LAP is based on the basic top-down ILP algorithm that has been suitably extended. In particular, coverage of examples is tested by using the abductive proof procedure defined by Kakas and Mancarella. Assumptions can be made in order to cover positive examples and to avoid the coverage of negative ones, and these assumptions can be used as new training data. LAP can be applied for learning in presence of incomplete knowledge and for learning exceptions to classification rules.

See the following paper for a detailed description of the system:

• E. Lamma, P. Mello, M. Milano, F. Riguzzi, "A System for Learning Abductive Logic Programs", Proceedings of the ILPS97 Workshop on Logic Programming and Knowledge Representation (LPKR97), J. Dix, L. M. Pereira, T. Przymusinski eds., October 1997.
• E. Lamma, P. Mello, M. Milano, F. Riguzzi, "Integrating Extensional and Intensional ILP Systems through Abduction", Proceedings of the 7th International Workshop on Logic Program Synthesis and Transformation (LOPSTR'97), July 1997.
• E. Lamma, P. Mello, M. Milano, F. Riguzzi, "A System for Learning Abductive Rules", Proceedings of APPIA-GULP-PRODE 97, M. Falaschi, M. Navarro, A. Policriti (ed.), June 1997.
• E. Lamma, P. Mello, M. Milano, F. Riguzzi, "Integrating Induction and Abduction in Logic Programming", Proceedings of the 2nd International Conference on Computational Intelligence & Neuroscience, March 1997
• F.Esposito, E.Lamma, D.Malerba, P.Mello, M.Milano, F.Riguzzi, G.Semeraro, "Learning Abductive Logic Programs", Proceedings of ECAI workshop on Abductive and Inductive Reasoning, Budapest, August 1996.

The system has been implemented in Sicstus Prolog 3#5. To install it, download the following files:

• learning system
• abductive proof procedure
• exception example
• incompleteness example
• instructions