智能算法推荐的社会风险及其法律规制  

Social Risk of Intelligent Algorithm Recommendation and Its Legal Regulation

作　　者：梅帅 Mei Shuai(School of Law,Renmin University of China,Beijing 100872,China)

机构地区：[1]中国人民大学法学院

出　　处：《科学．经济．社会》2022年第1期80-96,共17页Science Economy Society

基　　金：中国人民大学2020年度拔尖创新人才培育资助计划;国家社会科学基金重大项目“紧急状态的类型化和立法研究”(20&ZD175)。

摘　　要：从本质上看,智能算法推荐是以数据收集为基础,以决策运作为机理,通过数据运算,为用户提供决策参考的技术,其主要有协同过滤推荐、基于内容的算法推荐和混合算法推荐三种主要类型。它具有智能化、便利化和个性化特点的同时,也可能带来权益侵害、秩序妨害、技术异化、责任模糊等社会风险。因此,应以风险社会的法律理念为引领,按照公平原则、公开原则和社会利益原则的要求,加强对算法推荐合法性的前置审查,建立健全具体化的算法推荐场景规范,强化算法的行业监督与第三方评估,完善算法推荐的责任认定分担机制,更好地发挥智能算法推荐的积极功效,构建适应算法技术发展的法治秩序。Intelligent algorithm recommendation is the technology based on data collection,decision-making operation as the mechanism,and provide reference to users through data operation in essence,which has three main types including collaborative filtering recommendation,content-based algorithm recommendation and hybrid algorithm recommendation.It has the significant features of intelligence,convenience and individuation,and also probably brings social risks such as infringement of rights and interests,disturbance of order,alienation of technology,ambiguity of responsibility and so on.Therefore,it is essential to take the legal idea of risk society as the guide direction,according to the principles of fairness,openness and social benefit,reinforce the pre-examination of the legitimacy of algorithm recommendation,establish and consummate the specific algorithm recommendation scenario specification,strengthen the supervision and the third-party evaluation of algorithm recommendation,and improve the responsibility identification and allocation mechanism of algorithm recommendation,so as to give better function to intelligent algorithm recommendation and establish social order of the rule of law with adapting to the development of algorithm technology.

关 键 词：智能算法推荐 社会风险 法律规制 法治秩序 

分 类 号：D922.16[政治法律—宪法学与行政法学] G202[政治法律—法学]