Abstract:

With the rapid development of mobile social network, the problem of “mobile information overload” arises at the historic moment, a lot of complex mobile information causes users’ personalized service needs to be covered in mobile social networking services. In order to make a timely and accurate prediction of personalized user preference in mobile social network, we propose a method to acquire user preference based on multi-dimensional scenarios and a contextawareness model of mobile social networks. Firstly, according to the multidimensional scenarios of context, user cognitive behavior and quality of service, we respectively extract user preferences based on valid context, user’s effective recognition and the quality attribute of the service. Then, we construct a user-awareness model of mobile social network by fusing the multi-dimensional user preference. Finally, we use the method of hierarchical vector space to describe the user preference model, and update the model through update-processing mechanism based on the user-feedback information. The model has realized the quick prediction, dynamic tracking and updating of the user preferences in mobile social network, the accuracy of user preference prediction has been improved, which makes user personalized information needs to be better satisfied in mobile social network.

Key words: Multi-dimensional scenario, Mobile social networking, User preference, Context, Cognitive behavior