A Comparative Study Of The "Gate-keeping" Behaviors Of Algorithms In Personalized News Recommender System

With the rapidly development of artificial intelligence technology,the way of journalism is produced has been unprecedented changed.As the algorithm technology is gradually applied to the personalized news recommendation system,the "gatekeeper" mode in the traditional news production process is changed from the news selection under the "editing check" to the news distribution under the "algorithm check".Therefore,understanding the impact of new technologies on the “gate-off mode” have highly research value.This paper selects two personalized news recommendation platforms with great influence in China,which called "Today's headlines" and "Yidian information".There are three points that contributing to the research of the algorithmic control behavior,which are the algorithm as the "gatekeeper",The influence of the algorithm's gatekeeper behavior and the situstion of algorithm as the "gatekeeper".First of all,the first chapter summarizes the gatekeeper behavior mechanism of the personalized news recommendation system's algorithm,and sorts out the techniques,implementation process and check features of the algorithm.Secondly,by useing the content analysis method and the interview method,in the second chapter of the thesis,we compares the behavioral mechanism of the "headlines" and "Yidian information" algorithm from the three dimensions of the gatekeeper behavior's subject,the process and the standard.By comparing the similarities and differences between the two platforms,this paper summarizes the factors that affect the algorithm.The study found that the three factors: algorithmic values,the degree of capital and market tendencies,and the external role of government and public opinion,which influence the rules of the personalized news recommendation platform's algorithm.The third chapter further analyzes the impact of the algorithm's gatekeeper behavior.On the one hand,the control of the algorithm improves the efficiency of news distribution,promotes the return of the audience standard,and causes the evolution of the traditional gatekeeper mode.On the other hand,the control of the algorithm has also triggered a series of negative social impacts such as the lack of news values,the narrowing of user information,and the proliferation of vulgar content.Finally,this paper proposes a solution to the negative impact of the personalized news recommendation system's algorithm.The personalized news recommendation platform itself should improve the algorithm's gatekeeper behavior,improve content standards through optimization technology,and also make the algorithm maintain a certain width while "precisely pushing";users should improve their media literacy and form a "self-check" awareness.It is also necessary to diversify the reception of information to avoid narrowing the information;the government must promptly rectify the network chaos,improving the relevant laws and regulations,so that we can maintain a good network environment.