| In the Internet era,people are facing a serious problem of information overload.How to find useful information in the vast amounts of data has become a hot research topic.Information classification and search engine have solved this problem efficiently.Recommender system is another sounder way to solve the information overload problem.Compared to the former two,recommender system is more intelligent and active.The advantage is that recommender system can guide people to dig some information they really need by researching people s history and some other public data and doing calculate.News recommender system,which purpose is intelligence for users to recommend the news information they interested in,is one of the applications in recommender field.News recommender system faces to these problems which exist in recommender field,and not only that,these problems have their own special properties.Therefore,the key point of this thesis is to design and implement a news recommender system,and solve the problems faced by the field of news recommendation.Facing the recommended field problem also exists in the news recommendation,and there is a special field,so the focus of this paper is to design and implement a news recommendation system,and solve the problems faced by the field of news recommendation.Through the analysis of research status of the recommendation system and summary of the main problems,the main contents of this paper are as follows: firstly,the basic workflow of the recommendation system summary,analysis principle commonly used algorithm recommended in the system and the advantages and disadvantages;then according to the news recommendation data sparsity problems and user interest drift to solve the problem,through the existing scheme,design the collaborative filtering algorithm is improved and tested.Finally,the scheme of news recommendation system based on mobile Internet technology is designed,and finally the system is implemented.The experiments on the thesis proves that the recommendation method in this thesis is more effective in solving these problems that News recommender system faced on,and the tests proves that the news recommender system Based on Hadoop has better expansibility and real-time ability. |
