Reuse-Oriented Mapping Discovery for Meta-querier Customization.

With the tremendous growth in (semi-)structured information, in particular, the databases behind the deep Web, customized integration of data sources is highly desirable to accommodate various user needs. To build and maintain a potentially large number of customized integration systems (called meta-queriers), the first important step is to discover the mappings among their query forms for enabling interoperability between themeta-queriers and user-selected data sources. This paper proposes a reuse-oriented solution to mapping discovery by exploiting existing mappings. For facilitating mapping reuse, ontology-based and changedoriented models are introduced to abstract mappings respectively from the mapping peers (the mappings in the same domain) and the mapping evolution (the mappings from the previous versions). A human-friendly validation strategy is proposed in the pursuit of wide and active participation of non-technical volunteers. Our experimental results on real-world data sets confirm the feasibility and effectiveness of this solution.