This project evaluates different NLP approaches (rule-based, unsupervised, and supervised machine learning) across three core text mining tasks: sentiment analysis using VADER and SVM, topic classification using LDA, and named entity recognition using BERT and spaCy.