| { | |
| "paper_id": "W98-0110", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T06:05:11.504352Z" | |
| }, | |
| "title": "Synchronous TFG for Speech Translation", | |
| "authors": [ | |
| { | |
| "first": "Marc", | |
| "middle": [], | |
| "last": "Cavazza", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "University of Bradford Bradford", | |
| "location": { | |
| "addrLine": "BD7 lDP", | |
| "country": "United Kingdorn" | |
| } | |
| }, | |
| "email": "[email protected]" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "", | |
| "pdf_parse": { | |
| "paper_id": "W98-0110", | |
| "_pdf_hash": "", | |
| "abstract": [], | |
| "body_text": [ | |
| { | |
| "text": "The use of synchronous TAG for Machine Translation has been described by Abeille et al. [1990] and has resulted in several implementations [Prigent, 1994] , mainly developed using the XTAG system [Paroubek et al., 1992] . While we subscribe to the general arguments in favour of the use of TAG+ for Machine Translation, it appears that speech translation could constitute an ideal application of these ideas [Harbusch & Poller, 1994] . lt is actually easier to select specific areas where speech translation is both feasible and of practical impact (see e.g. the CST AR, VERBMOBIL and SRI Speech Translation projects). In this paper, we report the implementation of a minimal speech translation prototype based on synchronous TAG+ (more exactly, synchronous Tree Furcating Granunars or STFG), which has been developed as a direct extension of our TFG parser [Cavazza, 1998] .", | |
| "cite_spans": [ | |
| { | |
| "start": 73, | |
| "end": 94, | |
| "text": "Abeille et al. [1990]", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 139, | |
| "end": 154, | |
| "text": "[Prigent, 1994]", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 196, | |
| "end": 219, | |
| "text": "[Paroubek et al., 1992]", | |
| "ref_id": "BIBREF9" | |
| }, | |
| { | |
| "start": 408, | |
| "end": 433, | |
| "text": "[Harbusch & Poller, 1994]", | |
| "ref_id": "BIBREF7" | |
| }, | |
| { | |
| "start": 858, | |
| "end": 873, | |
| "text": "[Cavazza, 1998]", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction: Synchronous TAG+ for Machine Translation", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "Sheiber originally coined the tenn \"synchronous TAG\". They described synchronous derivation of semantic structures from tree operations carried on lexicalised trees. A synchronous TAG is thus a pair of two elementary trees, one representing the source language and the other a logical formula, which is also represented as a variant of TAG (and is lexicalised as weil). However, the term of synchronous TAG, when used for machine translation, actually subsumes different approaches and deserves some clarification. The initial presentation of TAG for Machine Translation by Abeille et al.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction: Synchronous TAG+ for Machine Translation", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "(1990] referred to synchronous TAG, though in fact it directly mapped lexicalised trees to one another, without making recourse to the \"semantic\" trees described by Shieber & Schabes [1990] . In that sense, it could be considered as a transfer formalism or a structural correspondence system [Kaplan et al\" 1989] . Further implementations by Prigent [1994] within the XTAG system [Paroubek et al\" 1992] have been based on an extended transfer paradigm, mapping between derivation trees in the source and target languages, thus introducing an intermediate representation.", | |
| "cite_spans": [ | |
| { | |
| "start": 183, | |
| "end": 189, | |
| "text": "[1990]", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 292, | |
| "end": 312, | |
| "text": "[Kaplan et al\" 1989]", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 342, | |
| "end": 356, | |
| "text": "Prigent [1994]", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 380, | |
| "end": 402, | |
| "text": "[Paroubek et al\" 1992]", | |
| "ref_id": "BIBREF9" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction: Synchronous TAG+ for Machine Translation", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "On the other hand, direct mapping between lexicalised trees has also been adopted in the STAG project . We would like to suggest, adopting a terminology from Prigent [1994] , that approaches based on the direct mapping between lexicalised trees should be renamed iso-synchronous. This would clearly indicate that the synchronous trees on which adjunction (resp. substitution) operations are carried are of the same kind. Our own implementation follows the isosynchronous approach, and is based on paired elementary trees in the TFG formalism [Cavazza, 1998 ].", | |
| "cite_spans": [ | |
| { | |
| "start": 158, | |
| "end": 172, | |
| "text": "Prigent [1994]", | |
| "ref_id": "BIBREF10" | |
| }, | |
| { | |
| "start": 542, | |
| "end": 556, | |
| "text": "[Cavazza, 1998", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction: Synchronous TAG+ for Machine Translation", | |
| "sec_num": "1." | |
| }, | |
| { | |
