Research Of Recommendation System Based On Context

In recent years, with the development of the Internet, online shopping become indispensable in people's lives, with much of the growth in electronic commerce website users and number of items, how to quickly recommend users sense of items of interest has become urgent problems. Traditional recommendation algorithms more attention to user rating, the user's context information under utilized, based on this topic in the in-depth study of the existing personalized recommendation technology, according to the user's context, research and design of the a collaborative filtering recommendation algorithm based on context.First of all, this paper describes the commonly used in personalized recommendation systems, especially collaborative filtering technology, and fully understand the principle and classification of collaborative filtering algorithm. The specific process of several classical algorithms are introduced in detail, and the advantages and disadvantages of these algorithms are analyzed. Then introduces the related technologies of the context aware recommendation system, and finally introduces the method of quality assessment of the commonly used recommendation algorithm.Secondly, this paper presents a contains the context similarity of context weight calculation method, in view of the existing context method to introduce the problem, given the combination of a context-pre filtering and context-modeling of context introduction method, and proposes a context based on similar degree of slope one algorithm, data to predict the fill, reduce the data sparsity, in order to reduce the adverse effects of data sparsity of the algorithm.Again, through the study of classic collaborative filtering recommendation algorithm.It is found that the traditional collaborative filtering algorithm in the presence of that context similarity is proposed a collaborative filtering recommendation algorithm based on context similarity. Compared with the traditional recommendation system, it is more accurate to mine the user's interest preference by using the context similarity, which greatly improves the accuracy of the recommendation.Finally, the performance of the improved model and algorithm is verified byexperiments, and the results are compared with other algorithms. The results show that the algorithm can get better results in a certain degree. Also tested the algorithm performance in different data sets to verify the robustness of the algorithm. The research can improve efficiency of collaborative filtering recommendation algorithm, for personalized recommendation technology in practice provides certain theory and method support.