Analyzing Bias of Comments on Political News Articles to Facilitate Transparent Online Communities

Abstract

Comments on the news articles can affect people's perceptions and behaviors. However, little is known about how people determine the degree of bias (DoB) of comments on political news articles. To address such bias issues, current platforms of news articles offer criteria to sort multiple comments on the news articles. However, little is known about whether the DoB of comments is reduced when publishers offer various criteria for sorting comments. We conducted surveys to identify how people determine DoB of comments on the news article, and how bias varies depending on how comments are sorted. The findings of this study revealed that there was a significant difference among the DoB by comments. Future work remains to develop an algorithm generating unbiased comments by using existing comments on political news articles and their DoB.