Applications of an Web information mining model to data mining and information retrieval tasks

We have developed a model to mine information in applications involving graph analysis. We demonstrate the model characteristics using a Web warehouse, where nodes represent Web pages and edges represent hyperlinks. In this paper we apply the model to data mining and information retrieval tasks. Related to data mining, we present views for clustering Web nodes and for finding frequent itemsets for association rule mining. Related to information retrieval, we present views for performing a simple query, for clustering the query results and an attempt to improve the quality of the ranking. The use of the model with these purposes demonstrates its modularity, flexibility, applicability and broadness in graph problems.