Learning to act: qualitative learning of deterministic action models | Journal of Logic and Computation | DeepDyve