Discriminative Lexical Semantic Segmentation with Gaps: Running the MWE Gamut

Discriminative Lexical Semantic Segmentation with Gaps: Running the MWE Gamut Nathan Schneider Emily Danchik Chris Dyer Noah A. Smith School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213, USA {nschneid,emilydan,cdyer,nasmith}@cs.cmu.edu Abstract We present a novel representation, evaluation measure, and supervised models for the task of identifying the multiword expressions (MWEs) in a sentence, resulting in a lexical seman- tic segmentation. Our approach generalizes a standard chunking representation to encode MWEs containing gaps, thereby enabling effi- cient sequence tagging algorithms for feature- rich discriminative models. Experiments on a new dataset of English web text offer the first linguistically-driven evaluation of MWE iden- tification with truly heterogeneous expression types. Our statistical sequence model greatly outperforms a lookup-based segmentation pro- cedure, achieving nearly 60% F1 for MWE identification. 1 Introduction Language has a knack for defying expectations when put under the microscope. For example, there is the notion—sometimes referred to as compositionality— that words will behave in predictable ways, with indi- vidual meanings that combine to form complex mean- ings according to general grammatical principles. Yet language is awash with examples to the contrary: in particular, idiomatic expressions such as awash with NP, have a knack for VP-ing, to the contrary, and defy expectations. Thanks to processes like metaphor and grammaticalization, these are (to various degrees) semantically opaque, structurally fossilized, and/or statistically idiosyncratic. In other words, idiomatic expressions may be exceptional in form, function, or distribution. They are so diverse, so unruly, so 1. MW named entities: Prime Minister Tony Blair 2. MW compounds: hot air balloon, skinny dip 3. conventionally SW compounds: somewhere 4. verb-particle: pick up, dry out, take over, cut short 5. verb-preposition: refer to, depend on, look for 6. verb-noun(-preposition): pay attention (to) 7. support verb: make decisions, take pictures 8. other phrasal verb: put up with, get rid of 9. PP modifier: above board, at all, from time to time 10. coordinated phrase: cut and dry, more or less 11. connective: as well as, let alone, in spite of 12. semi-fixed VP: pick up where <one> left off 13. fixed phrase: scared to death, leave of absence 14. phatic: You’re welcome. Me neither! 15. proverb: Beggars can’t be choosers. Figure 1: Some of the classes of idioms in English. The examples included here contain multiple lexicalized words—with the exception of those in (3), if the conven- tional single-word (SW) spelling is used. difficult to circumscribe, that entire theories of syn- tax are predicated on the notion that constructions with idiosyncratic form-meaning mappings (Fillmore et al., 1988; Goldberg, 1995) or statistical properties (Goldberg, 2006) offer crucial evidence about the grammatical organization of language. Here we focus on multiword expressions (MWEs): lexicalized combinations of two or more words that are exceptional enough to be considered as single units in the lexicon. As figure 1 illus- trates, MWEs occupy diverse syntactic and semantic functions. Within MWEs, we distinguish (a) proper names and (b) lexical idioms. The latter have proved themselves a “pain in the neck for NLP” (Sag et al., …

Content truncated to 500 words