Ranking Query Results in Cooperative Query Answering Based on the Knowledge Abstraction

Abstract

Cooperative query answering systems provide a user with approximate answers when processing a query. They use a knowledge representation framework to facilitate relevant information search. In this paper, we propose a metricized knowledge abstraction hierarchy (MKAH) which supports multi-level knowledge representation and quantitative distance measure among data values. We propose the method that calculates distances between arbitrary two nodes by incorporating the basic distances, and show the effectiveness of the hierarchy structure and distance measure. A prototype job search system that uses the MKAH has been implemented. We show through various experiments, that the MKAH provides rich semantic representation and high quality distance measure. The MKAH is appropriate for building up a large scaled system by integrating other systems