"Modelling and measuring trust in collaborative human-machine teams"

Human users are being increasingly supported by AI and data analytics tools in a wide range of tasks. Such human-machine teams are not symmetric, since the efficiency of the machines greatly surpasses that of the humans, whereas the decisions (especially, mission-critical ones) are predominantly made by the humans. This brings to the fore the need for accurately modelling, capturing, and calibrating trust in collaborative human-machine teams, as mistrust may lead to under-utilisation of machines and over-trust - to over-reliance and potentially erroneous decisions. This talk will overview our recent work on (i) temporal trust dynamics in human-machine teams; (ii) objectively modelling humans using physiological data; and (iii) robust collaborative human-machine teams in medicine and cyber-security.

Bio: Shlomo Berkovsky is a computer scientist, with deep theoretical and applied expertise in areas related to human-centric applications of Artificial Intelligence. His original research areas include user modelling and personalised technologies, which are at the intersection of the Machine Learning and Human-Computer Interaction disciplines. Berkovsky has extensive experience in leading research teams and projects. He established the Personalised Information Delivery team at the CSIRO, led the Interactive Behaviour Analytics team at Data61, and the Data Science and Product Analytics teams at Atlassian. Currently, he leads the Precision Health stream at the Centre for Health Informatics, Macquarie University. The stream focusses on the use of Machine Learning methods to develop patient models and personalised predictions of diagnosis and care, and studies how sensors can be deployed to predict medical conditions, and how clinicians and patients interact with health technologies. Berkovsky has a sustained record of research leadership, evidenced by publications at prestigious journals and conferences, steadily increasing citation track, conference Best Paper awards, keynote talks and tutorials, strong research funding stream, and leading large-scale research activities.