DocumentCode
442056
Title
Automatic text categorization based on angle distribution
Author
Liu, Tao ; Guo, Jun
Author_Institution
Sch. of Inf. Eng., Beijing Univ. of Posts & Telecommun., China
Volume
6
fYear
2005
fDate
18-21 Aug. 2005
Firstpage
3797
Abstract
In order to improve the performance of Chinese text categorization, a new Chinese text categorization method based on angle distribution is presented. The new method describes the text with a more precise model and proposed a new categorization algorithm by employing angle distribution. Simulation results on open Chinese text collection show that the precision and recall of most classes have been increased with reference to the classical method, and the macro average of precision and recall are both about 72 percents, which certificating the effectiveness and feasibility of the angle distribution-based algorithm.
Keywords
indexing; information retrieval; text analysis; angle distribution; automatic Chinese text categorization method; text similarity; Classification tree analysis; Content based retrieval; Distributed computing; Information retrieval; Machine learning; Machine learning algorithms; Nearest neighbor searches; Regression tree analysis; Text categorization; Web sites; Text categorization; angle distribution; similarity;
fLanguage
English
Publisher
ieee
Conference_Titel
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location
Guangzhou, China
Print_ISBN
0-7803-9091-1
Type
conf
DOI
10.1109/ICMLC.2005.1527601
Filename
1527601
Link To Document