Research objectives and content This project concerns both the object-oriented model of databases (which is promised, in the near future, to be referred to as the standard for next generation databases) and the different theories of imprecision. The notion of imprecision can be introduced in a database at two main levels: 1) in the data description (classical approaches allowing only for precise values, the aim is then to extend data management to partially unknown data), 2) in the queries (imprecise or flexible queries which enable database users to express preferences inside elementary conditions and priority between conditions). Some fuzzy object-oriented database models have been defined recently and now work is going on, attempting to define a unifying framework which could encompass these different-models. At the host institution an object-oriented database model, called UFO, has been defined, which allows us to handle fuzziness and uncertainty in data. The mapping of this model onto a relational model has also been described. Training content (objective, benefit and expected impact) The study of appropriate querying techniques for fuzzy and uncertain object-oriented database management systems is the next step; which merely has been addressed by now. This project aims at designing a flexible query language for the fuzzy and uncertain database model UFO (extended if necessary). The applicant being already involved in flexible querying (in the relational model of data), this project gives him the possibility to extend his research and to train himself to object-oriented environment.