File size: 31,602 Bytes
6fa4bc9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 |
{
"paper_id": "M91-1034",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T03:15:32.437444Z"
},
"title": "DESCRIPTION OF THE UNL/USL SYSTEM USED FOR MUC-3",
"authors": [
{
"first": "Jitender",
"middle": [
"S"
],
"last": "Deogun",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "Engineering University of. Nebraska -Lincoln Lincoln",
"location": {
"postCode": "68588-0115",
"region": "NE"
}
},
"email": "deogun@fergvax.unl.edu"
},
{
"first": "Vijay",
"middle": [
"V"
],
"last": "Raghavan",
"suffix": "",
"affiliation": {
"laboratory": "",
"institution": "University of Southwestern Louisian a Lafayette",
"location": {
"postCode": "70504-4330",
"region": "LA"
}
},
"email": "rghavan@cacs.usl.edu"
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "",
"pdf_parse": {
"paper_id": "M91-1034",
"_pdf_hash": "",
"abstract": [],
"body_text": [
{
"text": "The MuC-3 task consists of generating a database of filled templates with respect to messages tha t belong to a general topical domain . In particular, for the current phase, the message collectio n belongs to the domain of terrorist activities. On the one hand, a decision as to the relevance o f a message to a specified class of terrorist events should be made . If relevant, a predefined set of facts are to be extracted and placed as fills for appropriate slots of the template(s) created for this message . If not relevant, a template having a'*' as the fill in all but one slot, is created (see Appendix A for details) . Some aspects of the MUC-3 task are amenable to be solved by techniques typically employed in information retrieval (IR) . These techniques are especially designed to be applicable t o any domain . In contrast, there are other aspects of the problem that may require a great deal of language understanding, thus needing natural language processing (NLP) techniques . For the most part, NLP techniques may be considered domain dependent .",
"cite_spans": [
{
"start": 975,
"end": 980,
"text": "(NLP)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "BACKGROUN D",
"sec_num": null
},
{
"text": "The primary thrust of our effort has been to design and implement a system that employ s techniques typically found in m literature, augmented by basic search techniques available in fil e management systems . An important goal (for the time being) is to ensure that the system is domai n independent to the greatest extent possible . Consequently, certain slots which are not suitable,t o be handled by the chosen techniques are not filled ..",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "BACKGROUN D",
"sec_num": null
},
{
"text": "In the context of the MUC-3 task, slots fall into one of four categories depending on the typ e of fill that are applicable to them . Our system is designed to handle slots whose values are from a set-list . More specifically we process TYPE OF INCIDENT, CATEGORY OF INCIDENT, PERPETRATOR :",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "BACKGROUN D",
"sec_num": null
},
{
"text": "In addition, two slots whose fill s are of string type are also processed . These are PERPETRATOR : ID OF ORG(S) and LOCATION O F",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "CONFIDENCE, PHYSICAL TARGET TYPE, HUMAN TARGET TYPE, INSTRUMENT : TYPE(S), EFFEC T ON PHYSICAL TARGET(S) and EFFECT ON HUMAN TARGET(S) .",
"sec_num": null
},
{
"text": "As will be explained later, the system consists of an Indexing Module, a Learning Module, a Filtering Module, and a Template Filler Module . We had developed and experimentally validated indexing and learning techniques for use in the context of information retrieval and classification . These techniques were adapted to develop the indexing and learning modules for MUC-3 in additio n to the development and implementation of other modules . This site did not participate in eithe r MUC-1 or MUC-2 .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "INCIDENT .",
"sec_num": null
},
{
"text": "A popular strategy in IR is to formulate the problem of identifying items relevant to a subject area as one of conceptual categorization . The subject area(s) of interest is imagined as a concept or a class . Example items relevant to a certain concept are assumed to be given . Based on this informatio n and using techniques for learning from examples, a concept characterization rule, that is optimal i n a certain precise sense, is derived . In other words, retrieval of relevant items is actually viewed as a \"recognition\" problem .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "OVERVIEW OF THE SYSTEM",
"sec_num": null
},
{
