INTRODUCTION: Dance is not only able to strengthen the body but also an expression of art. It can not only express the culture of a nation or a country but also express the emotions of a country. Therefore, it is essential to utilize algorithms for the study of dance rhythm and melodic characteristics in today's world, and introduces a sentiment analysis algorithm for the study.OBJECTIVES: To disseminate our traditional dance culture, carry forward the spirit of our traditional art, enhance the creative level of our dance art, improve the current dance art in our country can not better apply the algorithm, and solve the problem that our current sentiment analysis algorithm can not be combined with art disciplines.METHODS: Use the neural network and deep learning in sentiment analysis to establish a sentiment analysis algorithm adapted; then use the sentiment analysis algorithm to calculate the in-depth filtering of the dance rhythm and melodic characteristics of the research object; finally, the heat map of the dance rhythm and melodic characteristics of the SRD is calculated according to the experiment of the algorithm.RESULTS: The core influencing factors of dance rhythm and melodic features are found to be attention mechanism and LMT through heat analysis (knowledge map); the experimental results using the sentiment analysis algorithm can be found to have a significant mediating effect on the joint enhancement of dance rhythm and melodic sense.CONCLUSION: The development of dance art not only lies in communication and integration but also combination with contemporary computer technology; using sentiment analysis algorithms can better analyze the dance rhythm and melodic characteristics; therefore, the level of algorithm application in the field of dance art should be improved.