| "text": "The overall prototype aims at demonstrating real-time speech translation of average 10-15 word sentences in spoken sublanguage areas. lt relies on off-the-shelf software both for speech recognition and text-tospeech synthesis. The speech recognition system used in our experiments is the Nuance system (from Nuance Communications), with a British English database. Speech synthesis is based on a Text-To-Speech system, in our case TTS-SDK for French (from Learnout & Hauspie). Our system takes as input an ASCil string in the source language (English), as produced by the speech recognition system, and outputs an ASCII string in the target language (French), which is passed to the 'ITS system. This is not to say that our system could be equally applied to the translation of written sublanguages, as the size and syntactic complexity of spoken and written sublanguages differ significantly 1 \u2022 The first step consists in selecting the relevant trees from the source language input. This corresponds to the lexical filtering step of the source grarnmar, and is equivalent to the construction of a set of tree stacks [Cavazza, 1998] . Each tree in the source language is associated a tree in the target language with appropriate mappings from source to target trees at roots, anchors and leaves (see below). The result is a set of candidate forests in the source language to be parsed. For each source forest, there exists an associated forest in the target language. However, it is the processing of the source language forest that fully determines the operations to be carried on the target forest. Parsing a forest involves tree fusion on the basis of adjacent categories, as described in [Cavazza, 1998] . Whenever a pair of adjacent trees (tl, t2) in the source forest undergoes a fusion operation (substitution or furcation), a synchronous operation is carried between the target pair (tl', t2'). In this way, the construction of the target sentence directly proceeds from the analysis of the source sentence. We do not resort to incremental generation of the target sentence, but delay output until the source forest has been entirely and successfully parsed.", | |
| "cite_spans": [ | |
| { | |
| "start": 1118, | |
| "end": 1133, | |
| "text": "[Cavazza, 1998]", | |
| "ref_id": "BIBREF3" | |
| }, | |
| { | |
| "start": 1693, | |
| "end": 1708, | |
| "text": "[Cavazza, 1998]", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "Simple difference in constructs between French and English, like those described in Abeille et al.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "[1990] are handled by linking arguments in the source and target node. Processing arguments in the source forest will then lead to the correct attribution of arguments in the target forest (even though their order might differ, as the parsing algorithm only relies on the source forest order). This also applies to the translation of idioms or when a simple word in the source (resp. target) language does correspond to an idiomatic construction in the target (resp. source) language. Differences in word order for adjectives, like in la clef bleue vs. the blue key, are directly reflected in the tree representations, where \"bleue\" is a left auxiliary tree *N and \"blue\" a right auxiliary tree. As a result the (N, *N) pair in the source language is matched to a (N, N*) pair for which there would be no fusion. But, because the tree operation is determined by the source forest pair, it is sufficient to adapt the fusion procedure to detect this and perform the correct operation. There are several differences in our formalisation and our implementation with respect to the original description of Abeille et al. [1990] . We establish links only between root nodes and between leaves, hence not relating nodes which are intemal to the source and target trees (e.g., \"VP\" nodes in [Abeille et al. 1990] ). This is partly due to the fact that we do not make use of intemal categories such as VP and NP, following in that sense both the description given by Abeille (1994) for French and the TFG philosophy, which aims at limiting tree depths. Another difference is that we restrict links between the source and target trees to nodes bearing the same syntactic category. This currently limits our ability to process some structural discrepancies, as in the example John gave a weak cough I John toussa faiblement, where an N*based (left) furcation in the English tree (N*-weak) would correspond to a *V-based (right) furcation in the French tree (*V-faiblement) [Abeille et al., 1990] . However, the system is currently able to process a subset of structural discrepancies. This is illustrated by figure 1., where parsing the source forest for the sentence the right door lacks a handle produces as an output il manque wie poignee a la porte de droite. Adopting the terminology of Dorr [1994] for translation divergences, we should be able to take into account mainly thematic, structural (e.g. \"shoot-NO\" vs. \"tirer-sur-NO\") and some lexical divergences. However, these points would necessitate further investigation due to the small size of our experiments.", | |
| "cite_spans": [ | |
| { | |
| "start": 1116, | |
| "end": 1122, | |
| "text": "[1990]", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 1283, | |
| "end": 1304, | |
| "text": "[Abeille et al. 1990]", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 1458, | |
| "end": 1472, | |
| "text": "Abeille (1994)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 1962, | |
| "end": 1984, | |
| "text": "[Abeille et al., 1990]", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 2281, | |
| "end": 2292, | |
| "text": "Dorr [1994]", | |
