Introducing Syntactic Structures into Target Opinion Word Extraction with Deep Learning

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 983–994, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 983 Word class flexibility: A deep contextualized approach Bai Li1,4, Guillaume Thomas2, Yang Xu1,3,4, Frank Rudzicz1,4,5 1 Department of Computer Science, University of Toronto, Toronto, Canada 2 Department of Linguistics, University of Toronto, Toronto, Canada 3 Cognitive Science Program, University of Toronto, Toronto, Canada 4 Vector Institute for Artificial Intelligence, Toronto, Canada 5 St. Michael’s Hospital, Toronto, Canada {bai, yangxu, frank}@cs.toronto.edu guillaume.thomas@utoronto.ca Abstract Word class flexibility refers to the phe- nomenon whereby a single word form is used across different grammatical categories. Ex- tensive work in linguistic typology has sought to characterize word class flexibility across languages, but quantifying this phenomenon accurately and at scale has been fraught with difficulties. We propose a principled method- ology to explore regularity in word class flex- ibility. Our method builds on recent work in contextualized word embeddings to quan- tify semantic shift between word classes (e.g., noun-to-verb, verb-to-noun), and we apply this method to 37 languages1. We find that contextualized embeddings not only capture human judgment of class variation within words in English, but also uncover shared ten- dencies in class flexibility across languages. Specifically, we find greater semantic variation when flexible lemmas are used in their domi- nant word class, supporting the view that word class flexibility is a directional process. Our work highlights the utility of deep contextual- ized models in linguistic typology. 1 Introduction In natural languages, lexical items can often be used in multiple word classes without overt changes in word form. For instance, the word buru in Mundari can be used as a noun to denote ‘mountain’, or as a verb to denote ‘to heap up’ (Evans and Osada, 2005). Known as word class flexibility, this phenomenon is considered one of the most challenging topics in linguistic typology (Evans and Levinson, 2009). We present a compu- tational methodology to quantify the regularity in word class flexibility across languages. 1Code and data to reproduce the experimental findings are available at: https://github.com/SPOClab-ca/ word-class-flexibility. There is an extensive literature on how lan- guages vary in word class flexibility, either di- rectly (Hengeveld, 1992; Vogel and Comrie, 2000; Van Lier and Rijkhoff, 2013) or through related notions such as word class conversion (with zero- derivation) (Vonen, 1994; Don, 2003; Bauer and Valera, 2005a; Manova, 2011; S¸tekauer et al., 2012). However, existing studies tend to rely on analyses of small sets of lexical items that may not be representative of word class flexibility in the broad lexicon. Critically lacking are system- atic analyses of word class flexibility across many languages, and existing typological studies have only focused on qualitative comparisons of word class systems. We take to our knowledge the first step towards computational quantification of word class flexi- bility in 37 languages, taken from the Universal Dependencies project (Zeman et al., 2019). We focus on lexical items that can be used both as nouns and as verbs, i.e., noun-verb flexibility. …

Content truncated to 500 words