Ensemble-based Semantic Lexicon Induction for Semantic Tagging

First Joint Conference on Lexical and Computational Semantics (*SEM), pages 199–208, Montr´eal, Canada, June 7-8, 2012. c⃝2012 Association for Computational Linguistics Ensemble-based Semantic Lexicon Induction for Semantic Tagging Ashequl Qadir University of Utah School of Computing Salt Lake City, UT 84112, USA asheq@cs.utah.edu Ellen Riloff University of Utah School of Computing Salt Lake City, UT 84112, USA riloff@cs.utah.edu Abstract We present an ensemble-based framework for semantic lexicon induction that incorpo- rates three diverse approaches for semantic class identification. Our architecture brings together previous bootstrapping methods for pattern-based semantic lexicon induction and contextual semantic tagging, and incorpo- rates a novel approach for inducing semantic classes from coreference chains. The three methods are embedded in a bootstrapping ar- chitecture where they produce independent hypotheses, consensus words are added to the lexicon, and the process repeats. Our results show that the ensemble outperforms individ- ual methods in terms of both lexicon quality and instance-based semantic tagging. 1 Introduction One of the most fundamental aspects of meaning is the association between words and semantic cate- gories, which allows us to understand that a “cow” is an animal and a “house” is a structure. We will use the term semantic lexicon to refer to a dictionary that associates words with semantic classes. Se- mantic dictionaries are useful for many NLP tasks, as evidenced by the widespread use of WordNet (Miller, 1990). However, off-the-shelf resources are not always sufficient for specialized domains, such as medicine, chemistry, or microelectronics. Fur- thermore, in virtually every domain, texts contain lexical variations that are often missing from dic- tionaries, such as acronyms, abbreviations, spelling variants, informal shorthand terms (e.g., “abx” for “antibiotics”), and composite terms (e.g., “may- december” or “virus/worm”). To address this prob- lem, techniques have been developed to automate the construction of semantic lexicons from text cor- pora using bootstrapping methods (Riloff and Shep- herd, 1997; Roark and Charniak, 1998; Phillips and Riloff, 2002; Thelen and Riloff, 2002; Ng, 2007; McIntosh and Curran, 2009; McIntosh, 2010), but accuracy is still far from perfect. Our research explores the use of ensemble meth- ods to improve the accuracy of semantic lexicon in- duction. Our observation is that semantic class as- sociations can be learned using several fundamen- tally different types of corpus analysis. Bootstrap- ping methods for semantic lexicon induction (e.g., (Riloff and Jones, 1999; Thelen and Riloff, 2002; McIntosh and Curran, 2009)) collect corpus-wide statistics for individual words based on shared con- textual patterns. In contrast, classifiers for semantic tagging (e.g., (Collins and Singer, 1999; Niu et al., 2003; Huang and Riloff, 2010)) label word instances and focus on the local context surrounding each in- stance. The difference between these approaches is that semantic taggers make decisions based on a sin- gle context and can assign different labels to differ- ent instances, whereas lexicon induction algorithms compile corpus statistics from multiple instances of a word and typically assign each word to a single semantic category.1 We also hypothesize that coref- erence resolution can be exploited to infer semantic 1This …

Content truncated to 500 words