Preliminary study of attention control modeling in complex skill training environments

Heejin Lim, John Yen

Research output: Chapter in Book/Report/Conference proceedingConference contribution


In complex skill-training systems, trainees are required to master multiple skills in a limited time, which may produce a large mental workload. Increased workload often affects performance, and trainees may get distracted or overloaded during training. Attention control is a critical activity in time-sharing environments with automatic tasks, and psychologists found that better attention control strategies can develop through training. Even though attention management is a key skill has to be acquired, it has not been considered to assess as a user model content sufficiently. In this paper, we propose an approach for attention-control modeling by detecting regular behavioral patterns that potentially explain the interdependency between primary and subtask performance. We can detect trainees' attention shift between tasks by interpreting the serial episodes of behaviors that have been uncovered. As a high attention needing training domain, we used Space Fortress game in which continuous input stream of ship maneuvering and intermittent event data are the source of the user model. We found the dependencies between these heterogeneous, multi-time streams and the point of attention shift. Domain experts or training coaches can infer the trainees' attention-control skill based on the detected rules of pattern that help them to instruct desirable strategies to handle multi subtasks.

Original languageEnglish (US)
Title of host publicationAdvances in Artificial Intelligence
EditorsAhmed Y. Tawfik, Scott D. Goodwin
PublisherSpringer Verlag
Number of pages13
ISBN (Electronic)9783540220046
StatePublished - 2014
Event17th Canadian Conference on Artificial Intelligence, Canadian AI 2004 - London, Canada
Duration: May 17 2004May 19 2004

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other17th Canadian Conference on Artificial Intelligence, Canadian AI 2004

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Preliminary study of attention control modeling in complex skill training environments'. Together they form a unique fingerprint.

Cite this