The development of lexical-semantic resources for natural language processing is urgent. Conceiving the computational systems which process natural languages as an exercise of "human language engineering", the development of such resources can benefit from the "knowledge representation models" provided by Knowledge Engineering. These models provide both the theoretical-methodological premises and the formal metalanguage for the computational treatment of word meaning. In this paper, we first present the linguistic-computational conception of "lexicon". Then, we describe the main knowledge representation paradigms emphasizing the theoretical approach to word meaning and the formalism related to each of them.
Lexical Semantics; Computational-Linguistic resource; knowledge representation; natural language processing