Human Cognition and Applied Decision Making Research Centre

Directors

Dr. Simon Dennis, Associate Professor Michael Lee and Dr. Daniel Navarro
School of Psychology
Faculty of Health Sciences, the University of Adelaide

A brief outline of some of our research is presented below:

Knowledge Representation

Representing and using knowledge about the environment is a crucial task facing any organism. One line of research in our unit investigates how people represent information, and how these representations can be learned (e.g., Lee, 2002; Navarro & Lee, 2003, 2004). This basic research has been applied to develop novel approaches to data visualisation, particularly in collaboration with the Defence Science and Technology Organisation (Lee, Butavicius, & Reilly, 2003).

Language

Language research in our unit has focused on sentence parsing processes and the extraction of propositional information from text (e.g., Dennis, 2004). Not only does this work provide basic insights into fundamental cognitive processes, but it has been applied in Defence and Education contexts. For example, systems for assisting vocabulary learning and for automated essay marking have been developed for deployment in Colorado schools. In addition, a prototype question-answering system has been developed, based on new theoretical models of the relational and semantic structure in language.

Decision-Making

Viewing human decision-making as a process of evidence-accumulation provides an elegant theoretical account for a range of phenomena. A basic theoretical contribution (Lee & Cummins 2004) showed how this approach can unify Bayesian and heuristic approaches to understanding human decision-making. We have applied evidence-accumulation models to develop novel and effective algorithms for text classification (Lee & Corlett 2003), and adaptive and scalable algorithms for prioritising e-mail (Lee, Chandrasena, & Navarro 2002).

Categorisation

Past theoretical work developed category-learning models able to accommodate sophisticated knowledge structures (Lee & Navarro, 2002; Navarro, in press). Our current research considers how human categorisation might approximate rational statistical ideals, and how different models of categorisation mimic one another.

Memory

Investigations of the structure and function of human memory span work on short-term memory, episodic memory, and semantic memory (e.g., Dennis & Humphreys, 2001). Applications include knowledge management systems and adapting user profiles to changing environments. For example, a recent collaborative research project with a telecommunications service provider aims to use models of human memory to understand people’s use of mobile phone applications.

Statistical Methods

Cognitive science progresses through the development of computational models of human cognitive processes. Accordingly, it is important to use modern statistical methods (e.g., Bayesian and information theoretic methods) for evaluating and comparing competing models. We do both basic research developing new model evaluation methods, such as “landscaping” and “parameter space partitioning” (Navarro, Pitt & Myung, 2004), and apply these sorts of methods to evaluate cognitive models in laboratory and applied contexts (e.g., Navarro & Lee, 2003, in press; Navarro, 2004).

References

Dennis, S. (2004). An unsupervised method for the extraction of propositional information from text. Proceedings of the National Academy of Sciences, 101, 5206-5213.

Dennis, S. & Humphreys, M. S. (2001). A context noise model of episodic word recognition. Psychological Review, 108, 452-477.

Lee, M. D. (2002). A simple method for generating additive clustering models with limited complexity. Machine Learning, 49, 39-58.

Lee, M.D., Chandrasena, L.H. & Navarro, D.J. (2002). Using cognitive decision models to prioritize e-mails. In W.G. Gray & C. D. Schunn, (Eds.), Proceedings of the 24th Annual Conference of the Cognitive Science Society, pp. 478-483. Mahwah, NJ: Erlbaum.

Lee, M.D., Butavicius, M.A., & Reilly, R.E. (2003). Visualizations of binary data: A comparative evaluation. International Journal of Human-Computer Studies, 59 (5), 569-602.

Lee, M. D. & Corlett, E. Y. (2003). Sequential sampling models of human text classification. Cognitive Science, 27, 159-193.

Lee, M.D., & Cummins, T.D.R. (2004). Evidence accumulation in decision making: Unifying the ‘take the best’ and ‘rational’ models. Psychonomic Bulletin & Review, 11, 343-352.

Lee, M. D., & Navarro, D. J. (2002). Extending the ALCOVE model of category learning to featural stimulus domains. Psychonomic Bulletin & Review, 9, 43-58.

Navarro, D. J. (in press). Analyzing the RULEX model of category learning. To appear in

Journal of Mathematical Psychology .

Navarro, D. J. (2004). A note on the applied use of MDL appproximations. Neural Computation, 16, 1763-1768.

Navarro, D. J. & Lee, M. D. (in press). An application of minimum description length clustering to partitioning learning curves. To appear in the proceedings of The 2005 IEEE International Symposium on Information Theory.

Navarro, D. J. & Lee, M. D. (2004). Common and distinctive features in stimulus representation: A modified version of the contrast model. Psychonomic Bulletin & Review, 11, 961–974.

Navarro, D. J. & Lee, M. D. (2003). Combining dimensions and features in similarity-based representations. In S. Becker, S. Thrun, and K. Obermayer (Eds.), Advances in Neural Information Processing Systems, 15 (pp. 67-74). Cambridge, MA: MIT Press.

Navarro, D. J., Pitt, M. A. & Myung, I. J. (2004). Assessing the distinguishability of models and the informativeness of data. Cognitive Psychology, 49, 47-84.