文献类型：期刊文献

英文题名：IMPROVED WEBPAGE CLASSIFICATION TECHNOLOGY BASED ON FEEDFORWARD BACKPROPAGATION NEURAL NETWORK

作者：Mu, Ruihui[1,2];Zeng, Xiaoqin[2]

第一作者：Mu, Ruihui;穆瑞辉

通讯作者：Mu, RH[1];Mu, RH[2]

机构：[1]Xinxiang Univ, Coll Comp & Informat Engn, Xinxiang 453000, Henan, Peoples R China;[2]Hohai Univ, Coll Comp & Informat, Nanjing 210098, Jiangsu, Peoples R China

第一机构：新乡学院计算机与信息工程学院

通讯机构：[1]corresponding author), Xinxiang Univ, Coll Comp & Informat Engn, Xinxiang 453000, Henan, Peoples R China;[2]corresponding author), Hohai Univ, Coll Comp & Informat, Nanjing 210098, Jiangsu, Peoples R China.|[1107118]新乡学院计算机与信息工程学院;[11071]新乡学院;

年份：2018

卷号：71

期号：9

起止页码：1236-+

外文期刊名：COMPTES RENDUS DE L ACADEMIE BULGARE DES SCIENCES

收录：;Scopus(收录号：2-s2.0-85055188699);WOS:【SCI-EXPANDED(收录号:WOS:000451496900011)】；

基金：This research was partially supported by the National Key Research and Development Plan Key Projects of China under Grant No 2017YFC0405800.

语种：英文

外文关键词：backpropagation; algorithm; dataset; feedforward; neural network

摘要：With the increasing number of web pages in the Internet, the problem of automatically classifying web page directories has become more and more pressing. This paper proposes an improved webpage classification method based on weighted terminology of words to achieve automatic and efficient classification of webpages. The method uses a set of functions to classify a web document. The improved weighting technique based on terminology as proposed in this paper has obvious advantages for feature selection and feature extraction. It can greatly reduce the large size of the classifier and improve the efficiency of web page classification. The algorithm proposed in this paper was run and tested on a benchmark data set. Test results show that compared to most of the existing term weighting techniques this technology is more accurate and has other advantages.