Abstract

In this report, we address the issue of automatization of bookmark management by means of information retrieval and machine learning techniques. We have designed a system that categorizes bookmarks in an automated way and thus facilitates adaptively their management. Organization of bookmarks is based on hierarchical categories, which are mapped to the structure of the Open Directory Project. The user’s hierarchy is built dynamically according to additions and removals of bookmarks. Furthermore, the system is designed to recommend related Web pages to the recent bookmarks. This service is based on similarities observed during the categorization process. The designed system has been fully implemented and named CaterBook.