Research On User Real-time Interest Adaptation Model And Personalized Recommendation Method Based On Geographical Context

The rapid development of the Internet has led to the rapid growth of network information.On the one hand,massive information provides users with a variety of information selection basis.On the other hand,the large amount and variety of Internet information makes it difficult for users to quickly find their actual needs in massive data.The emergence of personalized recommendation system has solved this kind of problem better,collaborative filtering algorithm is the core of personalized recommendation,and there are many experts and scholars still in constant exploration and research.This paper considers the problems of data sparseness and cold start faced by collaborative filtering algorithms,as well as the characteristics of context-aware and collaborative filtering algorithms for independent recommendation,from the perspective of users,establishing the user interest model based on geographic context,and a recommendation algorithm of hybrid collaborative filtering and association rules is proposing to improve the accuracy and efficiency of personalized information recommendation service.This paper specifically studied the following aspects:(1)Through the research on context perception,user interest model and personalized recommendation system at home and abroad,it is found that the user's interest preference is closely related to the user's current behavior and the geographical situation of the user.Through the research of situation definition,situation classification and context-aware computing service framework,the realization of the situational awareness,on the basis of which the concept of geographic context,abstract expression and other related knowledge,Finally,three kinds of personalized recommendation technology models based on geographic context perception are expounded,and the recommendation mode suitable for the research content of this paper is selected and explained.(2)User interest preferences and user behavior are in the process of change all the time,with strong dynamics and real-time.The comprehensive degree of acquired information is one of the key factors affecting the pros and cons of the model.Based on explicit interest items,implicit interest items and geographic scene interest items,a real-time interest adaptation model based on geographic context is constructed to better restore the user's real interest through the comprehensive expression of user interest.(3)Aiming at the problem that the traditional collaborative filtering algorithm considers the “user-project” scoring information and leads to sparse data and cold start,this paper adds a comment tag based on the scoring label and proposes a multi-dimensional tag-based collaborative filtering algorithm.The MovieLens dataset is compared with the traditional collaborative filtering algorithm.Experiments show that the improved algorithm is superior to the traditional collaborative filtering algorithm.The Apriori algorithm is used to analyze the association between geographic context and user behavior interest.The experiment proves that the method is feasible.Finally,the two algorithms are weighted into a personalized hybrid recommendation algorithm.(4)Designed a personalized recommendation service for travel information.The design of personalized information recommendation system based on user interest mainly includes recommendation system overall framework design,tourism database design,recommendation system application interface and other related design,in order to realize the real and effective recommendation of tourism resources.In the research of personalized recommendation,the association rule and collaborative filtering algorithm can solve the problem of data sparsity.The user interest model can effectively provide users with dynamic,real-time,efficient and suitable information for their own interests.Personalized information services that contain the user's current geographic context provide new ideas.