Multiagent system for construction dispute resolution (MAS-COR)

This paper develops theoretical foundation and implements technologies for generation of legal arguments based on precedent construction disputes. First, the authors simulated the process of legal discourse in construction disputes using a formal logic algorithm that is based on adversarial precedent law. In this regard: (1) facts associated with construction change order cases were factorized into binary, dimensional, and abstract factors; (2) relevance of the developed factors was associated with the disputing parties; (3) logical predicates and rules were generated based on the said factors; (4) factors were logically analyzed into distinct classifications; and (5) an 11 stage logical induction algorithm was used to show similarities, differences, strengths, and weaknesses between current and precedent construction disputes. Second, the authors created a multiagent system for construction dispute resolution (MAS-COR) that automates the developed algorithm. In this connection: (1) an agent-based role model was developed to represent the developed algorithms; (2) an agent-based role model was built to represent the developed algorithm; and (3) system implementation was carried using object-oriented programming on NetBean's integrated development environment. Using 30 previously arbitrated construction disputes, testing and validation steps were rigorously applied to assess the developed formal logic algorithm as well as the associated created agents and their integration into the MAS-COR system through syntactical debugging using theorem proving, model checking, and system testing. The results of this validation process illustrated that the system was capable of deriving significant legal arguments that help save time and effort of construction claim and dispute professionals while preparing the defense for their respective positions.