CALL FOR CONTRIBUTIONS

The AIPS-2002 Workshop in Knowledge Engineering
Tools and Techniques for AI Planning
April 24th, 2002
Toulouse, France

Knowledge Engineering (KE) in AI Planning has been defined as the process that deals with the acquisition, validation and maintenance of planning domain models, and the selection and optimization of appropriate planning machinery to work on them. Hence, KE processes support the planning process - they comprise of the off-line, knowledge-based aspects of planning that are to do with the application being built. In particular, the acquisition of domain knowledge may involve a prolonged analysis of the application domain by knowledge engineers resulting in a specification of domain structure, dynamics and control heuristics, as well as the environmental assumptions that will determine the kind of planner to be used. The overall goal of the workshop is to develop a clearer idea of the role of KE in planning systems, and to identify challenges for future research and development.

Specific issues of the Workshop

  • the scope, role and meaning of KE in AI Planning
  • tools and techniques for the creation, verification and validation, maintenance of planning domain models and planning systems
  • the development of front ends enabling non-planning-experts to build their own planning applications
  • the role of ontologies, domain modelling languages, domain description languages, plan languages, protocols for tool integration
  • the role of machine learning in the KE process e.g. in knowledge acquisition where techniques could be utilized to induce operator descriptions (to help non-expert users) or to acquire control knowledge (for planner tuning)
  • planning engineering environments for experimental/educational purposes (e.g. for assembling planning applications, such as plugging together various domain analysis tools / planners / execution architectures)
  • the use and relevance of tools, techniques, methodologies from the existing fields of knowledge acquisition / engineering to specific applications in AI planning
  • the effect of KE ideas on the direction of AI Planning research (or how relevant is knowledge-sparse planning research to knowledge-based planning?)
  • the classification of domains via their characteristics; the creation of taxonomies of planners, and the matching of planners to domains


Workshop Format

The workshop will be a one-day event organized in conjunction with AIPS-2002 and sponsored by PLANET.

Submissions

Papers should report on original research related to the specific issues of the Workshop. Papers (maximum of 5000 words) should be submitted by email to Lee McCluskey at lee@zeus.hud.ac.uk by February 11th, 2002, in gzip postscript or in pdf. All submissions will be reviewed by at least two referees and the referees' decision will be conveyed to the first author by March 15th. Final electronic versions of papers will be expected by March 22nd, 2002 at the latest, and will be placed on this website. Accepted contributors will be expected to present their paper at the workshop, which will be published in the workshop proceedings.

Workshop Chair

Lee McCluskey
Department of Computing and Mathematical Sciences,
The University of Huddersfield, Queensgate,
Huddersfield, West Yorks HD1 3DH, UK
lee@zeus.hud.ac.uk
tel: +44 1484 472247
fax: +44 1484 421106

Programme Committee

  • Susanne Biundo, Univ of Ulm, Germany
  • Brian Drabble, CIRL, Univ of Oregon, USA
  • Maria Fox, Univ of Durham, UK
  • Peter Jarvis, SRI, California, USA
  • David Wilkins, SRI, California, USA
  • Daniel Borrajo, University of Madrid, Spain
  • Ricardo Aler Mur, University of Madrid, Spain