LESSLEX: Linking multilingual Embeddings to SenSe representations of Lexical items

Abstract

We present LESSLEX, a novel multilingual lexical resource. Different from the vast majority of existing approaches, we ground our embeddings on a sense inventory made available from the BabelNet semantic network. In this setting, multilingual access is governed by the mapping of terms onto their underlying sense descriptions, such that all vectors co-exist in the same semantic space. For each term we have thus the 'blended' terminological vector along with those describing all senses associated to that term. LESSLEX has been tested on the conceptual similarity task: we experimented over the principal data sets for this task in their multilingual and cross-lingual variants, improving on or closely approaching state-of-the-art results. We conclude by arguing that LESSLEX vectors may be relevant for practical applications and for interdisciplinary research on conceptual and lexical access and competence.