Argumentation for Secure and Verifiable Agent Negotiation

Members

• Jamal Bentahar, Mohamed El-Menshawy, Maziar Gomrokchi, Babak Khosravifar, Ahmed Moazin & Wei Wan (Concordia University, Canada)
• Mohamed Mbarki, (Laval University, Canada)

Overview

The aim of this project is to investigate the use of argumentation theory, game theory, logic-based reasoning, formal verification techniques, and Web service standards for both improving agents’ negotiation abilities and developing a secure and verifiable computational environment for negotiating agents. This will contribute to the advancement of knowledge in the domain of autonomic computing by proposing a new and rigorous computational model for efficient, secure, and verifiable negotiation. Furthermore, the problem of resource negotiation via communication in a distributed setting is core to a number of applications, particularly the emerging semantic grid computing-based applications such as e-science and e-business. Grid computing is a type of parallel and distributed system that enables the sharing, selection, and aggregation of geographically distributed autonomous resources and services. This project aims to contribute to the advancement of research in this domain by applying the new model to such applications through concrete application scenarios serving as test-beds for the program using the modern developments in the domain of Web services.

Although negotiation issues in autonomic computing have been recently addressed by some researchers, the efficiency of such a negotiation is yet an open problem. In multi-agent systems, where environments are dynamic, agents should not only be able to negotiate resources and services, but they should be efficient in their negotiations by optimizing their choices. The relation between agent negotiation efficiency and the computational complexity should be addressed in order to build advanced negotiation environments. This is because negotiation efficiency requires the use of several computational resources that should be optimized in distributed settings.

Methodology

• Architecture of Reputation-Embedded Web Services Communities
  This architecture is designed to maintain the reputation of the communities of Web services. Here we assume that each web service has is associated with a community and do not function alone. The different components of the architecture are, User agent, Master agent, Provider agent, Extended UDDI, and Reputation system.

• Reputation Model
  This model is used by users and providers to estimate the reputation of a community and is based on responsiveness, inDemand, satisfaction and time recency.

• Logging Mechanism
  Reliable mechanism capable of managing malicious acts of agents.

• Experimental Results
  In our  implemented prototype, communities of Web services  are composed of distributed web services agents. The agent reasoning capabilities are implemented as Java modules. The testbed environment is populated with two agent types:

1. Service provider agents that are known as web services and gathered in a community.
2. User agents that are seeking for the best service provided by a web service.

The experiments consists of a series of empirical experiments tailored to show the adjustment of the communities’ reputation level.

Deliverables

• Babak Khosravifar, Jamal Bentahar, Ahmad Moazin, and Philippe Thiran (2010). Analyzing Communities of Web Services Using Incentives. In International Journal of Web Services Research (IJWSR), Volume 7 Number 3, pp. xxx-xxx, IGI Global.
• Babak Khosravifar, Jamal Bentahar, and Ahmad Moazin (2010). Analyzing the Relationships between some Parameters of Web Services Reputation. In Proceedings of the 8th IEEE International Conference on Web Services (ICWS 2010), Application and Industry Track, IEEE Press, Miami, Florida, USA, July 05 - 10, 2010.
• Babak Khosravifar, Jamal Bentahar, Ahmad Moazin, Zakaria Maamar and Philippe Thiran (2010). Analyzing Communities vs. Single Agent-based Web Services: Trust Perspectives. In Proceedings of the 7th IEEE International Conference on Services Computing (SCC 2010), IEEE Press, Miami, Florida, USA, July 05 - 10, 2010.
• Babak Khosravifar, Jamal Bentahar, Ahmad Moazin, and Philippe Thiran (2010). On the Reputation of Agent-based Web Services. In Proceedings of the 24th AAAI Conference on Artificial Intelligence (AAAI 2010), AAAI Press, Atlanta, Georgia, USA, July 11 - 15, 2010.
• Babak Khosravifar, Jamal Bentahar, Philippe Thiran, Ahmad Moazin, and Adrien Guiot (2009). An Approach to Incentive-based Reputation for Communities of Web Services. In Proceedings of IEEE 7th International Conference on Web Services (ICWS 2009), pp. 303-310, Los Angeles, USA, July 06-10, 2009.