"text": "Our system employs the above idea by mapping possible fill values of set-list type slots to concept s of interest. For example, in the context of the TYPE OF INCIDENT slot, fill values such as ARSON , MURDER, BOMBING, etc . are the concepts to be learned . Note that the question of whether concep t ARSON is applicable to a message is equivalent to deciding whether a message belongs to the messag e class identified by the label ARSON . Thus, the template filling task and the decision of whether a message is relevant to MUC-3 task are investigated as problems requiring conceptual categorization .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "OVERVIEW OF THE SYSTEM",
"sec_num": null
},
{
"text": "For each concept that is considered by the system to be applicable to a message, the system als o keeps track of the extent to which each of the paragraphs in the message contributed to this decision . Judicious use of this information enables various important activities such as the resolution of th e \"best\" fill for a slot from among alternatives, the linking of the fills to templates when more tha n one template must be generated for the same message, and the filling of the two string type slots .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "OVERVIEW OF THE SYSTEM",
"sec_num": null
},
{
"text": "The general architecture of the system is presented in Figure 1 . There are four major subsystems : Indexing Module, Learning Module, Filtering Module, and Template Filler Module . Each of thes e systems are outlined next .",
"cite_spans": [],
"ref_spans": [
{
"start": 55,
"end": 63,
"text": "Figure 1",
"ref_id": "FIGREF0"
}
],
"eq_spans": [],
"section": "OVERVIEW OF THE SYSTEM",
"sec_num": null
},
{
"text": "The function of the Indexing Module is to generate a representation for each message . A messag e is represented by a vector of weights . Each weight value either indicates the presence or absence o f a term in the message or the importance of a term to the message . A term is either a single-orde r term or a high-order term (i .e ., single terms or word combinations representing phrases) .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Indexing Modul e",
"sec_num": null
},
{
"text": "For the assignment of single terms to messages, the indexing module from the SMART Retrieval System [1] is used . This module utilizes a stop list to filter out the common words and the \"nostemming\" option is chosen . All terms that are assigned a weight larger than a threshold by thi s module are retained in the message representation vector .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Indexing Modul e",
"sec_num": null
},
{
"text": "For the purpose of phrase extraction, a modified version of the INDEX software, developed and implemented by Jones, et al. [2, 3] , is used . INDEX is used mainly to extract all possible substring s that are within certain minimum and maximum length specifications and are not substrings of othe r previously selected substrings . Several strategies for filtering these to identify \"good\" phrases ar e provided as a part of the software developed for the project .",
"cite_spans": [
{
"start": 109,
"end": 115,
"text": "Jones,",
"ref_id": null
},
{
"start": 116,
"end": 126,
"text": "et al. [2,",
"ref_id": null
},
{
"start": 127,
"end": 129,
"text": "3]",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Indexing Modul e",
"sec_num": null
},
{
"text": "Thus, each element of the vector representing a message corresponds to either a single term o r a phrase . The phrase identification is expected to be important as a precision improving device . This module also generates the system vocabulary, which consists of all the distinct single terms an d phrases used in representing the messages .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Indexing Modul e",
"sec_num": null
},
{
"text": "The function of the learning module is to derive the concept categorization rules for the variou s concepts of interest . Each rule is a vector of numeric weights, where the elements correspond to th e terms in the system vocabulary. This module also involves components for selecting a training set from the development set , identifying the concepts for which the training set has at least a minimum number of positiv e examples (i .e ., the learnable concepts) and preparing the grid file, which shows for each message i n the training set which of the learnable concepts are applicable . The source for this information is the set of key-templates manually generated for the 1300 messages in the development set .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Learning Module",
"sec_num": null
},
{
