The meaning of a word is important but for natural language Processing (NLP) systems the representation of the meaning is also an important consideration. We attempt to represent word senses by picking unambiguous labels for each sense of a word and p...
The meaning of a word is important but for natural language Processing (NLP) systems the representation of the meaning is also an important consideration. We attempt to represent word senses by picking unambiguous labels for each sense of a word and placing these labels, and their corresponding words, into a categorization of conceptual space. In particular, we use Rogets International Thesaurus (RIT) as controlled vocabulary to map English verbs to appropriate places in the thesaurus. This paper introduces the issues and suggests several algorithms with some discussions on future research for improvement.