User profile construction associated with a preferences database

The purpose of the user profile is to use the latter's preferences in order to personalize the responses to his requests and to carry out targeted recommendations. A user profile can be built both manually or automatically, except that the manual way remains time consuming and complex for the user. And that’s the reason why we present in this article an automatic method based on data mining techniques. The approach that we intend to suggest in this article consists of two main phases, the first consists of extracting the contextual preference rules, the second is the construction of the user profile which allows us to eliminate all the redundant and super fluous rules extracted from the first phase, the goal is to have a consistent and concise user profile. A consistent user profile ensures that the preference rules defining the profile match the maximum of this user's preferences and contradict few of them. A concise user profile implies that the profiles consists of a small number of preference rules. In this article, we will mainly define the concepts used in order to understand the extraction problem that we intend to solve and define formalization of the problem. We Validated Our Approach on a film database. The approach that we intend to suggest in this article consists of two main phases, the first consists of extracting the contextual preferencerules, the second is the construction of the user profile which allows us to eliminate all the redundant and superfluous rules extracted from the first phase, the goal is to have a consistent and concise user profile. A consistent user profile ensures that the preference rules defining the profile match the maximum of this user’s preferences and contradict few of them. A concise user profile implies that the profiles consists of a small number of preference rules.