Analysis of three methods for web-based opinion mining

Author

Ma, Hai-bing ; Geng, Yi-bing ; Qiu, Jun-rui

Author_Institution

Dept. of Political Work, Armed Police Force Political Acad. of Shanghai, Shanghai, China

Volume

2

fYear

2011

fDate

10-13 July 2011

Firstpage

915

Lastpage

919

Abstract

For the purpose of measuring semantic orientation of documents, we implemented an opinion mining tool which hybrids three different methods: The first one is based on semantic patterns, which simplify the structure of the natural language syntax; the second is based on the weighted sentiment lexicon, which used as semantic feature words; and the third one is based on traditional KNN or SVM text classification method. Our experiments show that each method has its own shorts and advantages.

Keywords

Internet; classification; data mining; natural language processing; support vector machines; text analysis; KNN; SVM text classification method; Web-based opinion mining; document; measuring semantic orientation; natural language syntax; opinion mining tool; semantic feature word; semantic pattern; weighted sentiment lexicon; Algorithm design and analysis; Feature extraction; Machine learning; Pattern matching; Semantics; Support vector machines; Text categorization; Natural language processing; semantic orientation; text classification;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics (ICMLC), 2011 International Conference on

Conference_Location

Guilin

ISSN

2160-133X

Print_ISBN

978-1-4577-0305-8

Type

conf

DOI

10.1109/ICMLC.2011.6016768

Filename

6016768