A large-itemset-based index structure for supporting personalized information filtering on the Internet

The World Wide Web creates many new challenges for information retrieval. Information Filtering (IF) can find good matches between the web pages and the users' information needs. In an information filtering system, users are, associated with profiles that describe what they need, while data are represented in the same form of user profiles. Comparing data with profiles, the users who are interested in the data are identified and informed. Therefore, a critical issue of the information filtering service is how to index the user profiles for an efficient matching process. In this paper, first, we propose a count-based tree method, to reduce the large storage space as needed by Yan and Garcia-Molina's tree method. Next, by applying the technique for mining association rules, we propose a large-item set-based method, the count-major large iternset, to further reduce the storage space. From our simulation results, the cost of storage space of our methods is less than that of the tree method.