Amharic Fake News Detection on Social Media Using Feature Fusion

These days, many people use social media as a source of information and medium of communication due to its easy to access, fast to disseminate and low cost platform. However, it also enables the wide propagation of fake news which causes economic, political, and social crises to the society. As a result, many researchers have been working towards detecting fake news. Most of the researches concerned on linguistic analysis of news content to identify its credibility, however fake news is also written intentionally to mislead users by mimicking true news. Beside this, Amharic is one of the under-resourced language that suffer from the benefits of fake news detection. To overcome the problem of fake news using content feature and under-resourced language, this study uses a feature fusion of linguistic and social context feature of the publisher information to detect Amharic fake news. The fusion-based experiment shows at least 94.13% and at most 98.7% relative error reduction over the content-based approaches.