Machine Learning-Based Sentiment Analysis

Abstract

Today, almost everyone uses some type of social media, most notably Facebook, Instagram, and Twitter. Through the use of this social media platform, everyone is able to express their own perspective, ideas, and emotions on any given issue. Tweets are the bite-sized chunks of text that users of the social networking service Twitter use to share and discuss their thoughts and opinions on a wide range of topics with their followers and the public at large. Twitter is one of the most well-known social media sites. Businesses can utilize this data to improve their services, increase customer satisfaction with their products, develop more effective strategies, and aid in the decision-making processes of their employees and clients. Tweets are a significant way for businesses to learn how their products are received by consumers, and the feedback they receive is invaluable in shaping future iterations. This study article focuses mostly on the development of sentiment analysis frameworks. Sentiment analysis allows us to read between the lines of someone's written words and understand what they're really thinking. The first stage in conducting sentiment analysis is collecting and organizing a large number of tweets or user thoughts from a social media platform. Using the Natural Language Processing Toolkit and a number of other approaches, we are classifying tweets as positive, negative, or neutral. Their sentiments guided the subsequent categorization of this tweet. Using Python's flask framework, the classified findings are presented in a variety of charts (including pie charts, bar charts, and line charts) and HTML pages.