Business managers need to respond to environmental changes both promptly and properly. To achieve that, environmental scanners play an important role in discovering and monitoring the information of interest (IOI). Once an environmental scanner detects a new IOI, it may trigger various actions such as event logging and user notification. In this project, we argue that business environmental scanning should be conducted in a continuous and resource-bounded manner. The scanner should continuously scan for IOI for the managers so that IOI (and its updates) may be comprehensively discovered. It should also control its consumption of resource (e.g. bandwidth of the computer network and services of the information servers on the Internet) in order to prevent the related network and servers from being exhausted. We thus plan to develop a multiagent framework ACES to achieve continuous and resource-bounded environmental scanning. The agents in ACES form an adaptive agent society in the sense that they automatically adapt their organization and specialty to (1) information needs of individual users, (2) resource limit of environmental scanning, (3) distribution of IOI in the environments, and (4) update behaviors of the IOI. We also plan to empirically evaluate the framework in two contexts: (1) a simulated information space on the World-Wide-Web, and (2) a prototype system running on the Internet for scanning financial environments. Major criteria for the empirical evaluation include effectiveness, completeness and timelines of information discovery and monitoring, which are both essential and significant for business administration.
Keywords: Environmental scanning, adaptive agent, information discovery, information monitoring
Back to Research Project