Multi-Source Multi-Class Fake News Detection

Proceedings of the 27th International Conference on Computational Linguistics, pages 1546–1557 Santa Fe, New Mexico, USA, August 20-26, 2018. 1546 Multi-Source Multi-Class Fake News Detection Hamid Karimi, Proteek Chandan Roy, Sari Saba-Sadiya, and Jiliang Tang Department of Computer Science and Engineering, Michigan State University {karimiha,royprote,sadiyasa,tangjili}@msu.edu Abstract Fake news spreading through media outlets poses a real threat to the trustworthiness of infor- mation and detecting fake news has attracted increasing attention in recent years. Fake news is typically written intentionally to mislead readers, which determines that fake news detection merely based on news content is tremendously challenging. Meanwhile, fake news could con- tain true evidence to mock true news and presents different degrees of fakeness, which further exacerbates the detection difficulty. On the other hand, the spread of fake news produces various types of data from different perspectives. These multiple sources provide rich contextual infor- mation about fake news and offer unprecedented opportunities for advanced fake news detection. In this paper, we study fake news detection with different degrees of fakeness by integrating multiple sources. In particular, we introduce approaches to combine information from multiple sources and to discriminate between different degrees of fakeness, and propose a Multi-source Multi-class Fake news Detection framework MMFD, which combines automated feature ex- traction, multi-source fusion and automated degrees of fakeness detection into a coherent and interpretable model. Experimental results on the real-world data demonstrate the effectiveness of the proposed framework and extensive experiments are further conducted to understand the working of the proposed framework. 1 Introduction Given its negative impacts, fake news has been identified as a global threat (Webb et al., 2016). Detecting fake news has become increasingly important and can benefit individuals and even our society in many aspects. First, people will be well-informed about events and news and their political and social activities will not be misguided. Second, despite a few recent initiatives by some social media providers like Facebook, there is no systematic fake news detection by social media platforms. Third, identifying fake news is one step toward targeting financial incentives encouraging the spreaders running their “business”. For instance, Google attempts to stop providing its ad services to fake news websites (Wingfiled et al., 2016). Fourth, people would not lose their trust in web and social media. However, fake news detection is naturally challenging especially in the era of social media. First, fake news is usually written intentionally to mislead its readers and the content of fake news is rather diverse in terms of length, topics, and styles (Shu et al., 2017a). For example, fake news in social media is short, informal and is often related to newly emerging and time-critical events. Thus, since we have not gained enough insights into the nature of fake news, hand-crafted features based on the news content are generally not sufficient (Ruchansky et al., 2017). Second, to mock true news, fake news could mix false statements with true ones. For example, fake news can cite true evidence to support a non-factual claim (Shu et …

Content truncated to 500 words