Sentimental Analysis of Movie Reviews using Soft Voting Ensemble-based Machine Learning
Ali Athar, Sikandar Ali, Muhammad Mohsan Sheeraz, Subrata Bhattachariee, Hee‐Cheol Kim
Abstract
Sentimental analysis helps to classify a subject's sentiments (e.g., positive, negative, or neutral) automatically towards a specific topic, product, news, or any movie. Machine learning is a powerful technique of artificial intelligence (AI) to control the increasing demand for accurate sentimental analysis. The analysis of sentiment on social networks, such as Facebook or Twitter, has become a powerful source of learning about the user's opinion and it has a wide range of applications in the same field. However, the accuracy and efficiency of sentimental analysis are being impeded by different challenges faced in the field of Natural language processing (NLP). In this paper, we have proposed a state-of-the-art soft voting ensemble (SVE) approach to perform sentimental analysis of movie reviews. Five different well-known machine learning (ML) classifiers have been used for this purpose, namely Logistic Regression (LR), Naïve Bayes (NB), XGBoost (XGB), Random Forest (RF), and Multilayer Perceptron (MLP). Our proposed ensemble approach outperformed all other classifiers by giving an overall accuracy, precision, recall, and f1-score of 89.9%, 90.0%, 90.0%, and 90.0%, respectively.