The main objective of an information retrieval system (IRS) is to provide relevant information in response to the user's query. On the one part, the relevance of a response concerns its exactness compared with the user's query. On the other part, it concerns its correspondence with the user's knowledge level and his preferences. One of the major contributions in this area of personalization of the system's response is by taking into consideration each user's specificity. We propose the use of explicit user model where the system's solution will be determined by the knowledge on the user. The user's activities are recorded as documents. The method we adopt for information retrieval combines query by criteria and information analysis. We have also proposed architecture for cooperative information retrieval. This architecture allows two users to share their experience in the process of information retrieval and for the interpretation of the system's result, on distant machines. The proposals were implemented in two systems : STREEMS and METIORE. STREEMS manages information on trees while METIORE manages information on bibliographic references.