论智能互联网时代的算法歧视治理与算法公正  

On the Governance of Algorithmic Discrimination and the Establishment of Algorithmic Fairness in the Age of Smart Internet

作　　者：杨学科 YANG Xueke(School of Law, Jilin University, Changchun 130012, China)

机构地区：[1]吉林大学法学院

出　　处：《山东科技大学学报（社会科学版）》2019年第4期33-40,58,共9页Journal of Shandong University of Science and Technology(Social Sciences)

基　　金：国家社科基金重大专项项目“核心价值观融入法治建设研究:以公正司法为核心的考察”(17VHJ007);教育部人文社会科学研究专项项目“新时代中国特色社会主义法治思想研究”(18JF210)

摘　　要：智能互联网时代到来,也是算法时代的到来,与之相随的算法歧视也正在成为互联网和算法治理的难题。一般而言,内置性编码凝视,有限、不相关或不正确的训练数据,算法黑箱建模方法或三者的不同组合都可造成算法歧视。面对算法歧视的危害,可以通过技术治理上的算法歧视纠偏和制度治理上的算法公正立塑来化解。其中算法歧视技术治理可在多样性加入、透明度、问责制上着手。算法公正的实现则可主要围绕着算法公正价值观、多元共治、算法素养三个方面来进行。The arrival of the smart Internet era also represents the arrival of the era of algorithms.The accompanying algorithmic discrimination is posing a difficult problem to Internet and algorithmic governance.In general, built-in encoding gaze, limited, irrelevant or incorrect training data, algorithmic black-box modeling methods or any different combination of the above three can cause algorithmic discrimination.Algorithmic discrimination correction in the perspective of technology governance and the establishment of algorithmic fairness in the perspective of institutional governance could be applied to solve algorithmic discrimination.Among them, the technology governance of algorithmic discrimination can start from the diversity of inclusion, transparency or accountability system.The establishment of algorithmic fairness can mainly revolve around three aspects: the algorithmic fairness values, multivariate co-governance, algorithmic literacy, to solve algorithmic discrimination.

关 键 词：智能互联网 算法歧视 算法黑箱 算法公正 算法素养 人工智能 

分 类 号：N031[自然科学总论—科学技术哲学] G301[文化科学]