Probabilistic Coordination Disambiguation in a Fully-Lexicalized Japanese Parser

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 306–314, Prague, June 2007. c⃝2007 Association for Computational Linguistics Probabilistic Coordination Disambiguation in a Fully-lexicalized Japanese Parser Daisuke Kawahara National Institute of Information and Communications Technology, 3-5 Hikaridai Seika-cho, Soraku-gun, Kyoto, 619-0289, Japan dk@nict.go.jp Sadao Kurohashi Graduate School of Informatics, Kyoto University, Yoshida-Honmachi, Sakyo-ku, Kyoto, 606-8501, Japan kuro@i.kyoto-u.ac.jp Abstract This paper describes a probabilistic model for coordination disambiguation integrated into syntactic and case structure analy- sis. Our model probabilistically assesses the parallelism of a candidate coordinate structure using syntactic/semantic similari- ties and cooccurrence statistics. We inte- grate these probabilities into the framework of fully-lexicalized parsing based on large- scale case frames. This approach simulta- neously addresses two tasks of coordination disambiguation: the detection of coordinate conjunctions and the scope disambiguation of coordinate structures. Experimental re- sults on web sentences indicate the effective- ness of our approach. 1 Introduction Coordinate structures are a potential source of syn- tactic ambiguity in natural language. Since their in- terpretation directly affects the meaning of the text, their disambiguation is important for natural lan- guage understanding. Coordination disambiguation consists of the fol- lowing two tasks: • the detection of coordinate conjunctions, • and finding the scope of coordinate structures. In English, for example, coordinate structures are triggered by coordinate conjunctions, such as and and or. In a coordinate structure that consists of more than two conjuncts, commas, which have var- ious usages, also function like coordinate conjunc- tions. Recognizing true coordinate conjunctions from such possible coordinate conjunctions is a task of coordination disambiguation (Kurohashi, 1995). The other is the task of identifying the range of co- ordinate phrases or clauses. Previous work on coordination disambiguation has focused on the task of addressing the scope am- biguity (e.g., (Agarwal and Boggess, 1992; Gold- berg, 1999; Resnik, 1999; Chantree et al., 2005)). Kurohashi and Nagao proposed a similarity-based method to resolve both of the two tasks for Japanese (Kurohashi and Nagao, 1994). Their method, how- ever, heuristically detects coordinate conjunctions by considering only similarities between possible conjuncts, and thus cannot disambiguate the follow- ing cases1: (1) a. kanojo-to she-cmi gakkou-ni school-acc itta went (φ went to school with her) b. kanojo-to she-cnj watashi-ga I-nom goukaku-shita passed an exam (she and I passed an exam) In sentence (1a), postposition “to” is used as a comi- tative case marker, but in sentence (1b), postposition “to” is used as a coordinate conjunction. To resolve this ambiguity, predicative case frames are required. Case frames describe what kinds of 1In this paper, we use the following abbreviations: nom (nominative), acc (accusative), abl (ablative), cmi (comi- tative), cnj (conjunction) and TM (topic marker). 306 Table 1: Case frame examples (Examples are writ- ten in English. Numbers following each example represent its frequency.). CS Examples ga I:18, person:15, craftsman:10, · · · yaku (1) wo bread:2484, meat:1521, cake:1283, · · · (broil) de oven:1630, frying pan:1311, · · · yaku (2) ga teacher:3, government:3, person:3, …

Content truncated to 500 words