Hossain, Muhammad IqbalAhmed, Md FaisalAlam, MahabubulDipta, Nabil FaieazIslam, Md RakibulWahid, Aunanna BinteRokti, Tanmin Alam2025-06-012025-06-0120252025ID 20101105ID 20201180ID 24341314ID 20201167ID 21101051http://hdl.handle.net/10361/26020Cataloged from PDF version of thesis.Includes bibliographical references (pages 56-58).This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science, 2025.Misleading information is mostly intended to trick and harm people is known as fake news. It is mostly designed to harm a person or organization’s reputation. In this era of information technology, the rapid dissemination of news through social media platforms such as Facebook, Instagram, Twitter and so on has become an integral part of our daily lives. It has also been seen how devastating consequences can be due to the spread of fake news. Rumors can create unimaginable havoc in real life. There are already some existing works that need more efficiency. To detect fake news from social media accu- rately with most accuracy, we intend to propose a deep learning-based approach. We have utilized two deep learning models LSTM and DNN, a hybrid model of DNN and LSTM, three advanced deep learning architectures such as DistilBERT+LSTM+DNN, DistilBERT+GRU+DNN, BERT+GRU+DNN and our TruthForge model to evaluate our work. Through our work, we aim to provide a robust tool for identifying and differentiat- ing between true and false news, thus advancing the accuracy of news verification.58 pagesenBRAC University theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission.Fake newsDeep learningSocial mediaTechnologyRumorsLSTMDNNGRUBERTPreprocessingFine-tuningHybridCognitive learning theoryThesis