Understanding the Impact of TikTok's Recommendation Algorithm on User Engagement

Abstract

The study investigates the impact of TikTok’s recommendation algorithms on content discovery and user engagement, utilizing a mixed-methods approach that integrates quantitative data analysis and qualitative interviews. The quantitative analysis involved examining a dataset of user interactions over six months, revealing that key features such as like ratios, trending hashtags, and video length significantly influence recommendation likelihood. Qualitative interviews with content creators and users provided insights into the perceived transparency and effectiveness of these recommendations. Our findings indicate that TikTok’s sophisticated blend of collaborative filtering and content-based filtering effectively personalizes content delivery, enhancing user engagement and democratizing content visibility. However, this also leads to content homogeneity and the reinforcement of echo chambers. The lack of algorithmic transparency emerged as a critical issue, affecting user trust and raising ethical concerns. Participants expressed a need for more clarity on how recommendations are generated and greater control over their content preferences. The study underscores the importance of balancing user engagement with ethical considerations. It advocates for the development of transparent and user-centric algorithms that not only engage users but also promote diverse content and ensure fair information dissemination. Future research should focus on long-term impacts of algorithmic recommendations and explore interdisciplinary approaches to enhance algorithmic accountability and transparency.