A Chinese News Archiving Approach Based on Incremental Naive Bayesian Learning

被引：0

作者：

Wang, Baofeng ^{[1
]}

Long, Shun ^{[2
]}

机构：

[1] Jinan Univ, Dept Japanese Studies, Guangzhou 510632, Guangdong, Peoples R China

[2] Jinan Univ, Dept Comp Sci, Guangzhou 510632, Guangdong, Peoples R China

来源：

2014 2ND INTERNATIONAL CONFERENCE ON ECONOMIC, BUSINESS MANAGEMENT AND EDUCATION INNOVATION (EBMEI 2014), VOL 39 | 2014年 / 39卷

关键词：

News archiving; Feature extraction; Naive Bayes classifier; Incremental learning;

D O I：

暂无

中图分类号：

C [社会科学总论];

学科分类号：

03 ; 0303 ;

摘要：

Although various approaches have been proposed to solve the demanding news archiving problem in the past few decades, few provides results satisfying enough to be deployed in practice. We propose in this paper an approach based on both incremental learning and naive Bayesian learning. An improved feature extraction approach is developed, before a Naive Bayes classifier be modified in consideration of efficiency in practice. Preliminary experiments carried out on Nanfang Daily news archive show that these above modifications significantly improve the efficiency and solve the overflow problem in news archiving.

引用

页码：134 / 139

页数：6