Dynamic Personalization for Meta-queriers

There exist an increasing number of online databases searchable only through query interfaces. Meta-queriers with an integrated query interface are a practical solution to effective searching of information from multiple data sources. The diversity of user demands and their frequent changes necessitate the need for dynamic and personalized meta-queriers. This paper presents a mapping management system, called M-Repository, to facilitate the implementation of meta-queriers with two innovative features: personalized source selection and dynamic reconfiguration. To support such two features, M-Repository is designed to contain two unique functionalities: 1) reuse-oriented mapping discovery, 2) semantic-based mapping organization. This paper describes the models and implementation structure of M-Repository. Experiments are conducted on real-world query forms, and the results show that the approach is feasible and effective.