Research Interests

My research focuses mainly at the following areas:

• Social network. Social media like Facebook, Twitter, LinkedIn, etc, have become a major trend over the Web 2.0 as well as an important communication vector. Their fast and unprecedented success has introduced several challenges for service providers as well as for their users. While the former have to face a tremendous flow of user generated content, the latter struggle to find relevant data that match their interests: they usually have to spend a long time to read all the content received, trying to filter out relevant information. My research work focuses on searching scalable structures/algorithms for filtering information sent/received according to the user interests and also for discovering new relevant content based on the social graph and the user context.
• Multi-dimensional indexing. How to index trajectories of mobile objects or large amount of multimedia documents (audio or video files)? How to scale using a distributed architecture? We work on dynamic indexes that scale for very large amount of multidimensional data and can be deployed on a cluster of machines.
• Text-indexing. Many solutions were proposed to index text. We distinguish in main memory approaches like KMP, Karp-Rabbin or Boyer-Moore, and disk-based approaches like the n-gram index or the String B-tree. We propose new solutions for both kinds of indexing with promising results.
• Data privacy. To test and validate new applications developers need realistic data so final tests are generally performed on excerpts from the on-going production databases. With the recent phenomenon of the externalization of any development and test appear two problems: (i) information in many databases is proprietary and must be protected and (ii) existing proposals lack an automatic detection of the sensitive data. So we work on an automatic detection of the values to be scrambled and an automatic propagation to other semantically linked values.