Ontology Enhancement for Including Newly Acquired Knowledge About Concept Descirptions and Answering Imprecise Queries

Abstract

This chapter presents a text-mining-based ontology enhancement and query-processing system. The key ideas introduced here are that of learning and including imprecise concept descriptions into ontology structures. This is essential for ontology-based text information extraction since it is not necessary that text description of the concepts or user-specified descriptions will exactly match stored concept descriptions. The traditional property-value framework for concept description has been extended to a property-value-qualifier framework for this purpose. The system also supports ontology enhancement by identifying, defining, and adding new precise and imprecise concept descriptions mined from text documents. The acquired knowledge is stored in a structured knowledge base for answering user queries. Since user queries may contain concept descriptions, which do not exactly match stored or known concepts, the query processor uses fuzzy reasoning for query processing. Each answer is accompanied by a confidence value that reflects its similarity to the original query concept.