Extraction of Semantic Clusters for Terminological Information Retrieval from MRDs
Gerardo Sierra, John McNaught · 2000
This paper describes a semantic clustering method for data extracted from machine readable dictionaries (MRDs) in order to build a terminological information retrieval system that finds terms from descriptions of concepts. We first examine approaches based on ontologies and statistics, before introducing our analogy-based approach that lets us extract semantic clusters by aligning definitions from two dictionaries. Evaluation of the final set of clusters for a small set of definitions demonstrates the utility of our approach. 1. Background The majority of lexicographers recognise the need for dictionaries that, contrary to the alphabetical ordering of entries, help users to look for a word that has escaped their memory even though they remember the concept. From a semantic point of view, Baldinger (1980) identifies two kinds of dictionaries. The semasiological one corresponds to the viewpoint of the person interpreting the speaker, thus one starts with the form of the expression to look for the meaning. It is