Design And Implementation Of Personalized Recommendation System Based On User Context

With the continuous development and innovation of today's information technology and the Internet technology, For consumers, when they are facing of the wide variety of today's mass of information, how to find out the information they need from the information they are interested in has become a thorny problem to be resolved. Confronted with the problem of information overload, recommended system technology has long been recognized as a key technology to solve the problem of information overload.Location-based Services (LBS), incorporates a new dimension on the basis of traditional applications, namely location, as one of the hottest application services in the field of recommendation system. But it searches information under the merely general geographic location information, therefore, ignores the degree of association with search results and the user interest hobby. On the basis of the geographic information system service technology of High German, the user's historical behavior preferences and user location scenarios are introduced. The first step is to filter the services around the user after identifying the location scenarios and filtrating based on the similarity, maybe restaurants, theaters and hotels, then filtering item similarity, eventually a list accorded with the user's preferences is pushed. However, due to the increasing number of driving user, the parking spaces are gradually being considered as one of the factors to choose the item of interest. The system has not only provided the detailed information of user's preferable item but also designed the prompt function of parking space around the related item.Some relevant design techniques of this system are introduced at the beginning. Next, an elaborated explaining of the specific design and realizations of every modules of the system are provided. And at last, after conducting some tests, results showed that the accuracy of this recommendation system could reach 81.56%, and the accuracy of the prompt facility of park lot status could reach 87%. Therefore, the realization of this system has mitigated the problem of the Information Overload and the difficulties of parking to some extent.