Impact of Personalized Recommendation on Today’s News Communication through Algorithmic Mechanism in the New Media Era

In recent years, the continuous innovation of technology has greatly contributed to the great changes in the media industry. The rapid development of big data and artificial intelligence technologies has enabled people to transition from the era of new media to the era of intelligent media. While the automation of news production brings broad prospects for intelligent media, it also accelerates the challenge of information explosion. Facing the massive amount of news and information, how to get the information users want quickly has become a big problem. In order to solve the audience’s information anxiety, personalized news recommendation system is born. In fact, news gate-keeping is an important part of news distribution. In the era of smart media, algorithmic distribution has impacted the original distribution mode and brought challenges to news gate-keeping. Personalized news recommendation is one of the gate-keeping methods of intelligent media. At the same time, with the rapid development of the Internet and information technology, today’s society has entered a period of information explosion. In terms of news, the rapid development of the Internet has made it easier to publish and read news on the Internet. As a result, online news has become an important way for people to get information. However, the previous news websites had a large amount of news information, but only collected and consolidated the news. As a result, users were left to passively receive news information from news sites and find the content they needed. Consequently, although the Internet has a huge amount of complicated news information, it is unable to meet the diversified and personalized news needs of users. In order to solve this issue, researchers are constantly looking for solutions. The emergence of recommendation system is an effective measure to cope with the above problem. The mainstream models of recommendation systems are collaborative filtering model and content-based recommendation model. However, there are two essential problems with collaborative filtering. The first one is the cold start problem, and the second one is that the preference matrix of item users becomes sparse as the number of items and users grows. These two issues can seriously affect the recommendation accuracy of the recommendation system. As a result, a hybrid recommendation system is built by fusing common recommendation algorithms. This system can not only deliver personalized information to different users, but also compensate the shortcomings of a single algorithm to a certain extent. To be specific, the newly constructed hybrid recommendation system can push news of interest to users according to their demographic attributes, behavioral attributes, and interests, thus expanding the scope of news communication.