Supporting efficient development of cognitive models at multiple skill levels

Abstract

This paper presents X-PRT, a new cognitive modeling tool supporting activities ranging from interface design to basic cognitive research. X-PRT provides a graphical model development environment for the CORE constraint-based cognitive modeling engine [7,13,21]. X-PRT comprises a novel feature set: (a) it supports the automatic generation of predictive models at multiple skill levels from a single taskspecification, (b) it supports a comprehensive set of modeling activities, and (c) it supports compositional reuse of existing cognitive/perceptual/motor skills by transforming high-level, hierarchical task descriptions into detailed performance predictions. Task hierarchies play a central role in X-PRT, serving as the organizing construct for task knowledge, the locus for compositionality, and the cognitive structures over which the learning theory is predicated. Empirical evidence supports the role of task hierarchies in routine skill acquisition.

Towards standardisation of user models for simulation and adaptation purposes

“…The constraint-based optimising reasoning engine system (CORE) [22,34,86] takes a different approach to model cognition. Instead of a rule-based system, it models cognition as a set of constraints and an objective function.…”

Leveraging characteristics of task structure to predict the cost of interruption

“…For tasks with less prescribed sequences, task models for significant variations could be created or learned and our COI model could be applied to each of them. Though this would require a large effort, it may soon be possible to develop or adapt tools to automate much of the process (e.g., see task modeling tools discussed in [12,23,35]). …”

“…However, experimental studies have shown that familiarity with a task seems to have little effect on how users perceive its hierarchical structure [38], suggesting that the mental representations for tasks remain fairly stable. Still, skilled tasks are typically performed in larger chunks [35] and COI models should consider this effect. A possible solution is to extend our current COI model to include skill level as a predictor and to encode a COI value for each skill level at each boundary or other salient point in the task.…”

Generating automated predictions of behavior strategically adapted to specific performance objectives

“…al. [24]. Several systems are under development to support applied modeling including: CogTool [11], X-PRT [24], and User Modeling Design Tool [20].…”

“…In particular, we will show how multiple strategies (and therefore behavioral predictions) may be derived from a single task specification, by varying the performance objective function. Tollinger, et al reported work on automatically generating predictions at multiple levels of skill based on a single hierarchical task description [24]. The current work adds another dimension to the space of possible models that can be generated via this approach.…”