The Study On The Gatekeeping Behavior Of Personalized News Recommendation

In recent years,with the rise of aggregated news clients such as "Toutiao","Yidianzixun" and "Kuaibao",the personalized recommendation mechanism for news has quickly entered people's vision.It has become an important way and a new type of service for connecting users and their required information in the era of mass content.The personalized news recommendation mechanism,which is based on aggregated news clients,is devoted to satisfying users' different interests in news and information.It plays an important role in solving information overload,reducing information redundancy and improving the efficiency of information distribution.However,this kind of new information distribution and dissemination method driven by technology is mainly implemented through computers,and the issue of its content gatekeeping is worth considering and discussing.Taking Toutiao APP as an example,this paper tries to reveal how the personalized news recommendation checks the news content by examining its content production mechanism and distribution mechanism.The article first introduces the technical features of "Toutiao",the reasons for its rise,its media attributes,and its conduct of checks.On this basis,this paper clarifies that the main gatekeepers of personalized news recommendation include four categories:content author,audit editor,algorithm and user.They acted as different gatekeepers,and implemented the actual operations of the news content gatekeeping at different stages.The gatekeeping mechanism of personalized news recommendation can be divided into three categories:content creation,content verification and algorithm recommendation.The gatekeeping process can be described as three steps:content creation gatekeeping?content check gatekeeping-algorithm recommendation gatekeeping.The result of the gatekeeping triggers constant interaction between users and algorithms.After confirming the gatekeepers and analyzing the mechanism of the gatekeepers,This paper points out that the gatekeeping behavior of personalized news recommendation shows the characteristics of algorithmic dominance,human-computer integration,and standard remodeling.In personalized recommendation,the gatekeeper relationship between the algorithms and users can be better explained by the structural duality principle.In news algorithm recommendation,the user's choice of news information is not only the rational choice of users to meet their own needs,but also not just the restriction result of the algorithm content recommendation.It's the dual process of action and structure.The users' action is structured in the process of using the algorithm recommendation.At the same time,the actions of the users constantly reflect this structured feature through their own functions.The algorithm as a structure interacts with the user as an actor,constructing each other,and coexisting cyclically.At present,there are many hidden concerns in the content gatekeeping of personalized news recommendation,including the lack of value judgment and the distortion of content production at the content level.It also includes the production of information cocoons and filter bubbles,the construction of user interest and the user's group constraints at the user level.Only by clearly clarifying its gatekeeping behavior and finding out the root of the problem,can we "prescribe the right medicine" and solve the above problems.