Multi-Class Classification of Hindi Text using Machine Learning

Abstract

This study focuses on the multi-class classification of Hindi texts using ML techniques. The goal is to develop an efficient model that can accurately classify Hindi text documents into various predefined categories. We explore the preprocessing steps necessary for cleaning and preparing the text data, tokenization, stop-word removal, and stemming. Feature extraction methods such as TF-IDF and word embeddings are employed to represent the text data numerically. It includes ML algorithm like Supervised Learning algorithm (Naïve Bayes, SVM, Decision tree and Random Forest) are experimented with to determine the best-performing model. The dataset used for training and testing consists of a diverse range of Hindi texts from different domains. The experimental results provide insights into the effectiveness of different approaches and algorithms, shedding light on the challenges and opportunities of Hindi texts. Keywords: Multi-Class Classification, Hindi Texts, Machine Learning Algorithms, Natural Language Processing, Text Classification, Feature Extraction, Evaluation Metrics.