GlossBoot: Bootstrapping Multilingual Domain Glossaries from the Web

Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pages 528–538, Sofia, Bulgaria, August 4-9 2013. c⃝2013 Association for Computational Linguistics GlossBoot: Bootstrapping Multilingual Domain Glossaries from the Web Flavio De Benedictis, Stefano Faralli and Roberto Navigli Dipartimento di Informatica Sapienza Universit`a di Roma flavio.debene@gmail.com,{faralli,navigli}@di.uniroma1.it Abstract We present GlossBoot, an effective minimally-supervised approach to ac- quiring wide-coverage domain glossaries for many languages. For each language of interest, given a small number of hypernymy relation seeds concerning a target domain, we bootstrap a glossary from the Web for that domain by means of iteratively acquired term/gloss extraction patterns. Our experiments show high performance in the acquisition of domain terminologies and glossaries for three different languages. 1 Introduction Much textual content, such as that available on the Web, contains a great deal of information fo- cused on specific areas of knowledge. However, it is not infrequent that, when reading a domain- specific text, we humans do not know the mean- ing of one or more terms. To help the human understanding of specialized texts, repositories of textual definitions for technical terms, called glos- saries, are compiled as reference resources within each domain of interest. Interestingly, electronic glossaries have been shown to be key resources not only for humans, but also in Natural Language Processing (NLP) tasks such as Question Answer- ing (Cui et al., 2007), Word Sense Disambiguation (Duan and Yates, 2010; Faralli and Navigli, 2012) and ontology learning (Navigli et al., 2011; Ve- lardi et al., 2013). Today large numbers of glossaries are available on the Web. However most such glossaries are small-scale, being made up of just some hundreds of definitions. Consequently, individual glossaries typically provide a partial view of a given domain. Moreover, there is no easy way of retrieving the subset of Web glossaries which appertains to a do- main of interest. Although online services such as Google Define allow the user to retrieve defi- nitions for an input term, such definitions are ex- tracted from Web glossaries and put together for the given term regardless of their domain. As a re- sult, gathering a large-scale, full-fledged domain glossary is not a speedy operation. Collaborative efforts are currently producing large-scale encyclopedias, such as Wikipedia, which are proving very useful in NLP (Hovy et al., 2013). Interestingly, wikipedias also include man- ually compiled glossaries. However, such glos- saries still suffer from the same above-mentioned problems, i.e., being incomplete or over-specific,1 and hard to customize according to a user’s needs. To automatically obtain large domain glos- saries, over recent years computational ap- proaches have been developed which extract tex- tual definitions from corpora (Navigli and Velardi, 2010; Reiplinger et al., 2012) or the Web (Fujii and Ishikawa, 2000). The former methods start from a given set of terms (possibly automatically extracted from a domain corpus) and then har- vest textual definitions for these terms from the input corpus using a supervised system. Web- based methods, instead, extract text snippets from Web pages which match pre-defined lexical pat- …

Content truncated to 500 words