Sentiment Classification on Movie Reviews and Twitter: An Experimental Study of Supervised Learning Models

Abstract

Sentiment classification refers to the act of putting in for natural language processing and text mining strategies to distinguish subjective textual data. Due to the huge availability of online data that coincide with the growth of social media, there has been a big interest from researchers in sentiment analysis and its applications. In this paper, we review the state of the art to determine how the previous researches have addressed this task. we also introduce an empirical study on two annotated datasets; 25,000 IMDB movie reviews and 25,000 tweets, where we used nine supervised learning models, the next step was to implement a voting ensemble classifier using the top four models we get from the previous steps. In the end, we outline a benchmark evaluation, the results show that the ensemble classifier outperforms all the machine learning models.