "beginners", "and", "people", "who", "are", "interested", "in", "trading", "but", "in", "a", "stress", "free", "environment", "." ]
finerord796
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The solution offers FX and binary options brokers , the chance to launch within two weeks and zero setup costs . Answer:
The:O solution:O offers:O FX:O and:O binary:O options:O brokers:O ,:O the:O chance:O to:O launch:O within:O two:O weeks:O and:O zero:O setup:O costs:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "solution", "offers", "FX", "and", "binary", "options", "brokers", ",", "the", "chance", "to", "launch", "within", "two", "weeks", "and", "zero", "setup", "costs", "." ]
finerord797
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The package includes a platform , manager , tools , media kit and reporting system . Answer:
The:O package:O includes:O a:O platform:O ,:O manager:O ,:O tools:O ,:O media:O kit:O and:O reporting:O system:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "package", "includes", "a", "platform", ",", "manager", ",", "tools", ",", "media", "kit", "and", "reporting", "system", "." ]
finerord798
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: By sponsoring promotions and competitions , brokers can increase the popularity of their brand and eventually boost their business performance . Answer:
By:O sponsoring:O promotions:O and:O competitions:O ,:O brokers:O can:O increase:O the:O popularity:O of:O their:O brand:O and:O eventually:O boost:O their:O business:O performance:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "By", "sponsoring", "promotions", "and", "competitions", ",", "brokers", "can", "increase", "the", "popularity", "of", "their", "brand", "and", "eventually", "boost", "their", "business", "performance", "." ]
finerord799
In the list of tokens, identify 'Person', 'Location', and 'Organisation' and label each accordingly. If the entity spans multiple tokens, use the prefix B-PER, B-LOC, or B-ORG for the first token, and I-PER, I-LOC, or I-ORG for the subsequent tokens of that entity. The beginning of each separate entity should always be labeled with a B-PER, B-LOC, or B-ORG prefix. If the token does not fit into any of the three named categories, or is not a named entity, label it as 'O'. Each line should contain one token and its corresponding label, separated by a colon. Do not combine tokens on your own. The format for each line should be: 'token:label'. Text: The platform is available in English , Arabic , Farsi and Chinese . Answer:
The:O platform:O is:O available:O in:O English:O ,:O Arabic:O ,:O Farsi:O and:O Chinese:O .:O
[ "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O" ]
[ "The", "platform", "is", "available", "in", "English", ",", "Arabic", ",", "Farsi", "and", "Chinese", "." ]