A Machine Learning-Based Approach to Analyze Public Sentiment on Indian New Education Policy 2020

Abstract

Sentiment analysis is a circular process that helps to understand the evaluation of a text. People are interested in sharing what they know about an event and reading what others say about it in reviews posted to online media outlets. Sentiment analysis (SA) can be used for this sorting purpose. From all the reviews provided by different people, SA pulls out the structured-less text reviews relating to an item survey, an event, and so on and then classifies them as either positive, negative, or neutral. Polarity categorization is another term for this. This research examines and contrasts several machine-learning approaches to SA on the Twitter dataset. Using the NEP2020 Twitter dataset, the results are compared. Results are evaluated using several criteria: accuracy, precision, and recall. Results show that logistic regression outperforms competing machine learning methods.