Learnability of Description Logic Programs

Carin-ALN is an interesting new rule learning bias for ILP. By allowing description logic terms as predicates of literals in datalog rules, it extends the normal bias used in ILP as it allows the use of all quantified variables in the body of a clause. It also has at-least and at-most restrictions to access the amount of indeterminism of relations. From a complexity point of view Carin-ALN allows to handle the difficult indeterminate relations efficiently by abstracting them into determinate aggregations. This paper describes a method which enables the embedding of Carin-ALN rule subsumption and learning into datalog rule subsumption and learning with numerical constraints. On the theoretical side, this allows us to transfer the learnability results known for ILP to Carin-ALN rules. On the practical side, this gives us a preprocessing method, which enables ILP systems to learn Carin-ALN rules just by transforming the data to be analyzed. We show, that this is not only a theoretical result in a first experiment: learning Carin-ALN rules from a standard ILP dataset.

[1]  Bernhard Nebel,et al.  Terminological Reasoning is Inherently Intractable , 1990, Artif. Intell..

[2]  Bernhard Nebel,et al.  Reasoning and Revision in Hybrid Representation Systems , 1990, Lecture Notes in Computer Science.

[3]  R. Mike Cameron-Jones,et al.  FOIL: A Midterm Report , 1993, ECML.

[4]  Saso Dzeroski,et al.  Inductive logic programming and learnability , 1994, SGAR.

[5]  Franz Baader,et al.  Tableau Algorithms for Description Logics , 2000, TABLEAUX.

[6]  William W. Cohen,et al.  The learnability of description logics with equality constraints , 1994, Machine Learning.

[7]  Alexander Borgida,et al.  Computing Least Common Subsumers in Description Logics , 1992, AAAI.

[8]  Stephen Muggleton,et al.  Efficient Induction of Logic Programs , 1990, ALT.

[9]  Jörg H. Siekmann,et al.  Einführung in die Künstliche Intelligenz , 1982, KIFS.

[10]  Arno J. Knobbe,et al.  Propositionalisation and Aggregates , 2001, PKDD.

[11]  Ronald J. Brachman,et al.  An Overview of the KL-ONE Knowledge Representation System , 1985, Cogn. Sci..

[12]  Alex Borgiday.  On the Relative Expressiveness of Description Logics and Predicate Logics , 1996 .

[13]  William W. Cohen.  Pac-Learning a Restricted Class of Recursive Logic Programs , 1993, AAAI.

[14]  Ralf Küsters,et al.  Computing the Least Common Subsumer and the Most Specific Concept in the Presence of Cyclic ALN-Concept Descriptions , 1998, KI.

[15]  Alon Y. Halevy,et al.  Combining Horn Rules and Description Logics in CARIN , 1998, Artif. Intell..

[16]  Nicolas Helft,et al.  Induction as Nonmonotonic Inference , 1989, KR.

[17]  William W. Cohen,et al.  Learning the Classic Description Logic: Theoretical and Experimental Results , 1994, KR.

[18]  William W. Cohen.  Pac-Learning Non-Recursive Prolog Clauses , 1995, Artif. Intell..

[19]  C. Eline Rouveirol,et al.  Towards Learning in Carin-aln , 2000 .

[20]  Jörg-Uwe Kietz,et al.  An Efficient Subsumption Algorithm for Inductive Logic Programming , 1994, ICML.

[21]  Michael Frazier,et al.  CLASSIC Learning , 1994, COLT.

[22]  Stefan Wrobel,et al.  Transformation-Based Learning Using Multirelational Aggregation , 2001, ILP.

[23]  Werner Nutt,et al.  Tractable Concept Languages , 1991, IJCAI.

[24]  Katharina Morik,et al.  A Polynomial Approach to the Constructive Induction of Structural Knowledge , 2004, Machine Learning.