Design And Implementation Of News Recommendation System Based On Text Processing

The development of information technology makes people get information more and more conveniently,but it also brings some problems such as information overload.Meanwhile,with the rise of mobile Internet and the popularity of intelligent terminals,the users' requirements for precise personalized recommendation are also increasing.In the field of personalized news recommendation,it is an urgent problem to be solved that how to dig out the content of users' interest and make accurate recommendation from a lot of news information.In this context,this paper improves the content based recommendation algorithm through the study of personalized recommendation technology.Around the text modeling part of personalized news recommendation,the process of text processing is optimized by combining the topic model and vector space model to improve the accuracy of the news recommendation system.In the news recommendation system based on text processing in this paper,the method of text processing is the key,and the text content mining determines whether the analysis of the user's interest is accurate.Therefore,in the aspect of news text processing,the topic model is introduced to combine it with the traditional space vector model,and the text content of the news is excavated with two aspects of the topic and the key words.Meanwhile,the timeliness of news is also taken into account.News release time in the system also needs to be an important index for news text modeling.On this basis,it combines the news and the user behavior data to calculate the user's topic and keyword preferences,and constructs a more accurate user interest model by combining the context information of the users when produced the behaviors.At the same time,the process of news recommendation generation should also be improved with considering the timeliness of news to recommend.Finally,the optimized recommendation method is applied to the specific news recommendation system,with the overall framework of the system is given.The implementation details of the core modules of the recommendation system are described in detail.In terms of experimental verification,this paper validates news recommendation system from the aspect of performance.The proposed method based on text processing is better in performance indicators,such as accuracy and diversity.