"text": "The concept rule vectors are derived by employing the perceptron -learning algorithm [4] . Th e algorithm is simple and efficient . The procedure is incremental in that the rule can be updated as new examples become available. As long as a decision boundary exists, this algorithm is guaranteed to find one and terminate .",
"cite_spans": [
{
"start": 85,
"end": 88,
"text": "[4]",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "Learning Module",
"sec_num": null
},
{
"text": "Usually, the decision boundary constructed is a hyper-plane . However, since the system vocabulary includes phrases, and phrases incorporate dependency information between single terms, ou r result is equivalent to constructing a non-linear boundary. In the terminology of the connectionis t networks, we employ a single-layer, high-order perceptron . The single-layer option facilitates fast learning time, while the higher-order option enables the use of more powerful separation boundaries . Furthermore, the concept rule vectors are connectionist, rather than symbolic in nature . Such rules are more attractive when a large number of features are involved and when robustness agains t noisyness in features is crucial .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Learning Module",
"sec_num": null
},
{
"text": "In addition to concepts associated with slot fills, another concept known as \"optimal-query \" is also derived . This rule vector distinguishes messages that are not relevant to Muc-3 task from those that generate at least one template. The system is set up in such a way that the training set of messages for deriving this optimal-query vector can be different from that used for the othe r concepts.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Learning Module",
"sec_num": null
},
{
"text": "The Filtering Module is responsible for identifying concepts applicable to a set of test messages and deciding whether a message is relevant to the MuC-3 task . The major subsystems of this modul e are concerned with test message indexing, assessment of concept relevance and the evaluation of a rule base by means of an inference engine .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Filtering modul e",
"sec_num": null
},
{
"text": "The test message indexing involves the determination of which of the single terms and phrase s in the system vocabulary are contained in the message . This process generates a message vecto r that is matched against each of the concept rule vectors to determine the corresponding activatio n values . The distribution of the activation values for the test set of messages relative to each concept i s analyzed to determine a threshold . A concept is considered relevant to a message if the correspondin g activation value exceeds the threshold chosen for that concept . Depending on the concepts applicabl e to a message, the inference engine activates appropriate rules of the rule base, whose terminal symbol s correspond to the various concepts acquired . The rule base expresses the requirements in terms o f concept combinations that, when present in a message, imply that the message is relevant to Muc-3 task . The module also identifies for each message, the extent to which its paragraphs contribute d to the activation values relative to the different concepts . This result is referred to as the concep t vs . paragraph relevance vector .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Filtering modul e",
"sec_num": null
},
{
"text": "For slots of string fill type, a database of possible fill values, grouped by slot name, is provide d as input to this module . For each string in the database for which at least one match is found i n the message, the paragraphs in which a match is found and the frequency of its occurrence in each paragraph is determined .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Filtering modul e",
"sec_num": null
},
{
"text": "This module is responsible for generating one or more templates for each message determined to b e relevant by the Filtering Module and filling the slots on the basis of concepts and string filled tha t are activated .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Template Filler Modul e",
"sec_num": null
},
{
"text": "For each relevant message either the optimal-query concept is activated or one or more inciden t types are recognized along with a desired combination of concepts (or both) . In the case exactly one incident type is recognized, for each of the other slots the following is performed . If several concepts are activated for this slot and only one value is permitted, the one with highest activation value i s chosen ; otherwise, all values are filled .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Template Filler Modul e",
"sec_num": null
},
{
