Sentiment Analysis of Movie Reviews Using Machine Learning Techniques

Abstract

Movie reviews assist viewers in determining whether a film is worth their time. Sentiment analysis is the procedure of investigating digital text to calculate whether the emotional tone of a word is favorable, unfavorable, or neutral. In the proposed study we Used IMDb Dataset., Because IMDb one of the most well-known internet databases for movies and people. This gives users access to a huge and varied dataset for sentiment analysis. and make the data overwhelming numerous measures such as word clouds and text stemming methods. Natural language processing (NLP) takes employed toward develop the suggested prototype because movie comments lack grammatical structures, and experiments have been conducted to come up to the current investigation with already existing learning model. We also applied some machine learning classifiers such as Logistics Regression (LR), Multinomial Naïve Bayes (MNB), Support Vector Classifiers (SVC), Decision Tree (DT), and Random Forest (RF). In addition, the proposed approaches are 5-fold cross-validation to obtain the accuracy rate as well as hyperparameter tuning in separate classifiers to allocate the finest parameters. The applied approaches presentation was assessed to regulate “Accuracy”, “Precision”, “Recall” and “f-score”. at what time all methods were likened, “Support Vector Classifier” gives uppermost correctness of 89.41%