基于MKL-SVM的网络购物评论分类方法

购物网站在线评论系统收集了大量的顾客评价。支持向量机（SVM）是一种有效的文本分类方法,可以用于跟踪和管理顾客意见,但是SVM存在训练收敛速度慢,分类精度难以提高等缺点。文章提出利用异质核函数性的不同特性,解决支持向量机（SVM）数据泛化学习能力弱的问题,提高SVM的分类精度,通过对顾客购物评论进行分类,解决购物网站海量顾客评论分析的问题,帮助企业及时进行顾客反馈,提升服务水平。

A classification method of online reviews based on MKL-SVM

An online shopping website accumulates a large number of customer reviews for goods and enterprise services. Support Vector Machine （SVM） is an efficient classification method and can be used to track and manage customer reviews. But SVM has some weaknesses, for example, its slow speed of training convergence and uneasy raise of classification accuracy. The author presents the use of heterogeneous nuclear function of different characteristics, which may resolve SVM＇s problem of weak generalization ability to learn and improve SVM classification accuracy. Through classification of online customer reviews, shopping sites may resolve the issues of critical analysis of mass data, and effectively help enterprises to improve service levels.