Research on book rating prediction based on GA-XGBoost

With the development of digital and fragmented reading, people’s demand for high-quality books is becoming stronger and stronger. The comprehensive score of books on the network platform has become the first step for people to understand the comprehensive quality of books. In order to further understand the scoring mechanism of books on the network platform and deeply explore the relationship between book scoring and other variables on the network platform, this paper first combined with relevant data and literature, and adopted Pearson coefficient correlation analysis to conduct the correlation between scoring and other variables, so as to retain and screen variables. Then, XGBoost algorithm based on genetic algorithm is used to analyze the importance of variables, so as to determine the internal logic of the network score of books, which is conducive to the efforts of all parties to improve the score of books and promote book sales.