Expressing Collaboration And Competition Among Abductive Logic Agents

This paper presents a language for coordinating several logic-based agents capable of abductive reasoning. The system is particularly suited for solving problems with incomplete knowledge, where agents may need to make reasonable hypotheses about the domain. We defined a simple declarative language to express agent behavior, and in particular, two forms of coordination: collaboration and competition. An example in the area of medical diagnosis is presented to show the features of the language and the behavior of the proposed architecture.