Rumor Detection in Social Media Using Machine Learning Methods

Throughout history, people have always wondered whether a word or writing is real. Whether or not that word was true depends on the person who spoke it. While the word of a trusted person was respected, the word of an unreliable person was believed to be rumor. Today, with the development of social networks, the amount of information circulating on the internet has increased rapidly. However, the reliability of this data is disputable. It is an undeniable fact that there will be gossip in the environment with so many unconfirmed data. Thus, efficiently detection of rumor is an important and hot topic. In this study, the detection of rumor in online social media is modeled as a classification problem, and the success of supervised machine learning methods in real data is assessed. OneR (One Rule), Naive Bayes, ZeroR, JRip, Random Forest, Sequential Minimal Optimization, and Hoeffding Tree algorithms are applied on rumor detection problem and comprehensive evaluation is provided.