Design And Implementation Of Personalized News Recommendation System Based On Hybrid Recommendation Algorithm

With the continuous development of information technology,we are now in an era of information explosion,but also faced with the problem of information overload,that is,it is difficult to find the news we are interested in from the huge news sources in the network.News service providers usually use personalized recommendation system to recommend to users.The traditional methods of personalized news recommendation include content-based recommendation,collaborative filtering recommendation and hybrid version of the two technologies.Traditional recommendation algorithms do not consider the time order of users in browsing records,and the reading order information of users'history records can better reflect the changes and diversity of users'interests in a period of time.Based on this problem,this paper designs and implements a personalized news recommendation system based on hybrid recommendation algorithm(1)In view of the fact that the traditional recommendation algorithm ignores the time sequence of users' browsing behavior,this paper proposes an improved recurrent neural network algorithm model,which uses attention based parallel convolutional neural network to aggregate users' interest features,and uses attention based recurrent neural network to mine hidden time series features.(2)In order to improve the problem that the single recommendation algorithm has insufficient recommendation effect in some application scenarios,this paper proposes a hybrid recommendation algorithm model,which uses the weighted hybrid recommendation strategy to fuse ALS model recommendation,content-based recommendation and improved recurrent neural network algorithm,and obtains the hybrid algorithm model by training the weight through logical regression.(3)In this paper,we use big data technology spark to design and implement a personalized news recommendation system based on hybrid algorithm model.The system test shows that the performance of the recommendation system is excellent.