Adaptive Group Recommendation Algorithm for Innovation and Entrepreneurship Learning Platforms Based on Attention Mechanisms

Abstract

Artificial intelligence technology greatly promotes the development of innovation and entrepreneurship learning platform, in order to effectively guide students to utilize the existing innovation and entrepreneurship learning resources and accurately match suitable competition questions, the study firstly proposes a new adaptive grouping algorithm, and then unfolds optimization and improvement of group recommendation algorithm based on the attention. The experimental results show that the group recommendation algorithm designed for the innovation and entrepreneurship learning platform achieves a root mean square error of 0.06, a coverage rate of 0.967, and better recommendation diversity. Meanwhile, the method achieves a better mean inverse rank and average precision mean under different datasets, which meets the students' adaptability needs for the competition questions; the hit rate of the recommendation results is as high as 91.32%, and the cumulative gain of normalized discount is 66.31%, which is the best recommendation performance. The design of the study can further improve the overall quality of innovation and entrepreneurship education and stimulate students' innovation and entrepreneurship motivation.