Lexifield: a system for the automatic building of lexicons by semantic expansion of short word lists

摘要

We present Lexifield, a fully automatic language-independent system for building domain-specific lexicons from a short list of terms defining the domain. Lexifield relies on a pre-trained word embedding model, a definition dictionary and a dictionary of synonyms. To evaluate this system, four lexicons have been generated: one lexicon in French for the topic "son" ("sound") and three lexicons in English for the topics "sound", "taste" and "odour". As compared to other word embedding-based systems and a state-of-the-art sensorial lexicon, Sensicon, our system achieves better precision and recall on reference lists extracted from manually created resources such as Roget's Thesaurus.