Research on Personalized Recommendation of Online Learning Resources for Innovation and Entrepreneurship Among College Students

To overcome the issues of high error rate in extracting interest preferences, low recommendation accuracy, and long recommendation task completion time in traditional personalized recommendation algorithms for online learning resources, a personalized recommendation method for innovation and entrepreneurship online learning resources for college students is proposed. The Flume framework is used to construct a data collection framework for capturing college students’ innovation and entrepreneurship online learning behavior, resulting in the collection of learning behavior data. The collected learning behavior data is then used to represent the features of innovation and entrepreneurship online learning resources, enabling the extraction of college students’ interest preferences. Improvements are made to the traditional collaborative filtering algorithm by incorporating the features of college students’ interest preferences, resulting in a personalized recommendation of innovation and entrepreneurship online learning resources for college students. Experimental results demonstrate that this method achieves a higher accuracy in extracting college students’ interest preferences, improves recommendation accuracy, reduces recommendation task completion time, and produces reliable recommendation results.