"text": "In the case more than one incident type is activated, the system must decide, for each activate d concept, to which incident type it is the closest . For this purpose, the concept versus paragrap h relevance vector is used . This vector contains the contribution of the various paragraphs in a messag e to the activation value of the slot fill relative to this message . The paragraph relevance vector of an activated slot fill, say CIVILIAN (from HUMAN TARGET TYPE), is compared to the vector associated with each of the activated incident types, say KIDNAPPING and MURDER. The strength of this match is then used to decide whether the fill CIVILIAN will be used in the KIDNAPPING or the MURDER template .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Template Filler Modul e",
"sec_num": null
},
{
"text": "If a message becomes relevant only due to optimal-query, then it enables other slots having activated fills to be filled even though no TYPE OF INCIDENT may have been activated .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Template Filler Modul e",
"sec_num": null
},
{
"text": "The system walkthrough explains how the message TsT1-MUC3-0099 is processed . The result obtained corresponds to the parameter settings used in our Option 4 (see report on UNL/USL : Muc-3 Test Results and Analysis) . In this option, Training Set 2 is used for determining the rule vector fo r optimal-query and Training Set 3 is used for the other concepts . The threshold used for decidin g whether a concept is activated is based on an analysis of the distribution of the activation values o f this concept relative to the test set messages (threshold setting T1) . Table 1 shows a list of all set-list type fills and those that are actually learnable on the basis of training set 3 . The concept rule vectors for each of these fills are constructed by using the indexin g and the learning module . The test message is indexed and the dot product of its representation vecto r with each of the concept rule vectors is computed . The activation values so obtained are compare d to the corresponding threshold values . Table 2 shows that, for the current message, the following five concepts are activated : BOMBING, TERRORIST ACT, TRANSPORT VEHICLE, SOME DAMAGE, and th e optimal-query. These concepts activate the appropriate leaf nodes of the AND/OR tree associated with the rulebase shown in Table 4 . This results in the root node getting the value \"true\" an d therefore, this message is termed relevant . For the current testing, the rule base is defined with all the concept weights being either 0 or 1 . The inference engine is, however, capable of handlin g any numeric weights between 0 and 1 . The vector representation for each of the paragraphs in the message are also multiplied by the concept rule vectors to obtain the paragraph vs concept relevanc e vector (Table 3) . This paragraph information is not useful in this case since neither several fills are activated for a slot for which only one fill is permitted nor is there an indication, in terms o f INCIDENT TYPE activations, that multiple templates should be created .",
"cite_spans": [],
"ref_spans": [
{
"start": 568,
"end": 575,
"text": "Table 1",
"ref_id": "TABREF2"
},
{
"start": 1019,
"end": 1026,
"text": "Table 2",
"ref_id": "TABREF3"
},
{
"start": 1296,
"end": 1303,
"text": "Table 4",
"ref_id": null
},
{
"start": 1774,
"end": 1783,
"text": "(Table 3)",
"ref_id": null
}
],
"eq_spans": [],
"section": "SYSTEM WALKTHROUG H",
"sec_num": null
},
{
"text": "For the two string fill slots the matching strings along with their occurrence frequency in th e various paragraphs is shown in Table 5 . The paragraph vector for BOMBING is found to match the paragraph vector of \"POLICE\" better (wrong decision!) . All 3 incident locations have a positiv e activation value with BOMBING . Since the location slot permits multiple fills, all three may b e retained . However, since \"PRC\" is not one of the South American countries, it is discarded .",
"cite_spans": [],
"ref_spans": [
{
"start": 128,
"end": 135,
"text": "Table 5",
"ref_id": null
}
],
"eq_spans": [],
"section": "SYSTEM WALKTHROUG H",
"sec_num": null
},
{
"text": "The filled template for this message is shown in Table 6 . This template most closely matches keytemplate that is numbered 2 (see Appendix H) . The paragraph relevance vector matching techniqu e needs to be refined as evidenced by the choice of \"POLICE\" as the perpetrator organization . Furthermore, template filler module should be refined to automatically determine and incorporate in th e filling process various dependencies between template fills . For example, \"POLICE \" is inconsistent with CATEGORY OF INCIDENT being TERRORIST ACT .",
"cite_spans": [],
"ref_spans": [
{
"start": 49,
"end": 56,
