In this paper, we describe a high-level behavior representation language (Herbal) and report new work regarding Herbal’s ACT-R compiler. This work suggests that Herbal reduces model development time by a factor of 10 when compared to working directly in Soar, ACT-R, or Jess. We then introduce a large ACT-R model (541 rules) that we generated in approximately 8 hours. We fit the model to learning data. The comparison indicates that humans performing spreadsheet tasks appeared to start with some expertise. The comparison also suggests that ACT-R, when processing tasks consisting of hundreds of unique memory elements over times spans of twenty to forty minutes, may have problems accurately representing the learning rates of humans. In addition, our study indicates that the spacing between learning sessions has significant effects that may impact the modeling of memory decay in ACT-R.

Developing intelligent agents and cognitive models is a complex software engineering activity. This article shows how tools to create intelligent agents can be improved by taking advantage of established software engineering principles such as high-level languages, maintenance-oriented development environments, and software reuse. We describe how these principles have been realized in the Herbal integrated development environment, a collection of tools that allows agent developers to exploit modern software engineering principles.