File size: 30,547 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
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
{
    "paper_id": "W98-0143",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T06:03:33.512584Z"
    },
    "title": "Consistent Grammar Development Using Partial-Tree Descriptions for Lexicalized Tree-Adjoining Grammars",
    "authors": [
        {
            "first": "Fei",
            "middle": [],
            "last": "Xia",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "Cognitive Science University of Pennsylvania 400A",
                "location": {
                    "postCode": "3401, 19104",
                    "settlement": "Walnut Street Philadelphia",
                    "region": "PA",
                    "country": "USA"
                }
            },
            "email": ""
        },
        {
            "first": "Martha",
            "middle": [],
            "last": "Palmer",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "Cognitive Science University of Pennsylvania 400A",
                "location": {
                    "postCode": "3401, 19104",
                    "settlement": "Walnut Street Philadelphia",
                    "region": "PA",
                    "country": "USA"
                }
            },
            "email": ""
        },
        {
            "first": "K",
            "middle": [],
            "last": "Vijay-Shanker",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "Cognitive Science University of Pennsylvania 400A",
                "location": {
                    "postCode": "3401, 19104",
                    "settlement": "Walnut Street Philadelphia",
                    "region": "PA",
                    "country": "USA"
                }
            },
            "email": ""
        },
        {
            "first": "Joseph",
            "middle": [],
            "last": "Rosenzweig",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "Cognitive Science University of Pennsylvania 400A",
                "location": {
                    "postCode": "3401, 19104",
                    "settlement": "Walnut Street Philadelphia",
                    "region": "PA",
                    "country": "USA"
                }
            },
            "email": "[email protected]"
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "",
    "pdf_parse": {
        "paper_id": "W98-0143",
        "_pdf_hash": "",
        "abstract": [],
        "body_text": [
            {
                "text": "An important characteristic of an FB-LTAG is that it is lexicalized, i.e., each lexical item is anchored to a tree structure that encodes subcategorization information. Trees with the same canonical subcategorizations are grouped into tree families. The reuse of tree substructures, such as wh-movement, in many different trees creates redundancy, which poses a problem for grammar development and maintenance {Vijay-Shanker and Schabes, 1992) . To consistently implement a change in some general aspect of the design of the grammar, all the relevant trees currently must be inspected and edited. Vijay Shanker and Schabes suggested the use of hierarchical organization and of tree descriptions to specify substructures that would be present in several elementary trees of a grammar. Since then, in addition to ourselves, Becker, {Becker, 1994) , Evans et al. {Evans et al., 1995), and Candito{Candito, 1996) have developed systerns for organizing trees of a TAG which could be used for developing and maintaining grammars.",
                "cite_spans": [
                    {
                        "start": 429,
                        "end": 443,
                        "text": "Schabes, 1992)",
                        "ref_id": "BIBREF5"
                    },
                    {
                        "start": 822,
                        "end": 844,
                        "text": "Becker, {Becker, 1994)",
                        "ref_id": "BIBREF0"
                    },
                    {
                        "start": 847,
                        "end": 885,
                        "text": "Evans et al. {Evans et al., 1995), and",
                        "ref_id": "BIBREF3"
                    },
                    {
                        "start": 886,
                        "end": 908,
                        "text": "Candito{Candito, 1996)",
                        "ref_id": "BIBREF1"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Our system is based on the ideas expressed in Vijay-Shanker and Schabes, (Vijay-Shanker and Schabes, 1992) , to use partial-tree descriptions in specifying a grammar by separately defining pieces of tree structures to encode independent syntactic principles. Various individual specifications are then combined to form the elementary trees of the grammar. Our paper begins with a description of our grammar development system and the process by which it generates the Penn English grammar as well as a Chinese TAG. We describe the significant properties of both grammars, pointing out the major differences between them, and the methods by which our system is informed about these language-specific properties. We then compare our approach to other grammar development approaches for LTAG such as the specification of TAGs in DATR {Evans et al., 1995) and Can-dito1s implementation {Candito, 1996) .",
                "cite_spans": [
                    {
                        "start": 73,
                        "end": 106,
                        "text": "(Vijay-Shanker and Schabes, 1992)",
                        "ref_id": "BIBREF5"
                    },
                    {
                        "start": 818,
                        "end": 851,
                        "text": "TAGs in DATR {Evans et al., 1995)",
                        "ref_id": null
                    },
                    {
                        "start": 856,
                        "end": 897,
                        "text": "Can-dito1s implementation {Candito, 1996)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "In our approach, three types of componentssubcategorization frames, blocks and lexical redistribution rules -are used to describe lexical and syntactic information. Actual trees are generated automatically from these abstract descriptions. In maintaining the grammar only the abstract descriptions need ever be manipulated; the tree descriptions and the actual trees which they subsume are computed deterministically from these high-level descriptions.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "System Overview",
