Our hypothesis is that when a user employs an IRS, he has an objective to achieve. This objective concerns his information need. In order to achieve this objective, the user generally does some activities using the IRS. The IRS propose to the user with solutions in response to the queries formulated by the user. The main task of an IRS is to provide the user with solutions that are relevant to his information need. This is termed personalization of information. The main axe of our study is the how to personnalize the system's response according to the user's objective. We propose the use of a user model for personnailizing the system's response. In our approach, the user model defines what to represent for each user. The activities of the user during the use of an IRS is recorded based on the user model. The analysis and synthesis of these activities are used to provide the user with more relevant solutions according to his objective. Three different applications have been developed to validate our approach of personalizing the system?s response and based on an architecture that we defined for a cooperative information retrieval. The three applications are METIORE_STREEMS, METIORE_LORIA and METIORE_REVUES. METIORE_STREEMS is an IRS for managing multimedia information on trees authorized for reforestation by the European Union (EU). The project was sponsored under the EU project LEONARDO. The second application, METIORE_LORIA is used for managing the publications of the computer science laboratory research center, LORIA, Nancy France. The third application METIORE_REVUES is used for the access and analysis of a collection of a journal called Relations Publiques Informations.