An AI learning technique that learns (logic) programs given a set of (input, output) examples representing correct behaviour, and also a set of examples showing incorrect behaviour.

----
Also a book:

http://www-ai.ijs.si/SasoDzeroski/ILPBook/ilp-front.jpg

The complete text has been made publicly available online by the authors Nada Lavrac and Saso Dzeroski:

* http://www-ai.ijs.si/SasoDzeroski/ILPBook/