Roget's Thesaurus: a Lexical Resource to Treasure
This work addresses lexical resource development for computational linguistics, but it is incremental as it adapts an existing thesaurus with established tools.
The paper tackled the creation of an electronic lexical knowledge base from Roget's Thesaurus by labeling semantic relations with WordNet, resulting in a qualitative and quantitative comparison and discussion of merging possibilities.
This paper presents the steps involved in creating an electronic lexical knowledge base from the 1987 Penguin edition of Roget's Thesaurus. Semantic relations are labelled with the help of WordNet. The two resources are compared in a qualitative and quantitative manner. Differences in the organization of the lexical material are discussed, as well as the possibility of merging both resources.