The volume of user generated content produced on micro-blogging platforms has grown at an extraordinary rate, and Twitter (now X) remains one of the most widely studied sources of short, opinion rich text. Automatically determining the polarity of such text, a task commonly known as sentiment analysis or opinion mining, has therefore become central to applications in marketing, brand monitoring, public health surveillance, politics and crisis management. This paper presents a consolidated review of sentiment analysis techniques applied to Twitter data, covering lexicon based approaches, classical supervised machine learning, deep neural architectures and, more recently, transformer based and large language model driven methods reported up to 2025. Thirty three representative studies are examined and compared in terms of feature representation, classifier, dataset and reported performance, and the persistent limitations of the field — informal language, class imbalance, domain dependence, weak handling of sarcasm and limited interpretability — are identified. To ground the review in a reproducible experiment, a comparative study is also reported on a publicly available Kaggle corpus of more than 14,000 tweets collected from the KFC and McDonald’s customer challenge. Tweets are normalised and vectorised using Term Frequency–Inverse Document Frequency (TF–IDF), and three supervised classifiers — Support Vector Machine (SVM), Random Forest (RF) and Decision Tree (DT) — are trained and evaluated using accuracy, precision, recall and F1-measure. The Decision Tree classifier attains the highest accuracy of 88.51%, followed by SVM at 87.71% and RF at 86.55%, while Random Forest returns the most balanced F1-measure. The review and the experiment together indicate that carefully engineered TF–IDF pipelines remain competitive and computationally inexpensive baselines, and they motivate a set of open research directions for future work.

Kundan Lal Vishwakarma, Prof. Nagendra Patel,"A Comparative Study of Supervised Machine Learning Techniques for Twitter Sentiment Classification" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(5): Page(399-405) September - October 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.