Аннотация
Short news videos have become a dominant format for digital news dissemination on social networks. Platforms such as TikTok, Douyin, YouTube, and Instagram increasingly rely on algorithmic recommendation mechanisms to distribute media content. This study examines the AI-driven creation and distribution strategies of short news videos produced by major media outlets in China and Kyrgyzstan. A dataset of publicly available short news videos published between 2023 and 2024 by CCTV, People's Daily, Kabar News Agency, and KTRK is constructed. Using natural language processing techniques and machine learning models, the study analyzes content features and evaluates their impact on distribution performance indicators, including views, shares, and engagement rates. The findings reveal significant cross-national differences in content structure, keyword density, and posting patterns, which influence algorithmic exposure and user interaction. The results contribute to understanding data-driven news production strategies in emerging digital media environments.