Design And Implementation Of A User Preference Extraction System Based On Ontology And Tags

With the rapid development of information technology, the internet has changed dramatically. Now, along with the development of integration of three networks and 3G technologies, mobile internet has ascended to the history stage. Lured and stimulated by the new chance, telecommunication operators begin to change their development concepts of single voice service, and turned their attention to the mobile network information service field, providing more data services for the user. With the growing number of services in the mobile network information service field, users need to spend a lot of time and battery power to find the services that they need through the handset. In addition, the user's consumption concept has become increasingly mature and rational, and the user's demands have become increasingly diversified. All of above increase the difficulty when users want to get their required services. The telecommunication operators' and service providers'traditional mode of information systems can not meet their service needs in this new field. The construction of personalized service system is a good idea to improve the existing system in the mobile network information services field, and also is an urgent need. The extraction of user preference is a key technology of personalized service system, but also is the key point to meet the needs of individual users.This paper analyzes and compares the existing user preference modeling techniques, and proposed a thought that extracting user preference from multiple perspectives. Mainly including:adding the conception of tag computing, proposed extraction of user preferences method based on tags; using the appropriate domain knowledge ontology to describe the services, proposed extraction of user preferences method based on ontology; through service domain ontology and user ontology optimization of user preferences; proposed the conception of atom preference.Tag-based extraction of user preference is using the service tag to build space vector, as a user model representation, with the help of some information retrieval algorithms to calculate the preference of the user to the tag.Ontology-based extraction of user preference is the improvement of tag-based. Experimental results showed that the semantic processing applied to the extraction of user preferences is necessary.