Title: Analysing social media sentiment: unravelling the trichotomy of positive, negative, and neutral sentiments in user comments

Authors: Reddy Sowmya Vangumalla; Yoonsuk Choi

Addresses: Department of Electrical and Computer Engineering, California State University, Fullerton, CA 92831, USA ' Department of Electrical and Computer Engineering, California State University, Fullerton, CA 92831, USA

Abstract: This study explores sentiment analysis of Twitter comments, focusing on neutral, negative, and positive attitudes. By applying advanced techniques such as feature engineering, data preprocessing, and machine learning, we aim to derive actionable insights. Our approach involves setting project goals, selecting data sources, and establishing infrastructure for analysis. After preprocessing, we utilise support vector machines (SVMs) for classification and evaluate the model with metrics like accuracy, precision, recall, and F1-score. Visualisation tools, including ROC curves and confusion matrices, help interpret the results. We discuss the limitations and suggest future research to enhance performance and address data quality issues.

Keywords: data analysis; decision-making; feature engineering; machine learning; sentiment analysis; social media; SVM; support vector machines; twitter; text preprocessing.

DOI: 10.1504/IJDATS.2025.150913

International Journal of Data Analysis Techniques and Strategies, 2025 Vol.17 No.4, pp.371 - 387

Received: 05 Jun 2024
Accepted: 25 Aug 2024
Published online: 05 Jan 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article