| "ref_id": "BIBREF4" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "Though the synchronous TAG approach to machine translation is essentially a kind of transfer formalism, we have augmented it with the inclusion of semantic f eatures in order to perf orm some form of syntactic disambiguation, mainly dealing with PP-attachments. These ambiguities are amenable to selectional restrictions, based on semantic features matching. lt could be argued that syntactic disambiguation is not strictly needed for French-to-English translation, as even incorrect attachments might generate correct translations (with similar ambiguities in the source and target languages). However, this would not be fully satisfactory and furthermore, accepting incorrect attachments would result in several forests being fully parsed before a result is produced.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "N* N* p N* N ~ N 1 1 1 1 1 Det A NQJ, V NJJ, det 1 1 1 . 1 door handle the right laclcs a . \u2022\u2022 .. . \u2022\u2022 \u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022\u2022 .!. \u2022\u2022\u2022\u2022\u2022\u2022 : '\u2022, : \u2022\u2022. p", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": ".: ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "\u2022\u2022~ N* *N ~! N* 1 /1 N l N Cl V N6iy~ 1 1 Det Det N", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Synchronous Processing of Source and Target Forests", | |
| "sec_num": "2." | |
| }, | |
| { | |
| "text": "A first version of the system has been developed and tested with a smaU vocabulary of Iess than 200 lexical entries. Constructs dealt with include idiomatic expressions, transitive/intransitive constructs, differences in word order, and a subset of translation divergences. The system is written in Comrnon LISP and runs on a SGI 02 with a RlOOOO processor at 150 MHz. The translation of a 10-15 word sentence is carried in 10-100 ms CPU time, depending on sentence complexity, essentially the number of PPattachments. Performance of the system is not related to the size of the lexicon but rather to the tree/word ratio, which determines the number of forests to be parsed during the analysis of a given sentence (see [Cavazza, 1998] ). This would make possible speech translation in user real-time (i.e., total time < 1 s), considering the time required by the speech recognition and speech synthesis components.This approach has been mainly developed for the translation of constraint:<I languages or application-related sublanguages. We do not claim it tobe appropriate for written language translation, which requires the ability to process much longer sentences and a !arger range of syntactic constructs. Further work would explore the usability of such a system in coUabor~tive multimedia applications. \u2022", | |
| "cite_spans": [ | |
| { | |
| "start": 719, | |
| "end": 734, | |
| "text": "[Cavazza, 1998]", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Preliminary Evaluation", | |
| "sec_num": "3." | |
| }, | |
| { | |
| "text": "i.e\" both in the average length of sentences andin the complexity of syntactic constructs.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "Non cornpositional discontinous constituents in Tree Adjoining Gramrnars", | |
| "authors": [ | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "Abeillts", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Y", | |
| "middle": [], | |
| "last": "Schabes", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1990, | |
| "venue": "Proceedings of the International Symposium on Discontinous Constituency", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "AbeilltS, A. & Schabes, Y., 1990. Non cornpositional discontinous constituents in Tree Adjoining Gramrnars. Proceedings of the International Symposium on Discontinous Constituency, Tilburg.", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Les Nouvel/es Sylltaxes: Grammaires d'unification et analyse dufrancais", | |
| "authors": [ | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "Abeillts", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1993, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "AbeilltS, A\" 1993. Les Nouvel/es Sylltaxes: Grammaires d'unification et analyse dufrancais. Paris: Armand Colin.", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "Using Lexicalized Tags for Machine Translation", | |
| "authors": [ | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "Abeillts", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Y", | |
| "middle": [], | |
| "last": "Schabes", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "A", | |
| "middle": [ | |
| "K" | |
| ], | |
| "last": "Joshi", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1990, | |
| "venue": "Proceedings of COLJNG'90", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "AbeilltS, A\" Schabes, Y. & Joshi, A.K., 1990. Using Lexicalized Tags for Machine Translation. \u2022 Proceedings of COLJNG'90, Helsinki.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "An Integrated Parser for TFG with Explicit Tree Typing", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Cavazza", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1998, | |
| "venue": "Proceedings of the Fourth TAG+ Workshop", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Cavazza, M\" 1998. An Integrated Parser for TFG with Explicit Tree Typing. Proceedings of the Fourth TAG+ Workshop, University of Pennsylvania (this volume).", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "Machine Translation Divergences: A Formal Description and Proposed Solution", | |
