Analysis and Evaluation of Sentiments in Online Communities

• Main Authors: Agarwal, Anamika, Sharma, Monika
• Format: Conference Proceeding
• Language: English
• Subjects: Analytical models; Ethical Considerations; Ethics; Natural Language Processing; Oral communication; Reviews; Sentiment analysis; Social Forums; Social networking (online)
• ISSN: 2769-2884
• DOI: 10.1109/ICRITO61523.2024.10522446

In this research paper, we focus on understanding people's feelings and opinions in online social forums, like social media platforms and discussion boards. These forums are crucial for conversations and sharing information online. Our main goals are to explain what sentiment analysis is, discuss the unique challenges of analyzing sentiments in social forums, review different methods used for sentiment analysis, and see how well they work in this context. We highlight the growing importance of sentiment analysis for grasping public opinions, spotting trends, and making better decisions in the fast-paced world of social forums. We explore the difficulties in sentiment analysis here, like dealing with casual language, understanding context, and the specifics of different topics discussed. By looking at various sentiment analysis techniques, such as rule-based methods, machine learning, and combinations of both, we aim to understand which ones are most useful for social forums. We also talk about the measures used to evaluate these techniques and stress the importance of having diverse and representative data to train these models. We provide examples of how sentiment analysis can be used, like monitoring brands or understanding political discussions. Finally, we discuss where future research in sentiment analysis for social forums could go, including improving models to handle messy text, making them work better across different topics, and considering ethical issues like privacy and bias. Our paper aims to contribute to the ongoing progress in sentiment analysis, offering insights into the challenges and opportunities in understanding user feelings in online social forums. We hope this will help guide future research to develop better sentiment analysis models for these platforms, which are both effective and ethical.