A Framework for Multiresolutional Knowledge Dissemination and Collection in Dynamic Agent Societies

The dynamic, distributed, and autonomous nature of agent-based applications make knowledge extraction and distributed processing a complicated task. This paper presents Fido, a programming model and framework implementation that simplifies this task by concealing the details of the agent society under a fixed processing pipeline. Fido developers are then free to focus on the business logic of the application with very little overhead and additionally benefit from multiresolutional result formation, distributed result caching, and other features of the overall framework