ANALYSIS AND CLASSIFICATION OF SENTIMENTS USING DUAL SENTIMENT FILTRATION

COMPUTER ENGINEERING

Sentiment analysis and opinion mining are used to determine sentiment expressed in text reviews. In area of sentiment domain analysis, Bag of Words i.e. BoW is the statistical machine learning approach which suffers from some fundamental issues those occurred during the problem of polarity shifting. Therefore, DSA i.e. Dual Sentiment Analysis is proposed to overcome the problem in classification of sentiments. It consists of two phases such as, Dual training (DT) which determines original as well as reversed training reviews from learning sentiment classifiers and Dual Prediction (DP) sort/classifies the reviews by considering both negative and positive sides of single review. Dependencies of DSA are removed by implementing corpus-based methodology which construct pseudo antonym dictionary. It is used for reversion of review. Along with DSA proposed system contributes features for inconsistency classifier by syntactic construction which helps to improve system performance.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.