Character extraction based on support vector machine

被引：0

作者：

Han, Bing ^{[1
]}

Lin, Hongfei ^{[1
]}

机构：

[1] Dalian Univ Technol, Dept Comp Sci & Engn, Dalian 116024, Peoples R China

来源：

RECENT ADVANCE OF CHINESE COMPUTING TECHNOLOGIES | 2007年

关键词：

characters; information extraction; feature extraction; bootstrapping; classification;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Our life has become increasingly close with internet. People not only want to find interesting information, but also try to find certain persons they are interested in. We defined eight kinds of characters in this paper, and realized the automatic characters extraction with the method of classification. This paper proposes four methods to select the feature words: the words around the characters, part of speech of those words, grouping characters and core feature words which are chosen by bootstrapping. In illustrations, classification is a suitable method for characters relation extraction and these methods of select feature words made the results greatly improved.

引用

页码：238 / 243

页数：6