Advancing NLP for Indian Languages: Sentiment Analysis and Chatbot Development

The research tackles the problems associated with sentiment analysis and chatbot development for Indian languages which primarily stem from language variations together with data mixing and limited dataset availability. Authors develop improved ways for sentiment analysis through dataset integration along with language-specific preprocessing such as custom Hindi stopword lists for preprocessing techniques. The research implements the Synthetic Minority Oversampling Technique for managing data imbalance. Author’s investigation utilized Logistic Regression and Support Vector Machine and Random Forest as machine learning models for testing. For the best result Support Vector Machine performed successfully. Data pre-processing techniques alongside data balancing methods improve model accuracy which promotes digital communication inclusivity for Indian languages according to the study.