A method for the acquisition of ontology-based user profiles

• The algorithm introduces simulated annealing to hierarchical clustering in the construction of user ontologies.
• The k-nearest-neighbor query method is combined with branch-and-bound search algorithm in the maintenance of user ontology.
• The method has less human intervention in creating user profiles and can make user profiles more strong expressive.

User profiles play an important role in information retrieval system. In this paper, we propose a novel method for the acquisition of ontology-based user profiles. In the method, the ontology-based user profiles can maintain the representations of personal interest. In addition, user ontologies can be automatically constructed. The method can make user profiles strong expressive and less manually interfered.