                "sec_num": "2"
            },
            {
                "text": "Subcategorization frames specify the category of the main anchor, the number of arguments, each argument's category and position with respect to the anchor, and other information such as feature equations or node expansions. Each tree family has one canonical subcategorization frame.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Subcategorization frames",
                "sec_num": "2.1"
            },
            {
                "text": "Blocks are used to represent the tree substructures that are reused in different trees, i.e. blocks subsume classes of trees. Each block includes a set of nodes, dominance relation, parent relation, precedence relation between nodes, and feature equations. This follows the definition of the tree descriptions specified in a logical language patterned after Rogers and Vijay-Shanker(Rogers and Vijay-Shanker 1 1994).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Blocks",
                "sec_num": "2.2"
            },
            {
                "text": "Blocks are divided into two types according to their functions: subcategorization blocks and transformation blocks. The former describes structural configurations incorporating the various information in a subcategorization frame. For example, some of the subcategorization blocks used in the development of the English grammar are shown in Figure 1 . 1 When the subcategorization frame for a verb is given by the grammar developer, the system will automatically create a new block (of code) by essentially selecting the appropriate primitive subcategorization blocks corresponding to the argument information specified in that verb frame.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 341,
                        "end": 349,
                        "text": "Figure 1",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Blocks",
                "sec_num": "2.2"
            },
            {
                "text": "The transformation blocks are used for various transformations such as wh-movement. These transformation blocks do not encode rules for modifying trees, but rather describe the properties of a particular syntactic construction. Figure 2 depicts our representation of phrasal extraction. This can be specialized to give the blocks for wh-movement, topicalization, relative clause formation, etc. For example, the wh-movement block is defined by further specifying that the ExtractionRoot is labeled S, the NewSite ha.s a +wh feature 1 and so on.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 228,
                        "end": 236,
                        "text": "Figure 2",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Blocks",
                "sec_num": "2.2"
            },
            {
                "text": "1 In order to focus on the use of tree descriptions and to make the figures less cumbersome, we show only the structural aspects and do not show the feature value specification. The parent, (immediate dominance), relationship is illustrated by a plain line and the dominance relationship by a dotted line. The arc between nodes shows the precedence order of the nodes are unspecified. The nodes' categories are enclosed in parentheses.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Blocks",
                "sec_num": "2.2"
            },
            {
                "text": "Rooi The third type of machinery available for a grammar developer is the Lexical Redistribution Rule (LRR). An LRR is a pair (r1 1 rr) of subcategorization frames, which produces a new frame when applied to a subcategorization frame s, by first matching2 the left frame ri of r to s, then combining information in rr and s. LRRs are introduced to incorporate the connection between subcategorization frames. For example, most transitive verbs have a frame for active(a subject and an object) and another frame for passive, where the object in the former frame becomes the subject in the latter. An LRR, denoted as passive LRR, is built to produce the passive subcategorization frame from the active one. Similarly, applying dative-shift LRR to the frame with one NP subject and two NP objects will produce a frame with an NP subject and an PP object. Besides the distinct content, LRRs and blocks also differ in several aspects:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "AnchorP(' VP' } / \\. PredP Anchol(' V' ) 1 Pre-0 (a} is_main_frame Rooi PredP Subject AncOO' P ~ Pred Object (b)jlred_ha.1_subj\u00abt (c~red_ha.1_obj\u00abt",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "\u2022 They have different functionalities: Blocks represent the substructures that are reused in different trees. They are used to reduce the redundancy among trees; LRRs are introduced to incorporate the connections between the closely related subcategorization frames.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "\u2022 Blocks are strictly additive and can be added in any order. LRRs, on the other hand, produce different results depending on the order they are applied in, and are allowed to be non-additive, i.e\" to remove information from the subcategorization frame they are being applied to, as in the procedure of passive from active.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "EQUATION",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [
                    {
                        "start": 0,
                        "end": 8,
                        "text": "EQUATION",
                        "ref_id": "EQREF",
                        "raw_str": "s, s, s, / \\ / \\ / \\ N\"! S N\"! S N\"! S /\"' /\"\". /\"\". NP j VP NP j VP NP j VP . /~ . /~ 1 / \" ' v. NP 1 PP 1 v. NP 1 1 PP 1 v. PP 1 j / \\ / \\ / \\ t p' NP 2 j p' NP 1 r, NP 1 1 1 1 l 1 \" .. \" (3) (b)",
                        "eq_num": "(c)"
                    }
                ],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "Figure 3: Elementary trees generated from combining blocks",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "181",
                "sec_num": null
            },
            {
