SherLock: A CNN, RNN-LSTM Based Mobile Platform for Fact- Checking on Social Media

Abstract

Abstract: Today, false news is easily created and distributed across many social media platforms. Due to that, people find it difficult to choose between right and wrong information on those platforms. Therefore, a strong need emerges to develop a factchecking platform to overcome this problem. Fact-checking means the process of verifying information. A CNN, RNN-LSTM based mobile solution has proposed from this study to verify information on social media including many features. CNN, RNN-LSTM based hybrid model ables to capture the high-level features and long-term dependencies from the input text. Some of the features of the mobile application includes fact-checking, daily news updates, news reporting and social media trends etc. The mobile solution is developed using Flutter as the front-end framework and Firebase as the back-end framework including REST APIs to gather daily news articles. The hybrid model achieved a 92% accuracy when checking the information circulating on social media.