MedLinker: Medical Entity Recognition and Linking combining Neural Language Models with Approximate Dictionary Matching
Adapting and improving LMMS for Entity Linking in the Medical Domain.
LIAAD at SemDeep-5 Challenge: Word-in-Context (WiC)
Solution for the WiC challenge based on LMMS (ranked 2nd place at end of challenge).
Language Modelling Makes Sense: Propagating Representations through WordNet for Full-Coverage Word Sense Disambiguation
Word Sense Disambiguation (WSD) based on BERT and sense embeddings with full coverage of WordNet.