Personalizing XML Search in PIMENTO

XML search is increasing in popularity as more and larger XML repositories are becoming available. The accuracy of XML search varies across different systems and a lot of effort is put into designing scoring functions tailored to specific users and datasets. We argue that there is no one scoring function that fits all and advocate incorporating user profiles into XML search to personalize query answers by accounting for user profiles. First, we propose a framework for defining user profiles and for enforcing them during query processing. Second, we adapt the well-known top-k pruning to account for user profiles. Finally, we present effectiveness and efficiency experiments which show that query personalization in XML search dramatically improves the accuracy of query results while incurring negligible processing overhead. This work is in the context of the Pimento project which aims at improving the relevance of searching structured and unstructured content.