Internet news sites have introduced numerous popularity-based recommendation techniques to solve the problem of an overload of information and enhance the efficiency of news users’ searches. However, news recommended by popularity-based mechanisms leans more toward soft subjects (i.e., sports and entertainment) and away from what is typically considered hard news (i.e., political, national, and economic news). One way to supplement the weakness of popularity-based recommendation systems is to develop personalized news recommendation systems. Personalization based on user history will allow users to access customized information more easily and quickly while mitigating information overload. However, if personalization is based only on users’ history, users will be exposed even more to the news that appeals to them and may not access topics that their histories say they have no interest in. At present, most studies of personalized recommendation systems have focused on either algorithms or on the positive effects of personalization; there is no empirical research on alleviation of the negative effects of personalization. This study suggests two information technology (IT) artifacts that may help users to evaluate their information consumption behavior and thus help them tailor this behavior to their needs and goals.

Abstract Only