Analyzing Public Sentiment on Demonetization Using SVM: A Machine Learning Approach

The Indian economy experienced significant disruption following the implementation of demonetization, a policy initiative aimed at eliminating black money, controlling inflation, and promoting financial inclusion. However, this currency ban generated widespread debate and polarized public opinion. This study analyzes public sentiment toward demonetization using social media data, specifically Twitter posts characterized by mixed sentiments, sarcasm, and nuanced linguistic expressions. We employ a PAD-SVM (Preprocessing-Analysis-Decision Support Vector Machine) approach comprising three stages: preprocessing, descriptive analysis, and prescriptive analysis. The preprocessing stage involves data cleaning, handling missing values, and feature extraction from tweet data. The descriptive analysis stage identifies key influencers and performs exploratory data analysis related to demonetization discourse. Subsequently, sentiment analysis is conducted to quantify user sentiments and assign polarity scores to individual tweets. Predictive modeling is then applied to forecast evolving public perception toward demonetization over time. This approach combines machine learning, statistical modeling, and natural language processing (NLP) techniques to process unstructured textual data and classify sentiments as positive, negative, or neutral. The integration of sentiment analysis with predictive analytics provides valuable real-time insights into public opinion dynamics and enables future trend forecasting regarding major economic policy interventions.