基于机器学习的电影评分预测研究

Research on Movie Score Prediction Based on Machine Learning

本文依据电影是人们日常生活中重要的娱乐方式之一,用户在选择观看一部电影前,通常会想通过已观看过用户对电影的评分或是评论来了解这部电影的是否值得观看的需求。评分预测(rating prediction)在个性化推荐研究领域中可以被理解为:被用来作为预测用户对那些尚没有评价过的电影的评分的研究问题。本文工作首先对电影数据集进行数据预处理,随后重点研究了支持向量机(SVM)回归预测对电影评分进行预测,实验结果MAE的值表明支持向量机(SVM)回归预测在电影评分预测中取得较好的预测。

In this paper,the film is one of the important ways of entertainment in people's daily life.Before people choose to watch a film,they usually want to know the quality of a film through film rating or comments.In the field of personalized recommendation research,rating prediction can be understood as a research problem that is used to predict users'ratings of movies that have not been evaluated.This paper first preprocesses the movie data set,and then focuses on the support vector machine(SVM)regression prediction to predict the movie score.The experimental results show that the MAE value of support vector machine(SVM)regres⁃sion prediction achieves better prediction in the movie score prediction.