A knowledge-engineering approach to the cognitive categorization of lexical meaning.

Abstract

A key challenge in natural language processing is to develop intelligent agents which can retrieve and manage knowledge efficiently as well as simulate human-level reasoning. Undoubtedly, the knowledge base plays a crucial role in such a cognitive architecture. The problem lies in the fact that most approaches to the computational treatment of the meaning of words are restricted to systems of binary lexical relations. The goal of this article is to describe, from the view of linguistics and cognitive science, the theoretical foundation which underlies the construction of the deep semantic representations in FunGramKB, a multipurpose lexico-conceptual knowledge base to be implemented in natural language understanding systems. Thus, the conceptual schemata of thematic frames and meaning postulates may not only provide a full-fledged formalization of lexical semantics in natural language processing but can also facilitate the comprehension of linguistic realizations in artificial intelligence.