Call For Paper Volume:6 Issue:11 Nov'2019 |

Use of DSA model by considering two sides of one single review

Publication Date : 16/09/2016

DOI : 10.21884/IJMTER.2016.3042.GE9J7

Author(s) :

Mrunmayi Chaudhari , Mrunmayi , Vishal .

Volume/Issue :
Volume 3
Issue 9
(09 - 2016)

Abstract :

Many a time the Bag-of-words (BOW) is measured as important part to mold text in arithmetical machine learning way in the sentiment analysis. Earlier BOW was used for text representation. Whereas on the other hand, sometimes the presentation of BOW remains limited due to some basic deficiencies in arranging the polarity shift problem. In this, firstly we propose a model called dual sentiment analysis (DSA), to deal with this problem for sentiment classification. In this we first propose a new data extension technique by creating a sentiment reversed review for each training samples and test reviews. Based on this, we propose a dual training algorithm in which we make use of original and reversed training reviews in pairs for training classifier, and a dual prediction algorithm for ordering the test reviews by considering two sides of one single review. And in this we also widen the DSA structure from polarity (positive-negative) classification to 3-class (positive-negative- neutral) classification, by considering the neutral reviews in addition. And then we finally develop a corpus-based method to construct a pseudo-antonym dictionary, which eliminates DSAs dependency on an outer antonym dictionary for review improvement The results express the utility of DSA in showing polarity shift in sentiment classification.

No. of Downloads :



Web Design MymensinghPremium WordPress ThemesWeb Development

Use of DSA model by considering two sides of one single review

July 16, 2016