With the rapid-fire development of the Internet and smart phone technology, a large number of short videos are participated through social platforms. Thus, videotape content analysis is a veritably important and popular work in machine literacy and artificial intelligence presently. Still, it's veritably delicate to dissect all aspects of videotape content firstly produced by large- scale druggies. How to screen out bad and illegal content from short videos published by a large number of druggies, elect high- quality videos to partake with other users, and ameliorate the quality of videotape on the distribution platform of the entire stoner is a top precedence. Grounded on this background, this paper focuses on optimizing videotape auditing to give introductory features for algorithm judgment, supporting original content and adding the distribution of new content, and strengthening homemade intervention combining algorithm recommendation with homemade recommendation. Four major aspects of the artificial training algorithm model bandy the optimization effect of artificial intelligence on the algorithm in order to give some guidance for the sustainable and healthy development of mobile short videotape reels and its impact on our Indian youth.