基于混合智能算法的网页文本分类仿真研究  

Simulation Research on Web Text categorization Based on Intelligence Algorithm

作　　者：李建军[1] 宋志章[1] 

机构地区：[1]齐齐哈尔大学,黑龙江齐齐哈尔160006

出　　处：《科技通报》2012年第6期152-154,共3页Bulletin of Science and Technology

基　　金：齐齐哈尔大学青年教师科学技术类科研启动支持计划项目(2011k-ZX01)

摘　　要：网页文本特征数常高达上万个,无用和冗余特征相当多,为提高网页文本分类精度,提出一种混合智能算法的网页文本分类方法。首先采用遗传算法对网页文本特征初步选择,然后采用蚁群算法对初步选择特征进行精细选择,最后采用K近邻算法建立文本分类器。结果表明,混合智能算法很好消除无用和冗余特征,提高了网页文本分类的精度,加快分类速度。The number of Webpage text feature is often high amounts,and has lot of useless and redundant features,in order to improve the webpage text classification accuracy,this paper presents a hybrid intelligent algorithm for webpage text classification method.Firstly,the genetic algorithm is used to select the webpage text features,and then ant colony algorithm is used to select the better features from the initial features,finally text classifier is built based on K nearest neighbor algorithm.The results show that the hybrid intelligent algorithm eliminates the useless and redundant features to improve the webpage text classification precision and increase the speed of classification.

关 键 词：网页文本 智能算法 分类方法 文本特征 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]