To date research on improving the state of multi-agent collaboration has only focused on the provision of grounding tools, technologies, protocols, standards and infrastructures that drive the Semantic Web and agent architectures. The basic cognitive and interactional requirements of agents have been neglected leading to the current state-of-the-art development of the Semantic Web whereby its full potential is constrained by the rigid state of multi-agent collaboration. This chapter illustrates and discusses an alternative approach to the development of the agent mediated Semantic Web. The fundamental premise of our approach is that enhancing agents cognitive and interactional abilities is the key to make the digital world of agents more flexible and adaptive in its role to facilitate distributed collaboration. The novelty of this research is that it adapts cognitive models from HCI to develop a heuristic framework called Cognitive Modelling of Multi-Agent Action (COMMAA) for modeling agents’ actions in an attempt to provide an architecture that improves the flexibility of Multi-agent interaction by promoting cognitive awareness. The results of the evaluation show an improved flexibility, interoperability and reusability of agents’ collective behaviours and goals.
An Alternative Approach to Network Demand Estimation: Implementation and Application in Multi-Agent Transport Simulation (MATSim)