把关与操纵:中国主流媒体与商业社交媒体推荐算法价值的比较研究  

Gatekeeping and Manipulation:A Comparative Study of the Values of Recommendation Algorithms in China's Mainstream Media and Commercial Social Media

作　　者：陈积银 胡睿心 CHEN Jiyin;HU Ruixin

机构地区：[1]西安交通大学新闻与新媒体学院,陕西西安710049

出　　处：《福建师范大学学报（哲学社会科学版）》2024年第2期105-117,170,171,共15页Journal of Fujian Normal University:Philosophy and Social Sciences Edition

基　　金：国家社会科学基金后期资助项目“智能推荐型视频产业价值链研究”(21FXWB003);科技部项目“智能信息传播与社会治理”(G2021170010L)。

摘　　要：算法把关使传统的新闻把关方式受到了冲击,同时也为互联网平台治理带来了挑战。通过连续跟踪爬取央视频、微博、抖音、微信视频号以及《今日头条》共18000余条热搜,从“量”与“质”两方面,以正面性、机构议程性、建设性、实时性、需要性、共情性、可分享性、世俗性和猎奇性这九个维度为分析框架进行综合性媒体算法价值探赜,分别考察中国主流媒体与商业社交媒体推荐算法价值之间的特点与异同点,以及其对中国媒介生态中的受众造成的影响。研究认为,主流媒体和商业社交媒体的热搜关键词都和算法把关有着直接关系,人工干预在其中起到了不同程度的辅助调控作用;主流媒体承担着更多的新闻价值和社会责任使命,信息迭代速率较低,在内容聚合方式上进行了严格把关,而商业社交媒体打造了一个表面上分散且民主的内容制作和选择系统,但其算法推荐主要遵循受众商品论和利益至上原则,其所谓的人工调控只是在迎合算法“流量至上”原则基础上对内容“合规性”的一种“审查”,而非发挥人类主观能动性的信息“筛选”。商业社交媒体“信息筛选主体和工作量的隐蔽转移”和“遵循资本逻辑的算法对受众隐秘规训”的特质,揭示了平台利益至上的本质,为中国网络平台治理提供了理论指导与实践参考。Algorithmic gatekeeping challenges traditional news curation methods and poses significant challenges to the governance of internet platforms.This study innovatively tracks and collects over 18,000 trending topics from China Central Television(CCTV)videos,Weibo,Douyin,WeChat video channels,and Toutiao,employing both quantitative and qualitative methods.It adopts a nine-dimensional framework for a comprehensive analysis of media algorithm values,focusing on positivity,institutional agenda-setting,constructiveness,timeliness,necessity,empathy,shareability,secularity,and sensationalism.The study explores the characteristics and differences in the recommendation algorithms values of China's mainstream media and commercial social media,as well as their impacts on the Chinese media audiences.The study finds that trending keywords in both mainstream and commercial social media are directly related to algorithmic gatekeeping,with varying degrees of human intervention playing a supplementary regulatory role.Mainstream media carry greater news value and social responsibility,exhibit a slower information iteration rate,and rigorously gatekeep content aggregation methods.In contrast,commercial social media have established a content creation and selection system that ostensibly appears decentralized and democratic.The algorithmic recommendations of these social media platforms predominantly follow the principles of audience commodification and a profit-centric approach.This so-called human intervention is essentially a“review”for compliance with the algorithm s“traffic supremacy”principle,rather than a genuine manifestation of human subjective agency in information“filtering”.The covert shift of“information filtering agents and workload”and“subtle audience discipline through capital logic algorithms”in commercial social media expose the platform s profit-driven nature,providing theoretical insights and practical guidelines for the governance of Chinese internet platforms.

关 键 词：算法把关 算法价值 网络平台治理 主流媒体 商业社交媒体 

分 类 号：G206[文化科学—传播学] G210.7