Network media news recommendation method based on high-dimensional auxiliary information

Abstract

The invention discloses a network media news recommendation method based on high-dimensional auxiliary information. The method comprises the steps of generating a main information feature vector according to a title and an abstract of news; generating a high-dimensional auxiliary information feature vector according to the content of the news; establishing a preference prediction model of the usernews and an optimization model predicted by the preference prediction model; performing parameter optimization on the optimization model; substituting the optimized parameters into the preference prediction model; and recommending news to the user according to the score calculated by the preference prediction model. According to the invention, historical preference behaviors of specific users arefully utilized to provide information guidance for recommendation; the two feature vectors provide more comprehensive feature extraction for news, in the preference prediction model, the contention loss function is minimum to achieve better fitting, and meanwhile, sparse regularization penalty is introduced to prevent overfitting in the model optimization process, so that the method has better accuracy and quicker timeliness for recommendation of specific users.

Description

technical field [0001] The invention belongs to the field of intelligent recommendation of text content, and in particular relates to a network media news recommendation method based on high-dimensional auxiliary information. Background technique [0002] In this era of rapid development of the Internet, there will always be a large amount of novel information coming to us. For Internet users, it is very cumbersome and unrealistic for them to filter interesting and high-quality information from it. The emergence of the search function has solved the problem of information screening and recommendation to a certain extent, but it is still far from enough. The search function requires users to actively provide keywords to filter massive information. When users cannot accurately describe their needs, the filtering effect of search engines will be greatly reduced. The emergence of personalized recommendation system has greatly reduced the contradiction between users' personaliz...

AI Technical Summary

Problems solved by technology

[0003] Network media news is an important means for users to quickly learn about the world. When it makes automatic news recommendations, it often encounters several problems. First, how to use the past behavior preferences of specific users to guide recommendations. 2. For the news recommendation, only a few keywords are featured, ignoring the information brought by other texts, which will lead to poor recommendation effect. Third, for the news or articles just reported, how to quickly judge whether to recommend to a specific user

Images