Keith Ransom

Keith Ransom

School of Psychology

Faculty of Health and Medical Sciences

Eligible to supervise Masters and PhD (as Co-Supervisor) - email supervisor to discuss availability.

I'm a Post Doctoral Researcher at the University of Adelaide in the School of Psychology where I collaborate with Dr. Rachel Stephens and supervise student research projects. I'm also a Postdoctoral Research Fellow at the University of Melbourne School of Psychological Sciences and member of the Computational Cognitive Science Lab led by Andrew Perfors and Charles Kemp.

My current research projects focus on two related fronts:

social meta-inference: this work seeks to extend theoretical models of human inference to models of communication and influence among individuals and groups, and to models of social consensus. This allows us to understand and predict patterns of bi-directional influence: namely, how people’s perception of the public information environment affects what they learn from information shared with them, and how patterns of inference and information transmission affect the public information environment.

explainable reasoning: Increasingly, public information sharing is heavily mediated through technology, including online social networks architectures and AI. Thus there is a growing need to understand how such mediation impacts human reasoning at scale: for example, how can people correctly calibrate their trust not only in each other, but in the technology, and in the evidence extant in the information environment. This research seeks to develop applied tools that monitor and promote calibrated trust and vigilance towards shared information, using easily comprehensible, yet high fidelity representations and explanations of data and reasoning at scale.

More broadly, my research looks at the challenges of everyday reasoning: how we learn about the world, and generalise from our experiences of it. I'm interested in how we infer and represent the structures and processes that lie beneath the data we do have and how we leverage these representations to reason beyond that data. I use a combination of computational modelling and simulation techniques as well as behavioural experiments to get at these issues in a variety of contexts and at multiple levels (individual, social group, and societal).

Keith Ransom, Rachel Stephens, Carolyn Semmler & Lewis Mitchell.
In association with: Andrew Perfors & Christopher Leckie.
Digi+ FAME program (Information Capability mission), University of Adelaide.  $99,050, 18 months.
Project: MAGPIE: Monitoring And Guarding the Public Information Environment.

Rachel Stephens, Keith Ransom, Rachel Searston, Zygmunt Szpak, Fernando Marmolejo-Ramos, Dragana Pittas.
Collaborative Research Fund, Defence Innovation Partnership.  $150,000, 12 months.
Project: Advancing SOCRETIS (“SOCial REasoning Tool & Interactive System”): An AI-enabled collaborative reasoning aid for the information environment.

Zygmunt Szpak, Wojciech Chojnacki, Rachel Stephens, Keith Ransom & Rachel Searston.
AI for Decision Making Program Round 2, Department of Defence and the Office of National Intelligence, delivered in partnership with the Defence Innovation Partnership in South Australia.  $100,000, 6 months.
Project: A tool for human-in-the-loop contextual anomaly detection

Rachel Stephens & Keith Ransom.
AI for Decision Making Program, Department of Defence and the Office of National Intelligence, delivered in partnership with the Defence Innovation Partnership in South Australia.  $20,000, 3 months.
Project tile: Testing human responses to AI fact-checking and uncertainty. Project ID: 167650398

Connect With Me
External Profiles

Other Links