算法推荐技术应用中的意识形态安全风险与防范策略

Ideological security risks and prevention strategies in the application of algorithm recommendation technology

作为数字时代主导信息传播的新技术——算法推荐技术正在重新建构社会的思想、行为和价值。算法推荐技术主导下的信息传播影响主流意识形态安全,其“去中心化”“信息茧房”“泛娱乐化”导致主流意识形态主导地位降低、内容供给不足、价值引领被消解、主流话语权被稀释。算法推荐技术应用的价值理性与工具理性失衡、网络平台及用户总体国家安全观的理念缺位、意识形态安全风险防范的监督和责任落实不足等多重因素导致了算法推荐技术应用中的意识形态安全风险。当前算法推荐技术应用中的意识形态安全风险防范应从以下四个方面予以改进:优化算法推荐下意识形态安全的技术支撑;增进社会公众对算法推荐技术的广泛了解和意识形态安全素养,特别是提升社会公众的总体国家安全观意识;健全算法推荐运行的意识形态安全监督制度;明确算法推荐技术运行中各主体的意识形态安全责任。

As a new technology that dominates information dissemination in the digital age,algorithm recommendation technology is reconstructing the thoughts,behaviors and valuesof society.Information dissemination dominated by algorithm recommendation technology affects the security of mainstream ideology.Its"decentralization","information cocoon"and"pan-entertainment"have led to a reduction in the dominant position of mainstream ideology,insufficient content supply,dissolution of value guidance,and dilution of mainstream discourse power.Multiple factors such as the imbalance between value rationality and instrumental rationality in the application of algorithm recommendation technology,the lack of the concept of the overall national security concept of network platforms and users,and the lack of supervision and responsibility for the prevention of ideological security risks have led to ideological security risks in the application of algorithm recommendation technology.The prevention of ideological security risks in the current application of algorithm recommendation technology should be improved from the following four aspects:optimize the technical support for ideological security under algorithm recommendation;enhance the public's extensive understanding of algorithm recommendation technology and ideological security literacy,especially enhance the public's awareness of the overall national security concept;improve the ideological security supervision system of algorithm recommendation operation;clarify the ideological security responsibilities of various subjects in the operation of algorithm recommendation technology.