| "authors": [ | |
| { | |
| "first": "B", | |
| "middle": [], | |
| "last": "Dorr", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1994, | |
| "venue": "Computational Linguistics", | |
| "volume": "20", | |
| "issue": "4", | |
| "pages": "597--633", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Dorr, B., 1994. Machine Translation Divergences: A Formal Description and Proposed Solution. Computational Linguistics, 20:4, pp. 597-633.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "Cnnstraining lexical selection across languages using TAG", | |
| "authors": [ | |
| { | |
| "first": "D", | |
| "middle": [], | |
| "last": "Egedi", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Palmer", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1994, | |
| "venue": "Proceedings of the Third International TAG Workshop", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Egedi, D. & Palmer, M., 1994. Cnnstraining lexical selection across languages using TAG. Proceedings of the Third International TAG Workshop, Paris. Technical report TALANA-RT-94-01, Universite Paris 7.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "Korean to English Translation Using Synchronous TAGs", | |
| "authors": [ | |
| { | |
| "first": "D", | |
| "middle": [], | |
| "last": "Egedi", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Palmer", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "H", | |
| "middle": [ | |
| "S" | |
| ], | |
| "last": "Park", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "A", | |
| "middle": [ | |
| "K" | |
| ], | |
| "last": "Joshi", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1994, | |
| "venue": "Proceedings of the First Conference of the association for Machine Translation in the Americas", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Egedi, D., Palmer, M., Park, H.S. & Joshi, A.K., 1994. Korean to English Translation Using Synchronous TAGs. Proceedings of the First Conference of the association for Machine Translation in the Americas, Columbia.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "Structural traoslation with synchronous rewriting systerns", | |
| "authors": [ | |
| { | |
| "first": "K", | |
| "middle": [], | |
| "last": "Harbusch", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "P", | |
| "middle": [], | |
| "last": "Poller", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1994, | |
| "venue": "Proceedings of the Third International TAG Workshop", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Harbusch, K. & Poller, P\" 1994. Structural traoslation with synchronous rewriting systerns. Proceedings of the Third International TAG Workshop, Paris. Tecboical report TALANA-RT-94-01, Universite Paris 7.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "Translation by Structural Correspondences", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [ | |
| "M" | |
| ], | |
| "last": "Kaplan", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "K", | |
| "middle": [], | |
| "last": "Netter", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "J", | |
| "middle": [], | |
| "last": "Wedekind", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "Zaeneo", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1989, | |
| "venue": "Proceedings of EACL'89", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Kaplan, R.M., Netter, K, Wedekind, J. & Zaeneo, A., 1989. Translation by Structural Correspondences. Proceedings of EACL'89, Manchester.", | |
| "links": null | |
| }, | |
| "BIBREF9": { | |
| "ref_id": "b9", | |
| "title": "XTAG -A Graphical Workbench for Developing Tree-Adjoining Grammars", | |
| "authors": [ | |
| { | |
| "first": "P", | |
| "middle": [], | |
| "last": "Paroubek", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Y", | |
| "middle": [], | |
| "last": "Schabes", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "Joshi", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1992, | |
| "venue": "Proceedings of ANLP'92", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Paroubek, P., Schabes, Y. & Joshi, A.K\" 1992. XTAG -A Graphical Workbench for Developing Tree- Adjoining Grammars. Proceedings of ANLP'92, Trento.", | |
| "links": null | |
| }, | |
| "BIBREF10": { | |
| "ref_id": "b10", | |
| "title": "Synchronous TAGs and Machine Translation", | |
| "authors": [ | |
| { | |
| "first": "G", | |
| "middle": [], | |
| "last": "Prigent", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1994, | |
| "venue": "Proceedings of the Third Intemational TAG Workshop", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Prigent, G., 1994. Synchronous TAGs and Machine Translation. Proceedings of the Third Intemational TAG Workshop, Paris. Technical report TALANA-RT-94-01, Universite Paris 7.", | |
| "links": null | |
| }, | |
| "BIBREF11": { | |
| "ref_id": "b11", | |
| "title": "Synchronous Tree--Adjoining Grammars", | |
| "authors": [ | |
| { | |
| "first": "S", | |
| "middle": [ | |
| "M" | |
| ], | |
| "last": "Schieber", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Y", | |
| "middle": [], | |
| "last": "Schabes", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1990, | |
| "venue": "Proceedings of COL!NG'90", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Schieber, S.M. & Schabes, Y., 1990. Synchronous Tree--Adjoining Grammars. Proceedings of COL!NG'90, Helsinki.", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": {} | |
| } | |
| } |