A Study on the Impact Mechanism of Short Video Recommendation Algorithms on User Content Engagement Duration from a Bidirectional Construction Perspective: A Case Study of Douyin's Hot Search Trends

DOI: https://doi.org/10.62517/jnme.202610407

Author(s)

Chenxi Yang

Affiliation(s)

King's College London, London, SW10 2RL, United Kingdom

Abstract

In the context of algorithm-dominated content distribution on short-video platforms, user engagement duration serves not only as a core performance metric for platforms but also as a key indicator of bidirectional interaction between users and algorithms. Existing research predominantly focuses on the unidirectional shaping of user behavior by algorithms, lacking systematic exploration of users counter-regulation of algorithmic behavior. This study adopts the bidirectional construction theory pas its framework, employs TikTok's trending topics as the research scenario, and utilizes a mixed methods approach combining quantitative and qualitative analysis-through questionnaire surveys, in-depth interviews, and statistical analyses using SPSS and NVivo-to investigate the bidirectional influence mechanisms between short-video recommendation algorithms and user content engagement duration. The findings reveal that TikTok's collaborative filtering and traffic-weighted mechanisms significantly impact user engagement duration; user behaviors such as full viewing, liking, swiping away, and negative feedback constitute core signals driving algorithmic iteration; and users with varying algorithmic literacy exhibit distinct differences in engagement patterns and algorithmic compliance strategies. By constructing a closed-loop model of "algorithm–user–algorithm", this study addresses the unidirectional limitations of existing theories and provides practical insights for platform algorithm optimization, trending content management, and algorithmic ethics governance.

Keywords

Bidirectional Construction; Short Video Recommendation Algorithms; Engagement Duration; TikTok Trending Topics; Algorithmic Compliance; Human-Computer Interaction

References

[1] Yu Guoming, Yang Ya. The Technical Logic, Value Paradox, and Regulatory Pathways of Algorithmic Recommendation [J]. Social Sciences Frontline, 2019(05):145-152. [2] Zhou Baohua. Computational Communication Studies: Communication Research as a New Paradigm [J]. News and Writing, 2019(01):29-36. [3] Peng L. The shaping and impact of algorithmic recommendations on user behavior [J]. Editors Friend, 2020(07):45-51. [4] Zhang Shiyue, Yu Guoming. Factors influencing short video user engagement duration - A dual perspective based on algorithmic perception and emotional experience [J]. Modern Communication (Journal of China Media University), 2021,43(08):112–118. [5] Wang Run. Information Cocoon and Algorithmic Taming: A Practical Study on the Agency of Short Video Users [J]. Journalism, 2022(03):56-65. [6] Li Biao, Zheng Xiaoran. Research on the Algorithmic Logic and Dissemination Mechanism of TikTok Hot searches [J]. News and Writing, 2022(07):41-49. [7] Wu Shiwen, Chen Mengjun. Bidirectional Construction: A Study on the Interaction Mechanism Between Algorithmic Technology and User Behavior [J]. International Journal of Journalism, 2023,45(02):67–84. [8] Zhang Fang, Li Hang. Research on Short Video Algorithmic Recommendation and User Immersion Behavior from a Human-Computer Interaction Perspective [J]. Modern Communication (Journal of China Media University), 2023,45(05):123-129. [9] Yu Guoming, Pan Jiabao. Algorithmic Taming: Reconstructing Human-Computer Relationships from the Perspective of User Agency [J]. Journal of Xinjiang Normal University (Philosophy and Social Sciences Edition), 2022,43(02):112–120. [10] Ding Wei, Tian Qiu. Variable rewards and attention capture: A psychological mechanism study of short video algorithms [J]. Journalism University, 2021(06):78–89+122. [11] Zhang Zhi an, Ran Zhen. Ethical Risks of Short Video Platform Algorithms and Their Governance Pathways [J]. Peoples Forum · Academic Frontiers, 2020(20):78-85. [12] Chen Changfeng, Huang Jiasheng. Algorithmic Bias and Information cocoons: The Ethical Dilemma of Intelligent Distribution [J]. News and Writing, 2020(01):35–42. [13] Budak S, Agrawal A, El Abbadi A. Limiting the Spread of Misinformation in Social Networks [J]. ACM Transactions on Knowledge Discovery from Data, 2016,10 (4):1-29. [14] Chen J, Zhang H, He X. Attentive Collaborative Filtering: Multimedia Recommendation with Item- and Component-Level Attention [C]//Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2017:335-344. [15] Pariser E. The Filter Bubble: What the Internet Is Hiding from You [M]. New York: Penguin Press, 2011:45-68.