Measuring Bias in News Websites, Towards a Model for Personalization

Abstract

This poster briefly elucidates on a crowdsourced exploratory study demonstrating the impact of common features of news websites' design on perceived bias. Type of news website, user characteristics, and the overall design, were shown to impact on perceived bias. The poster also reports the results of a novel method to validate and extend the initial results through comparative reevaluation. These confirmed the initial results and revealed additional significant findings. Lastly, the paper proposes a model of bias for personalizing news websites.