Publication - Using ODML to Model and Design Organizations for Multi-Agent Systems

Authors: Horling, Bryan; and Lesser, Victor
Title: Using ODML to Model and Design Organizations for Multi-Agent Systems
Abstract: In this paper, we introduce a new, domain-independent organizational design representation able to model and predict the quantitative performance characteristics of agent organizations. This representation, capable of capturing a wide range of multi-agent behaviors, can support the selection of an appropriate design given a particular operational context. We demonstrate the capabilities and efficacy of this language by comparing a range of metrics predicted by an ODML model to previously obtained empirical results from a real-world system. We then outline how such models can serve as the foundation for automated organizational design process by modeling a range of organizational possibilities.
Keywords: ANTs, Organizational Design
Publication: Proceedings of the Workshop on From Organizations to Organization Oriented Programming (OOOP 05), pp. 33 - 48
Editor: Boissier, Olivier; Dignum, Virginia; Matson, Eric; and Sichman, Jaime
Date: July 2005
Sources: PDF: /Documents/bhorling/OOOP05-ODML.pdf
Reference: Horling, Bryan; and Lesser, Victor. Using ODML to Model and Design Organizations for Multi-Agent Systems. Proceedings of the Workshop on From Organizations to Organization Oriented Programming (OOOP 05), Boissier, Olivier; Dignum, Virginia; Matson, Eric; and Sichman, Jaime, ed., pp. 33-48. July 2005.
bibtex:
@inproceedings{Horling-383,
  author    = "Bryan Horling and Victor Lesser",
  title     = "{Using ODML to Model and Design Organizations for
               Multi-Agent Systems}",
  booktitle = "Proceedings of the Workshop on From Organizations
               to Organization Oriented Programming (OOOP 05)",
  editor    = "Olivier Boissier and Virginia Dignum and Eric Matson and Jaime Sichman",
  pages     = "33-48",
  month     = "July",
  year      = "2005",
  url       = "http://mas.cs.umass.edu/paper/383",
}