Web forum mining based on user satisfaction

Author

Pokharel, Suresh ; Guha, Saikat

Author_Institution

ICT Program, Asian Inst. of Technol., Pathumthani, Thailand

fYear

2012

fDate

23-25 Nov. 2012

Firstpage

1

Lastpage

4

Abstract

Online discussion forums are popular resource for finding solutions for various problems. Solutions are found in the repositories of forums and their automated extraction is an important task. People are interested only in those forums which contain answers. Effective techniques for mining forums can significantly reduce searching and navigation time to find a solution. This paper describes a methodology for identifying answered threads from forum. To achieve this goal, question post is detected first and then the thread is classified as answered or unanswered by tracing satisfaction of the user. Based on the experiment performed, this technique showed significant improvement in the result.

Keywords

data mining; social networking (online); Web forum mining; answered threads; forum repositories; online discussion forums; question post; user satisfaction; Accuracy; Data mining; Discussion forums; Feature extraction; Message systems; Navigation; Support vector machines; Answered thread; Labeled sequential pattern; Online discussion forum; Question detection; User satisfaction;

fLanguage

English

Publisher

ieee

Conference_Titel

Internet (AH-ICI), 2012 Third Asian Himalayas International Conference on

Conference_Location

Kathmandu

ISSN

2157-0647

Print_ISBN

978-1-4673-2591-2

Type

conf

DOI

10.1109/AHICI.2012.6408454

Filename

6408454