Exploring evidence for shallow parsing

Exploring Evidence for Shallow Parsing � Xin Li Dan Roth Department of Computer Science University of Illinois at Urbana-Champaign Urbana, IL 61801 xli1@cs.uiuc.edu danr@cs.uiuc.edu Abstract Significant amount of work has been devoted recently to develop learning techniques that can be used to gener- ate partial (shallow) analysis of natu- ral language sentences rather than a full parse. In this work we set out to evalu- ate whether this direction is worthwhile by comparing a learned shallow parser to one of the best learned full parsers on tasks both can perform — identify- ing phrases in sentences. We conclude that directly learning to perform these tasks as shallow parsers do is advanta- geous over full parsers both in terms of performance and robustness to new and lower quality texts. 1 Introduction Shallow parsing is studied as an alternative to full-sentence parsing. Rather than producing a complete analysis of sentences, the alternative is to perform only partial analysis of the syn- tactic structures in a text (Harris, 1957; Abney, 1991; Greffenstette, 1993). A lot of recent work on shallow parsing has been influenced by Ab- ney’s work (Abney, 1991), who has suggested to “chunk” sentences to base level phrases. For ex- ample, the sentence “He reckons the current ac- count deficit will narrow to only $ 1.8 billion in September .” would be chunked as follows (Tjong Kim Sang and Buchholz, 2000): [NP He ] [VP reckons ] [NP the current account deficit ] [VP will narrow ] [PP �This research is supported by NSF grants IIS-9801638, ITR-IIS-0085836 and an ONR MURI Award. to ] [NP only $ 1.8 billion ] [PP in ] [NP September] . While earlier work in this direction concen- trated on manual construction of rules, most of the recent work has been motivated by the obser- vation that shallow syntactic information can be extracted using local information — by examin- ing the pattern itself, its nearby context and the local part-of-speech information. Thus, over the past few years, along with advances in the use of learning and statistical methods for acquisition of full parsers (Collins, 1997; Charniak, 1997a; Charniak, 1997b; Ratnaparkhi, 1997), significant progress has been made on the use of statisti- cal learning methods to recognize shallow pars- ing patterns — syntactic phrases or words that participate in a syntactic relationship (Church, 1988; Ramshaw and Marcus, 1995; Argamon et al., 1998; Cardie and Pierce, 1998; Munoz et al., 1999; Punyakanok and Roth, 2001; Buchholz et al., 1999; Tjong Kim Sang and Buchholz, 2000). Research on shallow parsing was inspired by psycholinguistics arguments (Gee and Grosjean, 1983) that suggest that in many scenarios (e.g., conversational) full parsing is not a realistic strat- egy for sentence processing and analysis, and was further motivated by several arguments from a natural language engineering viewpoint. First, it has been noted that in many natural lan- guage applications it is sufficient to use shallow parsing information; information such as noun phrases (NPs) and other syntactic sequences have been found useful in many …

Content truncated to 500 words