Multi-schema Matching Based on Web Structured Information Sources

Schema matching is widely used in many database applications, such as, data integration, data warehouse, data spaces, and ontology merging. In this paper, we propose multi-schema matching based on web structured information sources. There are two meanings at this point. Traditional matching techniques mainly address matching tasks between two attributes, namely pair wise-attribute correspondence. Thus, the first is that we will focus on find the semantic correspondence among multiple attributes, which is more difficult than pair wise-attribute correspondence. The main idea is to regard each attribute as a point in vector space and partition attributes into different set by clustering techniques. The attributes in the same cluster have the similar semantic. Second, we will use web sources that contain ample structured information to improve the quality of schema matching. We validate our approach with an experimental study, the results of which demonstrate that our approach is effective and has good performance.