Analyzing and Anticipating Cultural Transmission Pathways Through Short Video Networks

Abstract

Against the backdrop of Western media dominance, effectively disseminating Chinese culture and enhancing international understanding is increasingly important. This study introduces the gradient weighted dynamic correction (GWDC) algorithm—an adaptive optimization framework designed to improve the distribution of traditional cultural content on digital platforms. GWDC employs a dynamic time-decay factor to prioritize recent gradient information, reducing overfitting by 12% and increasing user engagement by 8–18%. The enhanced adaptive moment estimation with GWDC algorithm integrates adaptive moment estimation, improving learning efficiency by 20% on large-scale datasets. Validation on YouTube’s Chinese cultural video corpus shows that the GWDC-based model achieves an 18% higher completion rate for technical content and a 25% reduction in regional cultural bias. These innovations offer a data-driven framework for cross-cultural communication and provide actionable insights for content creators, platform developers, and policymakers.