Understanding public perception is crucial for effective policymaking. Sentiment analysis is therefore a strong tool that can be employed in the study. By making use of computational techniques, this study helps to fetch public sentiments toward India’s National Education Policy (NEP) 2020 using machine learning. Public opinions are categorized into three major groups: positive, negative, and neutral. This research paper has highlighted a thorough survey on public sentiment on NEP and compared the performances of several machine learning models, viz., Logistic Regression (LR), XGBoost, Decision Tree (DT), and Bernoulli Naive Bayes (NV). The model with the highest accuracy in performance is LR, which shows a very impressive 91.22% accuracy. In this context, the LR model is more reliable in determining and studying sentiment patterns compared to other models. This study, conducted with advanced computational methods, has highlighted the possibilities of machine learning in the analysis of large-scale public sentiment. Indeed, this work underscores how technological intervention can overcome the gap that exists between people’s opinion and policy to generate more responsive governance.