Ma main research topics concern data management, data modeling, information systems and information retrieval. Hereafter are some specific topics I have addressed the last 30 years, in collaboration with many of my PhD students and colleagues:

• Data modeling: My first contributions were on data modeling (relational modeling, semantic modeling and object modeling) and design tools for databases. I have proposed a rule-based design process that compiles multimodal specifications (first order formulas, natural language statements, graphical representations) and generate a coherent database schema with its set of integrity constraints, triggering procedures and view definitions.
• Active rules: Then I worked on active rules and their execution model within database systems. The contributions are two fold: formal rule specification and verification, execution model and optimisation issues. This work has been done within the ACTNET european network.
• Datawarehousing: A major topic I addressed  was datawarehouse design with a specific focus on the design of the data refreshment process based on actives rules. This work has been done within the DWQ european project.
• Data quality: This is a multidimensional problem that requires a sound specification of the dimensions, their interrelations and their respective evaluation procedures. The contribution was a metamodel that governs these definitions and serves as a monitoring plateforme for measuring data quality of business processes  (ANR QUADRIS project).
• Query personalisation: Querying massive data results in plethora answers among which only few are of interest to the user. The first contribution is a query reformulation algorithm based on user context and preferences that significantly reduces the size of answers as the experiments have shown on a large size data set (ImDB). The second contribution is the definition of a set of personalisation operators that can be used for web application engineering (ACI APMD project).  
• Recommender systems: This previous work has been extended to recommender systems based on the use of social data. The contributions  are on a specific use of folksonomies and on the definition of specific ranking metrics of the recommandations.
• Complex objects matching: Searching the web is generally a surface search based on keywords. Application to web services or graph documents does not take into account the deep semantics of the target objects. We proposed a new querying technics based on semantic subgraph isomorphism, the semantics being captured by the specific structure of the graphs. This work was done within the ANR AOC project and pursued in the CAIR project to address aggregate search.