                "text": "To generate elementary trees, we begin with a canonical subcategorization frame. The system will first generate related subcategorization frames by applying LRRs, then select subcategorization blocks corresponding to the information in the subcategorization frames, next the combinations of these blocks are further combined with the blocks corresponding to various transformations, finally, a set of trees are generated from those combined blocks, and they are the tree family for this subcategorization frame. Figure 3 shows some of the trees produced in this way. For instance, the last tree is obtained by incorporating information from the ditransitive verb subcategorization frame, applying t\u2022he dative-shift and passive LRRs, and then combining them with the wh-non-subject extraction block.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 512,
                        "end": 520,
                        "text": "Figure 3",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Tree generation",
                "sec_num": "2.4"
            },
            {
                "text": "3 Generating grammars",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Tree generation",
                "sec_num": "2.4"
            },
            {
                "text": "We have used our tool to specify a grammar for English in order to produce the trees used in the current English XTAG grammar. We have also used our tool to generate a large grammar for Chinese. In designing these grammars, we have tried to specify the grammars to refiect the similarities and the differences between the languages. The major features of our specification of these two grammars are summarized in Table   1 . By focusing on the specification of individual grammatical information, we have been able to generate nearly all of the trees (91.3% -638 out of the 699) from the tree families used in the current English grammar developed at Penn 3 . Our approach, has also exposed certain gaps in the Penn grammar. We are encouraged with the utility of our tool and the ease with which this large-scale grammar was developed.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 413,
                        "end": 422,
                        "text": "Table   1",
                        "ref_id": "TABREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Tree generation",
                "sec_num": "2.4"
            },
            {
                "text": "We are currently working on expanding the contents of subcategorization frame to include trees for other categories of words. For example, a frame \\11hich has no specifier and one NP complement and whose predicate is a preposition will correspond to PP -t P NP tree. We'll also introduce a modifier field and semantic fea-tures, so that the head features will propagate from modifiee to modified node, while non-head features from the predicate as the head of the modifier will be passed to the modified node.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Tree generation",
                "sec_num": "2.4"
            },
            {
                "text": "Evans, Gazdar and Weir (Evans et al., 1995) also discuss a method for organizing the trees in a TAG hierarchically, using an existing lexical representational system, DATR (Evans and Gazdar, 1989) . Since DATR can not capture directly dominance relation in the trees, these must be simulated by using feature equations.",
                "cite_spans": [
                    {
                        "start": 23,
                        "end": 43,
                        "text": "(Evans et al., 1995)",
                        "ref_id": "BIBREF3"
                    },
                    {
                        "start": 172,
                        "end": 196,
                        "text": "(Evans and Gazdar, 1989)",
                        "ref_id": "BIBREF2"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Comparison to Other Work",
                "sec_num": "4"
            },
            {
                "text": "There are substantial similarities and significant differences in our approach and Candito's approach, which she applied primarily to French and Italian. Both systems have built upon the basic ideas expressed in (Vijay-Shanker and Schabes, 1992) for organizing trees hierarchically and the use of tree descriptions that encode substructures found in several trees. The main difference is how Candito uses her dimensions in generating the trees. Her system imposes explicit conditions on how the classes appearing in the hierarchy can be combined, based on w hich dimension they are in. For example, one condition states that only a terminal node (leaf node of a hierarchy) of the second dimension can be used in constructing a tree. Therefore two redistributions (such as passive and causative) can be used in a single tree only when a new passive-causative terminal node is first created manually. In contrast, our approach automatically considers all possible applications of LRRs, and discards those that are inconsistent.",
                "cite_spans": [
                    {
                        "start": 212,
                        "end": 245,
                        "text": "(Vijay-Shanker and Schabes, 1992)",
                        "ref_id": "BIBREF5"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Comparison to Other Work",
                "sec_num": "4"
            },
            {
                "text": "We have described a tool for grammar development in which tree descriptions are used to provide an abstract specification of the linguistic phenomena relevant to a particular language. In grammar development and maintenance, only the abstract specifications need to be edited, and any changes or corrections will automatically be proliferated throughout the grammar. In addition to lightening the more tedious aspects of grammar maintenance, this approach 183 also allows a unique perspective on the general characteristics of a language. Defining hierarchical blocks for the grammar both necessitates and facilitates an examination of the linguistic assumptions that have been made with regard to feature specification and tree-family definition. This can be very useful for gaining an overview of the theory that is being implemented and exposing gaps that remain unmotivated and need to be investigated. The type of gaps that can be exposed could include a missing subcategorization frame that might arise from the automatic combination of blocks and which would correspond to an entire tree family, a missing tree which would represent a particular type of transformation for a subcategorization frame, or inconsistent feature equations. By focusing on syntactic properties at a higher level, our approach allows new opportunities for the investigation of how languages relate to themselves and to each other.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion",
