Extracting Mental State in Adversarial Knowledge Discovery

Wednesday, 26th May 2010, 10:00 am (PDCC Discussion Room)
Speaker: Professor David Skillicorn (School of Computing, Queen's University, Canada)
Title: Extracting Mental State in Adversarial Knowledge Discovery

Adversarial settings are those where the interests of those building the models and those being modelled do not align. This includes counterterrorism, law enforcement, and fraud detection, but also increasingly more mainstream areas such as customer relationship management.

One important form of data is text. Knowledge discovery from text includes such areas as topic detection and sentiment analysis,where models are derived from the content words. Recent advances in psychology have shown the existence of an internal state channel, carried by function words, and largely invisible to humans. Because this channel is subconsciously generated, it provides a direct insight into internal mental state that is hard to spoof. I will show how we have used this to explore deception in politics, court testimony, and radical forums.


David Skillicorn is a Professor in the School of Computing at Queen's University in Kingston, Canada. His research focuses on adversarial knowledge discovery, about which he has published extensively, including his recent book "Knowledge Discovery for Counterterrorism and Law Enforcement", Chapman and Hall, 2008.