With the rapid development of the Internet, traditional teaching methods have been unable to meet the learning needs of college teachers and students. First, we use network crawler technology to collect research data, complete data preprocessing work through a series of operations such as word splitting, deactivation of words, feature extraction, etc., and input the processed data into polynomial plain Bayesian classifier for training to realize the classification and analysis of music emotion features. On this premise, with the help of similarity algorithm and K nearest neighbor algorithm, the music course content recommendation algorithm is constructed. With the support of this paper’s algorithm and related development software, the design of music course content recommendation system in colleges and universities is completed, and the system is empirically analyzed. Compared with other systems, the real-time update delay and real-time recommendation delay of this paper’s system are shorter, the update delay is less than 1000ms, and the corresponding recommendation delay is less than 500, which verifies that this paper’s system has excellent operational performance, can bring students and teachers a comfortable experience, and promote the development of intelligent music teaching in colleges and universities.
