Abstract:
"The primary goal of this research endeavor is to develop a system that can effectively identify
fake news written in Romanized Sinhala by utilizing a bi-directional LSTM model. By
leveraging deep learning techniques and natural language processing, the project seeks to
enhance the precision and effectiveness of detecting deceptive information across various
online platforms. While bi-directional LSTM models have demonstrated promising outcomes
in various fields, this study focuses on their application specifically to Romanized Sinhala fake
news detection. By examining the potential advantages and optimizations of machine learning
techniques in this area, the project seeks to contribute to the advancement of reliable
information dissemination and combat the proliferation of false information."
