Enhancing query processing of information systems

Abstract

Current database and information systems only provide for inflexible query processing. In this paper we have developed techniques for supporting the answering of inadequate queries to information systems introducing certain level of intelligence and flexibility. Our approach uses the notion of α-cut in fuzzy set theory to provide acceptable approximate answers for a given inadequate query. We illustrate the proposed α-cut approach for a variety of different queries which may involve different types of selection conditions. The selection conditions considered may contain fuzzy terms measured on a numeric, nominal or linguistic scale. They may employ one or more linguistic hedges of (very), or of (fairly) for query modification. Fuzzy numbers and fuzzy intervals are introduced as representations of selection conditions of inadequate queries. It is not essential to construct extensive metadata into the existing database to implement the proposed approach. A user friendly system which makes use of the proposed theory has been developed. It can support selection conditions that contain linguistic hedges of (very) and (fairly), comparators (approximately), (more-than) and (less-than), and terms that are numeric, nominal or linguistic. The test results bear positive evidence to the ease of implementation of the proposed approach.