BTSVM在图书借阅平台推送信息中的应用

Application of BTSVM in pushing information of book lending platform

针对当前图书借阅平台推送信息量大、分类管理困难等问题,提出了一种基于参数优化二叉树支持向量机(BTSVM)的推送信息分类算法。采用统计分词法对文本信息进行预处理,形成高维特征集,再利用参数优化后的BTSVM算法,实现平台内各种推送信息的精确分类,从而对不同客户群体进行针对性的信息推送。实验结果表明,BTSVM算法较SVM和ELM方法具有更高的分类准确率。

In view of the current problems such as the large amount of information push and difficult classification management,an algorithm of push information classification based on parameter optimization of Binary Tree Support Vector Machine(BTSVM)is proposed.The statistical word segmentation method is adopted to preprocess the text information to form high dimensional sets.Through the parameter optimized BTSVM algorithm,the accurate classification of various push information in the platform is realized,so as to push targeted information to different customer groups.The empirical results show that BTSVM algorithm has higher classification accuracy than SVM and ELM methods.