基于词频统计的wap分类器设计与实现
首发时间:2012-10-17
摘要:移动互联网时代的到来给人们上网带来了方便,但同时由于互联网信息的多样性,人们搜索信息时往往会遇到返回的都是一些广告信息,或者是一些与自己搜索毫无关系的信息,甚至是一些不良信息。为了增强wap用户搜索体验,本文基于词频和支持向量机模型实现了一款wap资源分类器,完成了游戏、软件、视频、音频、图片、主题六大类的分类,最后通过实验结果分析验证,该分类器分类准确率在90%以上,召回率在80%以上,通过本文的wap资源分类器大大提高了用户的搜索体验。
For information in English, please click here
Design and Implementation of wap classifier based on word frequency statistics
Abstract:The arrival of the mobile Internet era has brought to us convenient, but at the same time due to the diversity of information on the Internet, people search for information is often despair because of that the search result often has the advertising information, or there is nothing information could use, even there would be some bad information. To enhance wap users search experience, a wap resources classifier which is based on word frequency and support vector machine model has completed in this paper. It complete the classification of the game software, video, audio, images, themes, and the accuracy rate of this wap classifier is more than 90%, the recall of this wap classifier is more than 80%. The wap resource classification has greatly improved the search experience for the users.
Keywords: wap classifier SVM machine learning
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于词频统计的wap分类器设计与实现
评论
全部评论0/1000