A Comparative Analysis of Machine Learning Based Sentiment Analysis

In today’s world, everyone expresses themselves through social media. The most popular platforms for sharing opinions on any issue are Twitter, Facebook, Youtube, IMDB, etc. We can examine people’s attitudes after evaluating messages, comments, responses, and reviews. Sentiment Analysis is a Natural Language Processing technique for analyzing texts and determining how people feel about them. The purpose of Sentiment Analysis is for the computer to be able to detect and express emotions. This work aimed to apply Machine Learning to discover the best accuracy for text-based sentiment Analysis. We have used two datasets in this project, one is a dataset of tweets, and the other is a dataset of movie reviews. We analyze each Machine Learning algorithm’s accuracy and find which algorithm gives the best accuracy in each dataset.