User Image Modeling and Interaction Optimization Strategies in Interaction Design of Short Video Parenting Platform Construction in Colleges and Universities

Abstract

With the rapid development of science and technology, students' behavior of using short video parenting platform to watch videos in their daily life is becoming more and more popular. Therefore, the article proposes a user classification method based on the improved FRM model for college short video parenting platform, determines the weights of the four indicators in the RFMS model through the combined assignment method, and then applies the K-means clustering algorithm to analyze the users by clustering. And on this basis, a short video recommendation algorithm based on association rules is proposed aiming to improve the accuracy of short video recommendation, and its effect is verified through experiments. Finally, an interaction optimization strategy based on college short video parenting platform is proposed in combination with the basic user profile. The article analyzes the changes in the distribution of topics and the number of topic types in the recommendation list obtained from the model of this paper and the collaborative filtering recommendation model. The results show that the user-based collaborative filtering recommendation model declines relatively sharply in the three indicators. The degree of decline becomes significantly smoother at [30,60], while the overall degree of decline of the three indicators of this paper's model is relatively smooth. From this, it can be concluded that the model in this paper has obvious advantages in improving the diversity of recommendation results.