Sentiment Analysis on Movie Reviews


  • P. Trupthi, Pandeti Madhura, Pooja Jehan, Pooja Polampalli, Aruna Kumara B


Sentimental analysis of movie reviews is basically excavating the judgements based on the existing movie analysis on the internet. This approach identifies and categorizes the subjective opinions to decide the attitude of the subject towards a particular text. These subjective opinions determine the satisfaction level of the subject. These subjective opinions also determine the success or the failure rate of the text. The polarity or the differences between these opinions are found using various pre-processing techniques i.e. filtering of data by converting the text into words, removing the stop words and negating the neutral opinions. And for the faster processing, Feature Extraction and Feature Selection techniques are used as it reduces the word count by eliminating redundant contents. The accuracy of these subjective opinions is governed by using certain classifiers such as Naive Bayes, Logistic Regression, Decision Tree, Random Forests, KNN, SVM, and Adaptive Boosting. Each classifier has a different accuracy rate processes based on the given code values. This paper focuses to find out which classifier gives the most accurate result.

Keywords:Sentiment Analysis, Polarity, Sentiments, Movie reviews, Feature extraction,Classifier, Feature selection, Accuracy.