中文摘要：

针对传统矩阵分解算法在处理海量数据时所面临的性能瓶颈以及大量数据的关键特征缺失问题,本文基于并行化矩阵分解算法对推荐系统效率进行提升,使用朴素贝叶斯分类算法提高推荐的准确率.首先基于TF-IDF算法构建图书评论的情感词典;然后结合朴素贝叶斯算法完善缺失关键特征的数据;最后使用并行化后的协同过滤推荐算法得到推荐结果.本文采用豆瓣读书网站上的真实图书评论数据进行实验验证,实验结果表明,分布式环境下的协同过滤推荐算法与朴素贝叶斯算法能够高效结合,显著提高推荐效率,准确度也有所提升.

英文摘要：

Aiming at solving the problem of the performance bottleneck and the key feature lack of the traditional matrix decomposition algorithm in dealing with massive data,based on the parallelization matrix decomposition algorithm,this paper improves the efficiency of recommendation systems,and the Naive Bayes classification algorithm is used to improve the recommendation accuracy.Firstly,based on the TF-IDF algorithm,the emotional dictionary of book reviews is built.Then,the key features of the data are improved with Naive Bayes algorithm.Finally,the recommended results are obtained using the parallel collaborative filtering recommendation algorithm.This paper uses the real book review data of Douban reading website to verify the experiment.The experimental results show that in the distributed environment,the recommendation algorithm based on the collaborative filtering and the Chinese emotional classifier algorithm based on Naive Bayes can be combined efficiently,the recommendation efficiency is greatly increased,and the accuracy is improved.