                "sec_num": "5"
            },
            {
                "text": "Matching occurs successfully when frame s is compatible with rz in the type of anchors, the number of arguments, their positions, categories and feat.nres . In other words, incompatible features etc. will block certain LR.fu from being applied.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "We have not yet attempted to extend our coverage to include punctuation, it-clefts, and a few idiosyncratic analyses that are included in the sixty trees we are not generating.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Patterns in metarules",
                "authors": [
                    {
                        "first": "Tilman",
                        "middle": [],
                        "last": "Becker",
                        "suffix": ""
                    }
                ],
                "year": 1994,
                "venue": "Proceedings of the 3rd TAG+ Conference",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Tilman Becker. 1994. Patterns in metarules. In Proceedings of the 3rd TAG+ Conference, Paris, France.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "A principle-ba.sed hierarchical representation of !tags",
                "authors": [
                    {
                        "first": "Marie-Helene",
                        "middle": [],
                        "last": "Candito",
                        "suffix": ""
                    }
                ],
                "year": 1996,
                "venue": "Proceedings of COLING-96",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Marie-Helene Candito. 1996. A principle-ba.sed hi- erarchical representation of !tags. In Proceedings of COLING-96, Copenhagen, Denmark.",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "Inference in datr",
                "authors": [
                    {
                        "first": "Roger",
                        "middle": [],
                        "last": "Evans",
                        "suffix": ""
                    },
                    {
                        "first": "Gerald",
                        "middle": [],
                        "last": "Gazdar",
                        "suffix": ""
                    }
                ],
                "year": 1989,
                "venue": "EACL-89",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Roger Evans and Gerald Gazdar. 1989. Inference in datr. In EACL-89.",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "Encoding Lexicalized Tree Adjoining Grammars with a Nonmonotonic Inheritance Hierarchy",
                "authors": [
                    {
                        "first": "Roger",
                        "middle": [],
                        "last": "Evans",
                        "suffix": ""
                    },
                    {
                        "first": "Gerald",
                        "middle": [],
                        "last": "Gazdar",
                        "suffix": ""
                    },
                    {
                        "first": "David",
                        "middle": [],
                        "last": "Weir",
                        "suffix": ""
                    }
                ],
                "year": 1995,
                "venue": "Proceedings of the 33rd Annual Meeting of the Association for Computational Linguistics(ACL '95)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Roger Evans, Gerald Gazdar, and David Weir. 1995. Encoding Lexicalized Tree Adjoining Grammars with a Nonmonotonic Inheritance Hierarchy. In Proceedings of the 33rd Annual Meeting of the Association for Computational Linguistics(ACL '95), Cambridge, MA.",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Obtaining Trees from their Descriptions: An Application to Tree Adjoining Grammars",
                "authors": [
                    {
                        "first": "James",
                        "middle": [],
                        "last": "Rogers",
                        "suffix": ""
                    },
                    {
                        "first": "K",
                        "middle": [],
                        "last": "Vijay-Shanker",
                        "suffix": ""
                    }
                ],
                "year": 1994,
                "venue": "Computational Intelligence",
                "volume": "10",
                "issue": "4",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "James Rogers and K. Vijay-Shanker. 1994. Obtain- ing Trees from their Descriptions: An Application to Tree Adjoining Grammars. Computational In- telligence, 10( 4).",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Structure sharing in lexicalized tree adjoining grammar",
                "authors": [
                    {
                        "first": "K",
                        "middle": [],
                        "last": "Vijay-Shanker",
                        "suffix": ""
                    },
                    {
                        "first": "Yves",
                        "middle": [],
                        "last": "Schabes",
                        "suffix": ""
                    }
                ],
                "year": 1992,
                "venue": "Proceedings of the 15th International Conference on Computational Linguistics (COL-ING '92)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "K. Vijay-Shanker and Yves Schabes. 1992. Struc- ture sharing in lexicalized tree adjoining gram- mar. In Proceedings of the 15th International Conference on Computational Linguistics (COL- ING '92), Nantes, France.",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "text": "Figure 1: Some subcategorization blocks ExtractionRoot /",
                "uris": null,
                "num": null,
                "type_str": "figure"
            },
            "TABREF1": {
                "text": "Major features of English and Chinese grammars",
                "html": null,
                "content": "<table/>",
                "type_str": "table",
                "num": null
            }
        }
    }
}