Automatic Music Labeling Algorithm based on Tag Depth Analysis

Abstract

With the rapid development of music social media, online music resources are rapidly increasing and music types are increasingly diversified. As an effective means to organize massive music data, rich music annotation information has become an important part of online music services. Automatic music tagging algorithm based on tag depth analysis is an automatic music tagging method that uses the concept of tag depth to classify songs. The algorithm starts with a set of songs, where each song is assigned one or more tags. The lengths of these markers are then compared and assigned to different categories. For example, if a song has three tags, it will be classified as pop / rock, dance and country music. If the song uses two tags, it will be classified as rock / pop and pop / rock, respectively. This process will continue until all songs have been classified by their tag depth and classified accordingly.