Analysis of douban movies based on regression and clustering

• 1、School of Information & Electronic Engineering, Zhejiang Gongshang University, Hangzhou 310020

Abstract：Since the number of moviegoers, movie-watching habits and movie subject matter are closely related to the quality of movies, we use big data technology to analyze Douban data. All the data used are from the official website of Douban. In this paper, linear regression and heat map are used to conduct correlation analysis on the film score, film year and number of participants in Douban film data set, which has a positive effect on the analysis of influencing factors of film quality. Through collaborative filtering algorithm, multiple users\' viewing records are processed to recommend a list of the most suitable movies for each user. At the same time, we use the k-means clustering algorithm to merge many movie types into several clusters, which greatly simplifies the collaborative filtering algorithm.

Keywords： data mining Film quality The recommended list Linear regression Thermal map analysis Collaborative filtering algorithm