Research On Negative Problems And Countermeasures Of Algorithm Recommendation From The Gatekeeper Perspective

In recent years,China’s short video industry has developed rapidly,which has launched many APPs such as Tiktok,Kuaishou,Wechat,etc.The 51 st “China Internet Development Statistics Report” shows that as of December 2022,the short video users exceeded one billion for the first time,with an App utilization rate of 94.8%.This shows a large scale of China ’ s short video users and a high demand for Internet entertainment.One of the reasons why these short video APPs are favored by users is that their algorithmic recommendation technology enables users to see what they are interested in accurately,and personalized algorithmic recommendation has become the mainstream distribution mode of short video content.At the same time,however,it has also brought some problems.The algorithm applied in short video has improved the efficiency of content distribution,but it also brings some difficulties for winnowing the content.In terms of the short video platform,the traditional winnowing model is no longer applicable to the massive content production mode,since the content producers(PGC,UGC),platform operators(algorithms,artificial),content consumers(short video users)are involved in the process of winnowing video content as gatekeepers.Based on the classical gatekeeper theory in communication science,this paper analyzes the content operation process and algorithmic recommendation mechanism in the short video app represented by Tiktok,summarizes the negative problems brought by algorithmic recommendation to the short video platform,and proposes strategies from the gatekeeper perspective.This paper first clarifies the problem from the perspective of the gatekeeper.By means of in-depth interview,this paper first clarifies the operation process and the technical logic of the algorithm recommendation used by the short video platform represented by Tiktok,and then summarizes the gate-keeping behaviors in the algorithm recommendation process by the questionnaire method,finding that in the algorithm recommendation,the gate-keeping subjects of the short video platform are not only limited to the platform,and there is an interacted relationship in the process of content production and distribution between users,platform and algorithm,who play the different by means of in-depth interviews roles.On this basis,this paper summarizes the problems brought by algorithmic recommendation to the short video platform with the help of participant observation and in-depth interview method,such as homogeneous and low-quality content proliferation brought by the algorithm recommending in terms of the content that users are interested in;users are immersed in the mimetic environment created by the algorithm,easily generating extreme views and thinking irrationally,which then leads to group and social polarization.The algorithm designer may have bias in designing the algorithm program due to subjective judgment or insufficient cognition,leading to algorithmic bias and information manipulation.Finally,this paper proposes countermeasures for the above problems from three levels,namely users,platforms and the state.For example,in terms of the proliferation of homogeneous and low-quality issues,the state should encourage more high-quality content,platforms should scientifically push heterogeneous information,and users should domesticate the algorithm through explicit and implicit feedback behaviors;as for the group polarization and social division,the state should strictly punish those people with comments of malevolence,platforms should encourage users to adopt “agenda setting”,and users should improve their media literacy and domesticate the algorithm;in terms of the algorithmic bias and information manipulation,the state should establish an algorithm evaluation and audit department,platforms should improve the transparency of the algorithm,and different subjects should strengthen their cooperation,etc.