Abstract
With recent development of technology, especially mobile devices has made the social networks accessible 24/7. Information spreading has become faster than ever, regardless of the credibility of this information. This brings unparalleled challenges in ensuring the reliability of the information. Misinformation spreading has a strong relation especially in the context of breaking news, where the information released gradual, often starting as unverified information. Automatically identifying rumors from online social media especially micro-blogging websites is an important research. Recent research in detecting rumors automatically on social networks have addressed many languages. In this article, we provide an overview of the research into rumors detection in social media which we divided into three groups: supervised based approaches, unsupervised based approaches, and hybrid approaches based on the type of the machine learning used in each approach. (C) 2018 The Authors. Published by Elsevier B.V.