Analysis of Teaching Mode of Music Major Students Based on Personalized Recommendation Algorithm

With the advent of the era of big data, the teaching practice model of music major has also changed. The emergence and growth of social media allow users to act freely. Therefore, on the basis of social tagging, this paper makes full use of the personalized description information and project content information carried by tags in collaborative tagging. A personalized music recommendation algorithm based on social networks is proposed. In social networks, complex interactive relationships are formed between users and music content, as well as between users themselves. By analyzing users’ historical behavior data (such as receiving and listening to playlists, liking, and sharing), we can calculate the similarity between users. For example, if two users frequently listen to the same song or artist, then their similarity is higher. Restart type random walk is an algorithm that iteratively searches on a graph, adding a “restart” mechanism to the traditional random walk. Specifically, during each walk, the algorithm has a certain probability (usually a preset small value, such as 0.15) to return to the starting node, rather than continuing to randomly select the next adjacent node. This mechanism helps the algorithm to explore new nodes while maintaining attention to the starting node, thereby avoiding getting stuck in local optima. A personalized mobile music recommendation algorithm based on the music genome is proposed based on the similarity of the node structure of two graphs and the random walk of restart type. Based on the similarity of interests between different users, a personalized mobile music recommendation algorithm is proposed to realize the personalized service of mobile terminals.