Research On Personalized User Model Based On Ontology

Personalized service is a serving manner which can satisfy user individual requirement. Through studying users' behavior and habits, we give users different service strategies. Ontology can provide a structuralize user model storage pattern and help system study user preference, so user interests would be satisfied much better. Personalized user model basing on ontology and automatic generation for user interest ontology have become new research task.At first, we compare several user model presentation methods, and indicate the excellence which ontology as a user model presentation method has. We gave an information retrieval model basing on ontology, and analyze the user model influence factors. Then we combine user preference with concept hierarchy for interest ontology, and improve personalized user model studying courses and concepts interest degree updating courses.Secondly, we give an automatic ontology generation method for user interest. FOGA can generate fuzzy ontology automatically from uncertainty data thorough formal concept analysis, concept hierarchy generation, fuzzy ontology generation and semantic representation conversion. We use fuzzy concept cluster algorithm, which associates every concept with object and attribute, and describes concept cluster to object and attribute sets.Moreover, we use approximating reasoning method to enrich fuzzy ontology generated, and integrate additional attribute from database into the ontology. It confirms concepts belonging to different classes, and associate class with objects and attributes relative to concepts, so to complete dynamic updating for the whole knowledge base.Finally, the fuzzy ontology automatic generation method mentioned above was proved by the experiment. We use Recall, Precision and F-Measure to evaluate cluster results, Relaxation Error to observe the keywords generation effects, and Average Uninterpolated Precision to estimate concept hierarchy structure.