抖音安全与信任中心在沪举办研讨会，学者共议算法向善与治理透明

Core Insights - Douyin's Safety and Trust Center held a communication meeting in Shanghai to promote algorithm transparency and the practice of "algorithm for good" through interaction between academia and the platform [1] Group 1: Event Overview - The communication meeting featured experts from various universities, including Fudan University and East China Normal University, discussing algorithm principles, content ecology, and platform responsibilities [3] - The event was divided into two parts: an introduction to Douyin's algorithm and content ecosystem, followed by a discussion addressing public concerns regarding core algorithm issues [3] Group 2: Expert Opinions - Professor Lu Ye from Fudan University highlighted the need for algorithms to cater to diverse user needs, emphasizing the importance of addressing criticism and the impact of algorithms on public discourse [3] - Professor Zhang Zhian stressed that public understanding of algorithms should consider the broader social context, advocating for problem-oriented approaches and collaboration with the public and academia [4] - Lecturer Huang Fuyuan suggested enhancing user control over recommendations to improve user experience and algorithm transparency [5] Group 3: Recommendations for Algorithm Governance - Director Fang Shishi from Shanghai Academy of Social Sciences noted that current algorithm transparency practices focus on technical explainability but lack accessibility for the general public, recommending a shift towards governance transparency [7] - Professor Zhou Zhou from Fudan University proposed incorporating more social objectives into recommendation algorithms to balance social responsibility and commercial interests [8] - Assistant Professor Wang Yilin emphasized the need for clear communication frameworks for creators, differentiating between short video and live streaming recommendation mechanisms [8] Group 4: Douyin's Commitment - Douyin's representative stated that the communication meeting is part of ongoing efforts to enhance algorithm transparency, with over 20 specialized communication meetings held in various cities this year [8] - The Douyin Safety and Trust Center launched in March, providing public access to algorithm principles and governance policies, along with an upgraded "Usage Management Assistant" for user content preference management [8]