Scheduled Speakers

Donald Stuss

Matthew Norton

Shiva Amiri

Igor Jurisica

Brendan Johns

Khaled El Emam

Peter Snelling

Dr. Brendan T. Johns, NSERC Post-Doctoral Fellow

Brendan Johns Brendan Johns is currently an NSERC post-doctoral fellow in the Department of Psychology at Queen’s University. He obtained his Ph.D. in 2012 from the Department of Psychological and Brain Sciences at Indiana University. He studies knowledge-acquisition processes and his work exploits large scale computational models of language and memory. He has published on a diverse range of topics, including semantic memory (knowledge), recognition memory (the process by which we recognize a previously studied item), perceptual inference, language processing, and semantic deficits in Alzheimer’s disease. He has won multiple awards for his research, including the Marr prize from the Cognitive Science Society, and the Castellan award from the Society for Computers in Psychology.


Linking Brain and Behaviour using Large Scale Computational Models of Knowledge


Recent research within computational cognitive science has led to the construction of large scale knowledge acquisition models that are capable of deriving rich semantic representations of words using cognitively-informed natural language processing techniques. These models have been proven successful at accounting for a large variety of human behaviour. However, the potential of these models is untapped, as they offer an efficient technique to analyze large amounts of clinical data, such as clinical diagnosis tasks, and relate them to other forms of data, such as verbal descriptions of a disease course or MRI scans. To demonstrate the power of these techniques, I will show how the models capture ongoing changes in memory in patients who are developing Alzheimer’s disease, even when behavioural variables are unable to do so. This line of research suggests that knowledge acquisition models provide an enticing opportunity for automated real-time monitoring, analysis, and diagnosis of verbally-based clinical data.