Paper
W. Van Laer, H. Blockeel, and L. De Raedt, "Inductive Constraint Logic
and the Mutagenesis Problem"
in Proceedings of the Eighth Dutch Conference on Artificial Intelligence,
ed. J.-J.Ch. Meyer and L.C. van der Gaag, 1996, pp. 265-276
[ps format]
Abstract
A novel approach to learning first order logic formulae
from positive and negative examples is incorporated in a system named ICL
(Inductive Constraint Logic). In ICL, examples are viewed as interpretations
which are true or false for
the target theory, whereas in present inductive logic programming systems,
examples are true and false ground facts (or clauses).
Furthermore, ICL uses a clausal representation, which corresponds to
a conjunctive normal form where each conjunct forms a constraint on
positive examples, whereas classical learning techniques
have concentrated on concept representations in disjunctive normal form.
We present some experiments with this new system on the mutagenesis problem.
These experiments illustrate some of the differences with other systems, and
indicate that our approach should work at least as well as the more classical
approaches.