Extracting meaning by idiomaticity: Description of the HSemID system at CogALex VI (2020)
(2020) COLING 2020 - Workshop on the Cognitive Aspects of the Lexicon — Location: Barcelona (8.December.2020)
Files
HSemIDSystemFinal2.pdf
Open Access - Adobe PDF
- 215.8 KB
Details
- Authors
- Abstract
- The HSemID system, submitted to the CogALex VI Shared Task is a hybrid system relying mainly on metric clusters measured in large web corpora, complemented by a vector space model using cosine similarity to detect semantic associations. Although the system reached ra-ther weak results for the subcategories of synonyms, antonyms and hypernyms, with some dif-ferences from one language to another, it is able to measure general semantic associations (as being random or not-random) with an F1 score close to 0.80. The results strongly suggest that idiomatic constructions play a fundamental role in semantic associations. Further experiments are necessary in order to fine-tune the model to the subcategories of synonyms, antonyms, hy-pernyms and to explain surprising differences across languages.
- Affiliations
Citations
Colson, J.-P. (2020). Extracting meaning by idiomaticity: Description of the HSemID system at CogALex VI (2020). COLING 2020 - Workshop on the Cognitive Aspects of the Lexicon, Barcelona. https://hdl.handle.net/2078.5/122212