A personalized recommender system based on users' information in folksonomies

Thanks to the high popularity and simplicity of folksonomies, many users tend to share objects (movies, songs, bookmarks, etc.) by annotating them with a set of tags of their own choice. Users represent the core of the system since they are both the contributors and the creators of the information. Yet, each user has its own profile and its own ideas making thereby the strength as well as the weakness of folksonomies. Indeed, it would be helpful to take account of users' profile when suggesting a list of tags and resources or even a list of friends, in order to make a more personal recommandation. The goal is to suggest tags (or resources) which may correspond to a user's vocabulary or interests rather than a list of most used and popular tags in folksonomies. In this paper, we consider users' profile as a new dimension of a folksonomy classically composed of three dimensions "users, tags, ressources" and we propose an approach to group users with equivalent profiles and equivalent interests as quadratic concepts. Then, we use quadratic concepts in order to propose our personalized recommendation system of users, tags and resources according to each user's profile. Carried out experiments on the large-scale real-world filmography dataset MovieLens highlight encouraging results in terms of precision.