We investigated how representations of the Syrian Refugee Crisis (2011-2021) differ across US partisan news outlets. We analyzed 47,388 articles from the online US media about the Crisis to detail differences in reporting between left- and right-leaning media. We used several natural language processing techniques to quantify differences in representations across partisan outlets and time. We used large-scale language models for single word translation, question answering, and cloze tests to document that the left-leaning media tended to frame refugees as child victims or people seeking a better future, while right-leaning media cast refugees simply as people or Islamic terrorists. We use econometric models to understand the narrative shift across partisan media. For certain events in the Crisis, we find that right-media tended to have lower sentiment and a higher offensive and hate speech score compared to other media. Based on our results, we provide several recommendations. Stakeholders may utilize our findings to intervene around refugee representations, and design communications campaigns that improve they way society sees refugees and possibly aid refugee outcomes.
Navin Kumar is an Postdoctoral Associate at the Yale School of Medicine. Navin completed a PhD at the Yale University Sociology Department, advised by Nicholas Christakis. His research harnesses the power of social computing, machine learning, and natural language analysis, grounded with insights from the social, behavioral, and clinical fields. He evaluates the developed computational approaches using diverse human-centered methods, deriving insights to improve health outcomes. Recent work details COVID-19 misinformation on Reddit following vaccine scientific announcements. Similar work centers on how vape/e-cigarette users produce and consume health misinformation.
Contact Heather.branch@yale.edu to register
Sponsoring Organization(s)
ISP