Collaborative Research: A Computational Framework for Agent-based Contracting

The goal of this research is to investigate and develop computational methods for supporting automated negotiation of contracts in business-to-business E-commerce applications. Such processes require the ability to negotiate over contracts that have scheduling constraints, interact with a highly distributed web of suppliers with different capabilities through the completion of the contracted work, and deal with failures in contract execution. These problems are modeled using a community of self-interested agents with limited rationality, each representing a business entity or a decision-maker. Algorithms and decision strategies will be developed for effective bid construction, bid evaluation, supplier selection, execution monitoring, and renegotiation of contracts among multiple agents. A distributed computational testbed will be developed and used for simulation, and analysis. The results of this research will contribute to the development of general computational models for multi-agent contracting. The research efforts will also produce a highly distributed market infrastructure with an agent population that can be used by industry and researchers for further investigations, as well as by educators as a tool for training the next generation of e-commerce specialists. The potential payoff of the research is high, given the projected size of the business-to-business and make-to-order e-commerce.