Using Hierarchical Clustering for Learning the Ontologies used in Recommendation Systems
Ontologies are being successfully used to overcome semantic heterogeneity, and are becoming fundamental elements of the Semantic Web. Recently, it has also been shown that ontologies can be used to build more accurate and more personalized recommendation systems by inferencing missing user's preferences. However, these systems assume the existence of ontologies, without considering their construction. With product catalogs changing continuously, new techniques are required in order to build these ontologies in real time, and autonomously from any expert intervention. This paper focuses on this problem and show that it is possible to learn ontologies autonomously by using clustering algorithms. Results on the MovieLens and Jester data sets show that recommender system with learnt ontologies significantly outperform the classical recommendation approach.