XML search personalization strategies using query expansion, reranking and a search engine modification.

As the amount of information increases every day and queries are most of the times short and ambiguous, search personalization techniques are becoming almost a must. These techniques retrieve relevant results closer to the user, covering in a better and easier way his/her information needs. This work is focused on improving the personalized retrieval process, trying to combine the advantages of all the approaches used for personalization: query reformulation, reranking of results and search engine modification. The proposed personalization techniques have been designed for XML retrieval, but they could easily be applied to non structured documents. The experimental results obtained from a user study using a parliamentary document collection support the validity of our approach.