When lexical semantics and word sense disambiguation meet lexicography

Abstract Lexical ambiguity, where a word has multiple related (polysemous) and/or unrelated (homographic) meanings, causes challenges in learning and teaching semi-technical medical vocabulary owing to its lack of consideration in current lexicographical resources, including wordlists and dictionaries. Academic wordlists fail to indicate polysemes and homographs, while conventional dictionaries number multiple meanings of a semi-technical medical word in a vertical format that is unlikely to showcase polysemous and homographic relations. This study focuses on a new lexicographical resource, named SemiMed, which addresses issues in wordlists and dictionaries arising from lexical ambiguity. Hsu’s (2013) Medical Word List (MWL), an academic list of semi-technical medical vocabulary, was selected as a starting point. A qualitative analysis underpinned by lexical semantic theories was conducted to analyse polysemes and homographs of MWL words. Then, a quantitative analysis that employed word sense disambiguation practices calculated MWL word meaning frequency. These analyses resulted in two key features, radial networks of meanings and technicality levels, that are expected to help SemiMed resolve issues in academic wordlists and conventional dictionaries caused by lexical ambiguity.