A New Rational Framework for Modelling Exploratory Device Learning ... but does it Fit with ACT-R?

We are investigating a new, rational framework for exploratory learning. We are applying it to model the exploratory learning of simple interactive devices — such as boxes with buttons and lights, or domestic central heating timers (Cox & Young, 2000) — although we would like to think that our framework is of considerably greater generality. (To some extent, exploration lies in the eye of the beholder, and it is possible to view many everyday cognitive activities as forms of exploration.) The framework promises to offer an account of various empirical phenomena in exploratory device learning, but — despite its rational basis — it is difficult to fit it onto ACT-R.