rScholar: An Interactive Contextual User Interface to Enhance UX of Scholarly Recommender Systems

Abstract

Scholarly recommender systems attempt to reduce the number of research resources or papers presented to scholars and predict the utility of resources for their scholarly tasks. Industry practitioners and academic researchers agree that the interface of a recommender system may have as profound an effect on users’ experience as the recommender’s algorithmic performance. Despite this, little attention has been given to User Interface and Interaction Design of scholarly recommender systems. Scholarly recommender systems rarely use contextual data, such as personal and situational characteristics, that can dramatically affect the user experience (UX) and effectiveness of SRSs. This research presents rScholar, a scholarly recommender system interface that utilizes User Interface and Interaction Design adequacy indicators as well as the user contextual data to enhance the user experience. The evaluation of rScholar is performed by user studies and expert review of feedback by users and by comparison to the UI of the recommendation display of Google Scholar.