Machine reading for relationships between autism research and the media

Autism represents a broad group of developmental disorders characterized by impaired social interactions. It has attracted widespread concern in academia and media. The reported causes of autism have an enormous impact on the people's opinion about autistic patients and various treatments from scientific literature and media news. In this study, we use open information extraction (OIE), a machine reading technique, to automatically extract causal relationships from autism researches and news reports. We explored the differences between the scientific literature and news about the reported etiologies with respect to autism. As a result, we find that scientific papers clearly emphasize physical and biological factors, while the media emphasizes social factors as the perceived cause of autism.