Web mining model and its applications for information gathering

Web mining is used to automatically discover and extract information from Web-related data sources such as documents, log, services, and user profiles. Although standard data mining methods may be applied for mining on the Web, many specific algorithms need to be developed and applied for various purposes of Web based information processing in multiple Web resources, effectively and efficiently. In the paper, we propose an abstract Web mining model for extracting approximate concepts hidden in user profiles on the semantic Web. The abstract Web mining model represents knowledge on user profiles by using an ontology which consists of both 'part-of' and 'is-a' relations. We also describe the details of using the abstract Web mining model for information gathering. In this application, classes of the ontology are represented as subsets of a list of keywords. An efficient filtering algorithm is also developed to filter out most non-relevant inputs. (C) 2004 Elsevier B.V. All rights reserved.