Hybrid attribute based sentiment classification of online reviews for consumer intelligence

2018 ◽  
Vol 49 (1) ◽  
pp. 137-149 ◽  
Author(s):  
Barkha Bansal ◽  
Sangeet Srivastava

This paper discusses an efficient algorithm for sentiment classification of online text reviews posted in social networking sites and blogs which are mostly in unstructured and ungrammatical in nature. Model proposed in this paper utilizes support vector machine supervised learning algorithm and fuzzy inference system for enhancing the degree of sentiment polarity of text reviews and providing multilevel polarity categories. Model is also able to predict degree of sentiment polarity of online reviews. The model accuracy is validated on twitter data set and compared with another earlier model.


Sign in / Sign up

Export Citation Format

Share Document