In 1991 Mark Weiser presented a vision of “calm” technology that adapts to its user’s lifestyle, and integrates with our everyday lives. Unfortunately, 25 years later we are as far from the realisation as it gets, thanks in a large part to our mobile devices. Mobile notifications may capture our attention anytime and anywhere. Yet, inappropriate notifications, arriving at inopportune times (e.g. at 3am), in embarrassing contexts (e.g. when in a meeting), with currently irrelevant content or from undesired contacts, can be counter productive, resulting in user frustration, increased task completion times and task errors. In this talk I will present our line of research on the identification of factors that impact user interruptibility and our solutions for automated recognition of opportune moments to interrupt. We ground our research in the cognition theory, then collect real-world data on mobile notification handling, employ machine learning to account for factors impacting interruptibility, and finally develop practical software for mobile attention management. The work includes publications from ACM UbiComp’15, ACM CHI’16 as well as the ACM UbiComp’14 best paper nominee.
Dr Veljko Pejović is an assistant professor and the vice dean for international affairs at the Faculty of Computer and Information Science, University of Ljubljana, Slovenia. His broader interests include mobile computing, HCI, resource-efficient computing, and the interaction of technology and society. Dr Pejović obtained his PhD in computer science from the University of California, Santa Barbara, USA, where he worked on mobile and wireless technologies for bringing connectivity to remote rural regions. Before joining the University of Ljubljana, Dr Pejović worked as a research fellow at the University of Birmingham, UK, investigating and developing mobile-based behavioural change interventions. He won the first prize at the Orange Data for Development Challenge 2013 for his work on using mobile phone data for epidemics modelling and containment strategy evaluation and the ACM UbiComp'14 best paper nomination for his work on mobile interruptibility inference. More details can be found at: http://lrss.fri.uni-lj.si/Veljko