What Was Written vs. Who Read It: News Media Profiling Using Text Analysis and Social Media Context

Predicting the political bias and the factuality of reporting of entire news\noutlets are critical elements of media profiling, which is an understudied but\nan increasingly important research direction. The present level of\nproliferation of fake, biased, and propagandistic content online, has made it\nimpossible to fact-check every single suspicious claim, either manually or\nautomatically. Alternatively, we can profile entire news outlets and look for\nthose that are likely to publish fake or biased content. This approach makes it\npossible to detect likely "fake news" the moment they are published, by simply\nchecking the reliability of their source. From a practical perspective,\npolitical bias and factuality of reporting have a linguistic aspect but also a\nsocial context. Here, we study the impact of both, namely (i) what was written\n(i.e., what was published by the target medium, and how it describes itself on\nTwitter) vs. (ii) who read it (i.e., analyzing the readers of the target medium\non Facebook, Twitter, and YouTube). We further study (iii) what was written\nabout the target medium on Wikipedia. The evaluation results show that what was\nwritten matters most, and that putting all information sources together yields\nhuge improvements over the current state-of-the-art.\n