SweetDeal: Representing Agent Contracts with Exceptions Using Semantic Web Rules, Ontologies, and Process Descriptions

  • Authors:
  • Benjamin N. Grosof;Terrence C. Poon

  • Affiliations:
  • Information Technology at the Sloan School of Management at Massachusetts Institute of Technology;Oracle Corporation

  • Venue:
  • International Journal of Electronic Commerce
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper addresses the problem of how to represent exception-handling provisions in automated knowledge-based e-contracts. In many natural language contracts, such provisions constitute the majority of the contract volume. The paper makes three novel contributions. (1) The rule-based SweetDeal e-contracting approach is extended to incorporate ontology knowledge (specifically, ontologies about business processes), and an application scenario about late delivery in manufacturing supply-chain management shows how it can be used to represent provisions for exception-handling. The contract rules are represented in the Situated Courteous Logic Programs knowledge representation encoded in RuleML, the leading approach to Semantic Web rules. The process ontologies are represented in description logic encoded in DAML+OIL, close predecessor of W3C's OWL, the leading approach to semantic Web ontologies. This system is the first to combine these emerging semantic Web standards with one another and for a practical e-business application domain. (2) A simple new technical mechanism for the semantic Web is presented that integrates ontology knowledge with rule knowledge: Rule predicates are defined by reference to ontological classes or properties. (3) Process ontology knowledge drawn from the MIT Process Handbook, a large repository used by practical industrial process designers, is represented in DAML+OIL/OWL, and some of this knowledge is formalized and used for the e-contracting application scenario. Stated more generally, the approach in this paper provides a foundation for representing and automating deals about services--in particular, for helping to search, select, and compose semantic Web services.