Estimating Neutrality of News Articles and Reactions on Twitter

Abstract

In recent years, many people have started to browse news articles on social networking sites and to use the reactions to news articles as a reference for understanding the news. However, owing to the bias of the news articles posted on social network services (SNSs) and the reactions to them, user misunderstanding of news has become a social problem. To address this problem, based on the idea that the neutrality of news articles and reactions can be estimated and presented to users, this paper proposes a metric for estimating the neutrality of news called "popularity value." Popularity value is calculated based on the strength of the daily interest in the news topic among users who responded to the news article on Twitter and the distribution of the responding users based on the strength of their daily interest. Through evaluation experiments, we show that the proposed popularity value is effective in predicting the neutrality of news articles posted on SNSs and reactions to them.