| WordNet is a lexical reference system that is widely used in the Natural Language (NL) community. The reason for its popularity is its broad and uniform coverage of general concepts that are fundamental to any NL System requiring universal knowledge. WordNet contains concepts, relations between concepts and descriptive glosses that define concepts. The descriptive glosses in WordNet contain a vast amount of semantic knowledge that currently remains undiscovered.;The aim of my work, presented in this thesis, is to extract the semantic information contained within WordNet glosses to transform WordNet into a comprehensive knowledge base. This extraction of semantic knowledge is done via semantic relations. This extracted semantic information is particularly beneficial in deriving axioms pertaining to WordNet concepts. This thesis provides an in-depth look at the process of semantic annotations and presents some of the applications of the semantically enhanced WordNet knowledge base. |
