Sentiment Analysis in Indian Regional Language (Kannada)
K P Suprith Patil, Anu D, Kishan K, G Bhavyashree, Abhishek Venkatesh · 2024
Sentiment analysis, which focuses on identifying and quantifying emotions in text to capture a writer's sentiments, whether they are positive or negative, plays a pivotal role in natural language processing (NLP) and information retrieval. Existing methods predominantly use English translations and are centered on binary sentiment classification, overlooking the rich literary tradition and cultural significance of Kannada. Using techniques like tokenization, data cleansing, stop word removal, stemming, and machine learning models specifically tailored for Kannada sentiment analysis, this paper proposes a new machine learning approach for the sentiment analysis of Kannada texts. We evaluated various classifiers, such as Linear Support Vector Classifier, Logistic Regression, SGD, K-Nearest Neighbors, Multinomial Naive Bayes, and Random Forest, by employing TF-IDF for data preprocessing and feature extraction.