"text": "Table 6",
"ref_id": null
}
],
"eq_spans": [],
"section": "SYSTEM WALKTHROUG H",
"sec_num": null
},
{
"text": "By proper modification of the stop list used during phrase extraction, phrases such as NO INJURY could be extracted . The optimal-query vector identifies relevant passages fairly accurately . Carefu l detailed analysis of individual instances should lead to many ideas for improvement . Output file conceptl 1 ARSON slot-3 concept2 2 ARSON THREAT slot-3 concept3 3 ../ MURDER slot -concept4 4 .~/ DEATH THREAT slot -concept5 5 ,/ BOMBING slot -concept8 8 BOMB THREAT slotconcept? 7 ,/ KIDNAPPING slot -concept8 8 KIDNAPPING THREAT slot -concept9 9 HIJACKING slot-3 conceptl0 10 HIJACKING THREAT ",
"cite_spans": [],
"ref_spans": [
{
"start": 287,
"end": 631,
"text": "Output file conceptl 1 ARSON slot-3 concept2 2 ARSON THREAT slot-3 concept3 3 ../ MURDER slot -concept4 4 .~/ DEATH THREAT slot -concept5 5 ,/ BOMBING slot -concept8 8 BOMB THREAT slotconcept? 7 ,/ KIDNAPPING slot -concept8 8 KIDNAPPING THREAT slot -concept9 9 HIJACKING slot-3 conceptl0 10 HIJACKING THREAT",
"ref_id": "TABREF2"
}
],
"eq_spans": [],
"section": "SYSTEM WALKTHROUG H",
"sec_num": null
},
{
"text": "CEV Cutoff Concept Name",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Concept Name",
"sec_num": null
}
],
"back_matter": [
{
"text": "Note : The highlighted predicates have true values . Lower case predicates are terminals, whereas th e ones in upper case are non-terminals . The organizations in TST1-MUC3-0099 #TST1-MUC3-0099 POLICE 1 :1 3 : 1 SHINING PATH 4 :1 5 :1 6 :1 7:1 8 : 1The locations in TST1-MUC3-0099 #TST1-MUC3-0099 LIMA 1 :1 2 :1 8: 1 PRC 2 : 1 SAN ISIDRO 2 : 1 EFFECT ON HUMAN TARGET(S) - Table 6 : The filled template for TsT1-MUC3-009 9",
"cite_spans": [],
"ref_spans": [
{
"start": 372,
"end": 379,
"text": "Table 6",
"ref_id": null
}
],
"eq_spans": [],
"section": "annex",
"sec_num": null
},
{
"text": "1 . Buckley, C . (1985) , \"Implementation of the SMART information retrieval system\", TR 85 -686 , Dept . of Computer Science, Cornell University, Ithaca, NY .2 . L . P. Jones, E . W . Gassie. Jr ., S . Radhakrishnan, \"INDEX : Statistical basis for an automatic conceptual phrase-index system\", Journal of American Society for Information Science, Vol . 41, pp . 87 -91, 1990 . 3 . L . P. Jones, E. W . Gassie. Jr ., S . Radhakrishnan, \"PORTREP : A portable repeated strin g finder\", Software Practice and Experience, Vol . 19, pp . 63 -77, 1989 . 4. Duda, R . O . and Hart, P. E . (1973) , Pattern Classification and Scene Analysis, Wiley, NY .",
"cite_spans": [
{
"start": 4,
"end": 23,
"text": "Buckley, C . (1985)",
"ref_id": null
},
{
"start": 339,
"end": 377,
"text": "Science, Vol . 41, pp . 87 -91, 1990 .",
"ref_id": null
},
{
"start": 518,
"end": 568,
"text": "Vol . 19, pp . 63 -77, 1989 . 4. Duda, R . O . and",
"ref_id": null
},
{
"start": 569,
"end": 588,
"text": "Hart, P. E . (1973)",
"ref_id": null
}
],
"ref_spans": [],
"eq_spans": [],
"section": "REFERENCE S",
"sec_num": null
},
{
"text": "The papers in this section provide two general perspectives on MUC-3 tha t came out of the evaluation task .The first paper describes an experimental syste m based on a statistical text categorization technique . The results of testing tha t system on the MUC-3 task help in assessing the difficulty of the task and th e appropriateness of text categorization as an element of a complete informatio n extraction system . The second one is a joint paper prepared by representatives o f seven of the participating sites on the subject of discourse analysis as it pertains t o MUC-3 . The desire to offer this paper arose from the common perception of th e discourse handling demands placed on the systems by the MUC-3 corpus and task .",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "PART IV: OTHER CONTRIBUTED PAPER S",
"sec_num": null
}
],
"bib_entries": {},
"ref_entries": {
"FIGREF0": {
"uris": null,
"type_str": "figure",
"text": "Block diagram of the system",
"num": null
},
"TABREF0": {
"type_str": "table",
"num": null,
"html": null,
"text": "",
"content": "<table><tr><td>Trainin g Set Messages</td><td>Indexing Module</td><td>s</td><td>Syste m Vocabulary</td></tr><tr><td/><td/><td>Indexed</td><td/></tr><tr><td/><td/><td>Training</td><td/></tr><tr><td/><td/><td>Set</td><td/></tr><tr><td/><td/><td>Message s</td><td/></tr><tr><td>Training Set Key Template</td><td>Learning Module</td><td>s. Concep t Rul e Vectors</td><td/></tr></table>"
},
"TABREF2": {
"type_str": "table",
"num": null,
"html": null,
"text": "List of Set List Slot Fill s",
"content": "<table><tr><td>CID</td></tr></table>"
},
"TABREF3": {
"type_str": "table",
"num": null,
"html": null,
"text": "",
"content": "<table/>"
}
}
}
} |