算法推荐新闻的技术创新与伦理困境:一个综述

Technical Innovation and Ethical Predicament of Algorithmic Recommendation News:A Summary

“算法主导下信息传播的社会影响与挑战”成为2018年度“中国十大学术热点”,算法推荐新闻也是学界关注的重要话题。大数据时代,新闻分发正在进行变革,对于算法推荐新闻的模式及其实践,从一开始就伴随是好是坏的争论。一方面认为,算法推荐新闻实现了信息超载时代用户需求信息的精准推荐,凸显了受众的主动地位;提高了内容分发效率,实现了对信息和人的精准与高效匹配,有利于增强用户黏性;为社会民意搭建了一个前所未有的社会基础平台。另一方面认为,算法推荐导致隐私侵犯、算法偏见和算法歧视;导致新闻娱乐化、低俗化与同质性;导致信息茧房和群体极化;导致价值观分化与社群区隔,不利于社会整合。对于算法推荐新闻,要深入思考数据的所有权和信息的选择权、对什么负责与对谁负责、工具理性和价值理性等几个问题,秉持宽容审慎和科学的态度进行研究。

"The social impact and challenges of information dissemination under the guidance of algorithm" has become "the top ten academic hotspots in China" in 2018, and algorithm recommendation news is also an important topic of academic concern. In the era of big data, news distribution is changing.For the model and practice of recommending news by algorithm, there is a debate about whether it is good or bad at the beginning. On the one hand, the algorithm recommendation news achieves the accurate recommendation of user demand information in the information overload era, highlighting the active status of the audience;improves the efficiency of content distribution, achieves the accurate and efficient matching of information and people, which is conducive to enhancing user stickiness;and builds an unprecedented platform for social public opinion. On the other hand, algorithmic recommendation leads to privacy violation, algorithmic bias and algorithmic discrimination, news entertainment, vulgarization and homogeneity, information cocoons and population polarization, value differentiation and community segregation, which are not conducive to social integration. For algorithmic recommendation news, we should think deeply about the ownership of data and the right to choose information, what and who are responsible for it, instrumental rationality and value rationality, and study it with a tolerant, prudent and scientific attitude.