Constructing a poor man's wordnet in a resource-rich world

Darja Fišer, Benoît Sagot
2015 Language Resources and Evaluation  
In this paper we present a language-independent, fully modular and automatic approach to bootstrap a wordnet for a new language by recycling different types of already existing language resources, such as machine-readable dictionaries, parallel corpora, and Wikipedia. The approach, which we apply here to Slovene, takes into account monosemous and polysemous words, general and specialised vocabulary as well as simple and multi-word lexemes. The extracted words are then assigned one or several
more » ... set ids, based on a classifier that relies on several features including distributional similarity. Finally, we identify and remove highly dubious (literal, synset) pairs, based on simple distributional information extracted from a large corpus in an unsupervised way. Automatic, manual and task-based evaluations show that the resulting resource, the latest version of the Slovene wordnet, is already a valuable source of lexico-semantic information. Keywords Wordnet development Á Multilingual lexicon extraction Á Word-sense disambiguation Á Distributional similarity Introduction Motivation. In the past decade, the role of lexical knowledge has increased in many areas of natural language processing as it has been shown that exploiting lexical
doi:10.1007/s10579-015-9295-6 fatcat:som24c2hn5bhzpgylpdbs2rrrq