Paper: Semantic distance and terminology structuring methods for the detection of semantically close terms

ACL ID W12-2403
Title Semantic distance and terminology structuring methods for the detection of semantically close terms
Venue Workshop on Biomedical Natural Language Processing
Session
Year 2012
Authors

The identification of semantically similar lin- guistic expressions despite their formal differ- ence is an important task within NLP appli- cations (information retrieval and extraction, terminology structuring...) We propose to de- tect the semantic relatedness between biomed- ical terms from the pharmacovigilance area. Two approaches are exploited: semantic dis- tance within structured resources and termi- nology structuring methods applied to a raw list of terms. We compare these methods and study their complementarity. The results are evaluated against the reference pharmacovigi- lance data and manually by an expert.