Democracy Orientation of News Recommender Systems: A Case Study in U.S. News Platform

Abstract

As society progresses and information increases, people increasingly rely on systems that filter information and recommend relevant content. This trend is evident in the news sector, with various platforms implementing news recommendation systems. While much research has focused on improving user retention rates and click-through rates, there has been less discussion on the type of news that should be recommended. This study aims to identify the values that news recommendation systems should uphold in a democratic society and to evaluate the news recommendation systems of U.S. news portals. We selected Google, MSN, and Yahoo for evaluation, applying the democratic index measurement method from Hwang’s 2023 study. The results revealed significant differences among these websites. This study also analyzes how website algorithms and user interfaces affect news recommendation outcomes and compares these findings with Korean news portals. This research validates the applicability of previous models and suggests elements for future news recommendation systems to consider.