Clustering Structured Web Sources: A Schema-Based, Model-Differentiation Approach

Abstract. The Web has been rapidly “deepened” with the prevalence of databases online. On this “deep Web,” numerous sources are structured, providing schema-rich data– Their schemas define the object domain and its query capabilities. This paper proposes clustering sources by their query schemas, which is critical for enabling both source selection and query mediation, by organizing sources of with similar query capabiliabstraction, this problem is essentially clustering categorical data (by viewing each query schema as a transaction). Our approach hypothesizes that “homogeneous sources” are characterized by the same hidden generative models for their schemas. To find clusters governed by such statistical distributions, we propose a novel objective function, model-differentiation, which employs principled hypothesis testing to maximize statistical heterogeneity among clusters. Our evaluation shows that, on clustering the Web query schemas, the model-differentiation func...

• Type: Conference
• Year: 2004
• Where: EDBTW
• Authors: Bin He, Tao Tao, Kevin Chen-Chuan Chang
• Keywords: Data– Their Schemas, EDBTW 2004, Query Schemas, Similar Query Capabiliabstraction, Software Engineering