In an age of personalisation, the deliberate pursuit of serendipity within news recommendation architecture serves the important functions of stimulating more engaged and receptive readerships by satisfying an ingrained desire for novelty and surprise, while simultaneously guarding against negative algorithmic tendencies such as overspecialisation and popularity bias.

As a PhD candidate in computational communication science, my research centres on the design and contextual specification of real-time news recommender systems and their corresponding capacities to foster serendipitous news consumption experiences.

Through app-based experiments, this work explores the extent to which we may import recent innovations in social media recommender system design (e.g., short-term interest modelling) and implementation (e.g., implicit user signalling) to produce more dynamic and responsive news environments, ultimately shedding light on the impacts of shifting algorithmic intermediation on resultant patterns of news consumption and user experience.

This work is performed as part of the larger NEWSFLOWS project, in which we study the effects of feedback loops within the journalistic environment on citizens’ attitudes and beliefs.