Automatic Enrichment of WordNet with Common-Sense Knowledge

Automatic Enrichment of WordNet with Common-Sense Knowledge Luigi Di Caro, Guido Boella University of Turin Corso Svizzera 185 - Turin - Italy dicaro@di.unito.it, boella@di.unito.it Abstract WordNet represents a cornerstone in the Computational Linguistics field, linking words to meanings (or senses) through a taxonomical representation of synsets, i.e., clusters of words with an equivalent meaning in a specific context often described by few definitions (or glosses) and examples. Most of the approaches to the Word Sense Disambiguation task fully rely on these short texts as a source of contextual information to match with the input text to disambiguate. This paper presents the first attempt to enrich synsets data with common-sense definitions, automatically retrieved from ConceptNet 5, and disambiguated accordingly to WordNet. The aim was to exploit the shared- and immediate-thinking nature of common-sense knowledge to extend the short but incredibly useful contextual information of the synsets. A manual evaluation on a subset of the entire result (which counts a total of almost 600K synset enrichments) shows a very high precision with an estimated good recall. Keywords: Semantic Resources, Semantic Enrichment, Common-Sense Knowledge 1. Introduction In the last 20 years, the Artificial Intelligence (AI) commu- nity working on Computational Linguistics (CL) has been using one knowledge base among all, i.e., WordNet (Miller, 1995). In few words, WordNet was a first answer to the most important question in this area, which is the treatment of language ambiguity. Generally speaking, a word is a symbolic expression that may refer to multiple meanings (polysemy), while distinct words may share the same meaning. Syntax reflects gram- mar rules which add complexity to the overall communi- cation medium, making CL one of the most challenging research area in the AI field. From a more detailed perspective, WordNet organizes words in synsets, i.e., sets of words sharing a unique mean- ing in specific contexts (synonyms), further described by descriptions (glosses) and examples. Synsets are then struc- tured in a taxonomy which incorporates the semantics of generality/specificity of the referenced concepts. Although extensively adopted, the limits of this resource are some- times critical: 1) the top-down and general-purpose nature at the basis of its construction lets asking about the actual need of some underused meanings, and 2) most Word Sense Disambiguation approaches use WordNet glosses to under- stand the link between an input word (and its context) and the candidate synsets. In recent years, natural language understanding traced the line of novel and challenging research directions which have been unveiled under the name of textual entailment and question answering. The former is a form of inference based on a lexical basis (to be not intended as in Formal Semantics) while the latter considers the last-mile goal of every computational linguists’ dream: asking questions to a computer in natural language as with humans. As a matter of fact, these tasks require an incredibly rich semantic knowledge containing facts related to behavioural rather than conceptual information, such as what an object may or may not do or what may happen with …

Content truncated to 500 words