Identification and Resolution of Chinese Zero Pronouns: A Machine Learning Approach

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 541–550, Prague, June 2007. c⃝2007 Association for Computational Linguistics Identification and Resolution of Chinese Zero Pronouns: A Machine Learning Approach Shanheng Zhao and Hwee Tou Ng Department of Computer Science National University of Singapore 3 Science Drive 2 Singapore 117543 {zhaosh, nght}@comp.nus.edu.sg Abstract In this paper, we present a machine learn- ing approach to the identification and reso- lution of Chinese anaphoric zero pronouns. We perform both identification and resolu- tion automatically, with two sets of easily computable features. Experimental results show that our proposed learning approach achieves anaphoric zero pronoun resolution accuracy comparable to a previous state-of- the-art, heuristic rule-based approach. To our knowledge, our work is the first to per- form both identification and resolution of Chinese anaphoric zero pronouns using a machine learning approach. 1 Introduction Coreference resolution is the task of determining whether two or more noun phrases refer to the same entity in a text. It is an important task in discourse analysis, and successful coreference resolution ben- efits many natural language processing applications such as information extraction, question answering, etc. In the literature, much of the work on corefer- ence resolution is for English text (Soon et al., 2001; Ng and Cardie, 2002b; Yang et al., 2003; McCal- lum and Wellner, 2005). Publicly available cor- pora for coreference resolution are mostly in En- glish, e.g., the Message Understanding Conference tasks (MUC6 and MUC7)1. Relatively less work has 1http://www-nlpir.nist.gov/related_ projects/muc/ been done on coreference resolution for Chinese. Recently, the ACE Entity Detection and Tracking (EDT) task2 included annotated Chinese corpora for coreference resolution. Florian et al. (2004) and Zhou et al. (2005) reported research on Chinese coreference resolution. A prominent phenomenon in Chinese coreference resolution is the prevalence of zero pronouns. A zero pronoun (ZP) is a gap in a sentence which refers to an entity that supplies the necessary information for interpreting the gap. An anaphoric zero pro- noun (AZP) is a zero pronoun that corefers to one or more overt noun phrases present in the preced- ing text. Zero pronouns occur much more frequently in Chinese compared to English, and pose a unique challenge in coreference resolution for Chinese. For example, Kim (2000) conducted a study to compare the use of overt subjects in English, Chinese, and other languages. He found that the use of overt sub- jects in English is over 96%, while this percentage is only 64% for Chinese, indicating that zero pronouns (lack of overt subjects) are much more prevalent in Chinese. Chinese zero pronouns have been studied in lin- guistics research (Li and Thompson, 1979; Li, 2004), but only a small body of prior work in com- putational linguistics deals with Chinese zero pro- noun identification and resolution (Yeh and Chen, 2004; Converse, 2006). To our knowledge, all pre- vious research on zero pronoun identification and resolution in Chinese uses hand-engineered rules or heuristics, and our present work is the …

Content truncated to 500 words