Understanding public sentiment using social media activities: a machine learning approach

Abstract

People use different social media platform to express their thoughts in their profiles or write comments on various aspects of world. Then a huge amount of data is generated and are stored on the internet in an unorganized, unstructured way. In this age of digital technology, million even trillion numbers of data are generated and it has become difficult to analyze those data manually. But sentiment analysis is the scientific way to extract valuable insights from the data, analyzing data and preparing those for using in various purposes. In the world, a large number of people live with different languages as their medium to communicate with others and achieve other objectives. This study was conducted to analyze sentiment of Bangla content from social media data. This study aims to analyze sentiment in two general categories: Positive and Negative. It basically defines content carries what emotions, either positive or negative of people in a particular context. This sentiment analysis approach is conducted using Machine Learning and Natural Language Processing (NLP) techniques. The outcome of this study is to develop a model that can accurately measure emotional states from social media content. This finding can have a bold effect on understanding the public opinion on various issues. From these findings, it can be valuable in real-world events in business, culture, society and more.