Multi-agent learning. In multi-agent learning, a group of distributed, learning agents improves its group performance through collective experience. The agent-based paradigm is ideal for supporting distributed and collaborative ontology learning. Agents are able to locate and translate disparate referenced concepts and improve concept precisions as they enhance their experience from others. It is proven that agents are able to: 1) learn diverse ontologies; 2) locate, share, and integrate knowledge; 3) improve group performance through experience; and 4) introduce novel approach and novel algorithms.