Tweeting Truth: Investigating False Information on Twitter Using Data Analysis and a Proposed Detection Model

The emergence of social media platforms has revolutionized information dissemination and news gathering across various fields. Despite its benefits, it has also made it easier for false information to spread rapidly, necessitating the need to identify and put a stop to it. It has also become a breeding ground for spammers who propagate fake content. The aim of this study is to classify news as authentic, false, or uncertain and to prevent the spread of misleading and inaccurate information. We aim to provide a comprehensive understanding of the current landscape of Twitter fake news detection by conducting an in-depth analysis of the latest state-of-the-art models, which includes a comparison of their effectiveness as well as an assessment of their strengths and weaknesses. Following that, we have presented the problem statement, research objectives, and proposed methodology for detecting and preventing false information. To gain a deeper understanding of the prevalence and nature of fake news on social media, we conducted exploratory data analysis on Twitter, which is currently the most widely used platform for sharing information and engaging with others. After thoroughly examining the dataset, we have found a considerable presence of false tweets, emphasizing the pressing need for more effective methods to detect and counteract misinformation. These findings have important implications for social media platforms, news organizations, and policymakers. The motivation behind this work is to fight against the spread of inaccurate information and to tackle the problem of spammers who promote the fake content. Our objective is to offer an advanced solution that utilizes state-of-the-art technology to achieve exceptional performance and effectively address this problem. The innovative hybrid deep learning method we propose integrates cutting-edge natural language processing capabilities including NoR_BeRT, Du_BRO, and MAC_BiGNet at different stages such as feature extraction, feature selection and classification, ensuring superior precision.