The Study On Problems And Regulatons Of Algorithm Recommendation From The Perspective Of Gatekeeper Theory

In the new media era,every new communication technology is born with an iterative update of the form of communication.With the development of big data and Internet technology,information has grown exponentially,but it also makes it difficult for users to quickly find the information they need in massive information.Algorithm recommendation is an effective way to solve this problem.and it is widely used in the field of news communication.In particular,the birth of a series of mobile news apps featuring algorithmic recommendations,such as Jinritoutiao and Tiantiankuaibao,opened up a new era of content distribution,and personalized news recommendations for users became the mainstream of content distribution.At the same time,algorithm recommendations also bring a series of social problems.In fact,the algorithm recommendation can be regarded as a gatekeeping process,and the technical means are used to determine what the users read.Therefore,this study combined the theory of the gatekeeper to analyze the gatekeeping phenomenon in the algorithm recommendation process.We discussed the technical logic and its operational process,then analyzed the gatekeeping phenomenon in combination with the algorithm recommended operational process.The research found that in the algorithm recommendation environment,the gatekeeping process forms a closed-loop system,and the interaction between the various gatekeepers.Secondly,we analyzed the negative problems in the operation of recommendation algorithms through interviews with relevant internet practitioners.In terms of content,homogenization and advertising flooding are the main problems.The subjective judgment of the algorithm designer may embeds the bias into the algorithm system,and the operator's intervention may cause information manipulation.The positive feedback mechanism of the algorithm will aggravate the phenomenon of "filter bubbles".When users indulge in the pseudo-environment created by their own interests,it is easy to make the public lose their reflection on public events,pay more attention to information than to meaning,resulting in inadequate understanding of the real world.The algorithmic system relies on users data,which also brings about privacy violation.Finally,this study discussed the current relevant laws and regulations,industry norms,based on this,we put forward reasonable suggestions for the prevention of negative problems in algorithm from the two aspects of technology evaluation and audit in algorithm design and data and algorithm transparency in the operation process.