This paper introduces a new methodology that combines human learners and inductive machine learners, which learn from data, into a cooperative multi-agent learning system. Educational cooperative learning techniques are used to model computational cooperative learning systems, which are used to facilitate the discovery of new knowledge to be used for classification purposes. The combination of humans and machines into a cooperative multi-agent learning system entails the marriage of the data-driven inductive machine learning approach with the knowledge-driven traditional knowledge-acquisition method. Results show that this approach succeeds in addressing the knowledge-acquisition bottleneck.