Extraction of Implicit Knowledge from WordNet
Wim Peters · 2003
Lexical knowledge databases such as WordNet contain much semantic information that is left implicit. In order to make maximal use of these resources it is important to make this implicit semantic information explicit. Metonymy and regular polysemy constitute a type of implicit ontological knowledge. This paper describes the semi-automatic extraction of systematically related word senses from WordNet by exploiting its hierarchical structure, and the identification of relations that link these on the basis of the glosses. 1.