© 2020 Elsevier Ltd Although algorithms have been widely used to deliver useful applications and services, it is unclear how users actually experience and interact with algorithm-driven services. This ambiguity is even more troubling in news recommendation algorithms, where thorny issues are complicated. This study investigates the user experience and usability of algorithms by focusing on users' cognitive process to understand how qualities/features are received and transformed into experiences and interaction. This work examines how users perceive and feel about issues in news recommendations and how they interact and engage with algorithm-recommended news. It proposes an algorithm experience model of news recommendation integrating the heuristic process of cognitive, affective, and behavioral factors. The underlying algorithm can affect in different ways the user's perception and trust of the system. The heuristic affect occurs when users' subjective feelings about transparency and accuracy act as a mental shortcut: users considered transparent and accurate systems convenient and useful. The mediating role of trust suggests that establishing algorithmic trust between users and NRS could enhance algorithm performance. The model illustrates the users' cognitive processes of perceptual judgment as well as the motivation behind user behaviors. The results highlight a link between news recommendation systems and user interaction, providing a clearer conceptualization of user-centered development and the evaluation of algorithm-based services.