A Big Data-Driven Hybrid Recommendation Method for Accurate News Media Content Personalization

This paper deeply analyzes the application of big data models in news media content recommendation and proposes a new hybrid recommendation method that combines deep learning (RNN) and a recommendation algorithm based on news type preference characteristics. The method uses technologies such as data mining and machine learning to deeply mine user information, such as user behavior data and interest preferences, to achieve more accurate news media content recommendations for users and solves the problems of insufficient accuracy and lack of in-depth user interest mining in news media content recommendations of traditional content recommendation algorithms. This paper elaborates on the specific application of the big data model in news content recommendation. The experimental results show that the recommendation accuracy of the proposed hybrid recommendation (HR) algorithm is 95.9% and 93.8% in the news of beauty cosmetics and electronic products, respectively, which is much higher than the other three recommendation algorithms. At the same time, the average click-through rate of the hybrid recommendation algorithm in the news media content recommendation is increased by 20%. Compared with FM, CF and ARR recommendation algorithms, the proposed method has obvious advantages, which verifies that the proposed method can effectively solve the problems of traditional recommendation algorithms, improve the accuracy and diversity of news media content recommendation, and bring more traffic and revenue to news media organizations.