How Fake News Affect Trust in the Output of a Machine Learning System for News Curation.

摘要: People are increasingly consuming news curated by machine learning (ML) systems. Motivated studies on algorithmic bias, this paper explores which recommendations of an curation system users trust and how is affected untrustworthy stories like fake news. In a study with 82 vocational school students background in IT, we found that able to provide ratings distinguish trustworthy quality from recommendations. However, single story combined four rated similarly as five stories. The results could be first indication benefit appearing context. also show the limitations users’ abilities rate system. We discuss implications for user experience interactive