2019 Vol.2 Oct. No.5

• Title: Multimedia music trend forecasting research
• Name: Femida Judith
• Company: University of Lausanne
• Abstract:

  In order to solve the inaccurate prediction results in traditional prediction methods, a multimedia music trend forecasting method is proposed and designed. According to the statistical principle, the music on-demand rate mining model is constructed. Combined with the change of music on-demand rate, a rapid clustering model of music data is established. According to the clustering characteristics of music data, an analysis model of multimedia music trends is constructed. The Markov distance principle is used to combine the analysis results of music trends to establish a predictive algorithm model for music trends. The prediction of music trends is realized by the above four models. The prediction experiment of music popular trend in the VOD system shows that the prediction accuracy of the multimedia prediction method is 26.71% higher than that of the traditional method, and it is effective.

• Keyword: multimedia; music trend; on-demand rate; prediction method; music rapid clustering; algorithm model;
• DOI: 10.12250/jpciams2019050541.Multimedia music trend forecasting research[J]. Computer Informatization and Mechanical System, 2019, vol. 2, pp. 99-105.
• Citation form: Femida Judith

.