Experimental comparison of graph-based relational concept learning with inductive logic programming systems

Authors:
Jesus A. Gonzalez;Lawrence B. Holder;Diane J. Cook
Affiliations:
Instituto Nacional de Astrofisica, Optica y Electronica, Puebla, Mexico;Department of Computer Science and Engineering, University of Texas at Arlington, Arlington, TX;Department of Computer Science and Engineering, University of Texas at Arlington, Arlington, TX
Venue:
ILP'02 Proceedings of the 12th international conference on Inductive logic programming
Year:
2002

Citing 7
Cited 2

Conceptual structures: information processing in mind and machine

Conceptual structures: information processing in mind and machine
Efficient top-down induction of logic programs

ACM SIGART Bulletin
Stochastic Complexity in Statistical Inquiry Theory

Stochastic Complexity in Statistical Inquiry Theory
Semantic Networks in Artificial Intelligence

Semantic Networks in Artificial Intelligence
Graph-Based Data Mining

IEEE Intelligent Systems
Empirical and theoretical analysis of relational concept learning using a graph-based representation

Empirical and theoretical analysis of relational concept learning using a graph-based representation
Substructure discovery using minimum description length and background knowledge

Journal of Artificial Intelligence Research

Computing graph-based lattices from smallest projections

KONT'07/KPP'07 Proceedings of the First international conference on Knowledge processing and data analysis
Learning closed sets of labeled graphs for chemical applications

ILP'05 Proceedings of the 15th international conference on Inductive Logic Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

We compare our graph-based relational concept learning approach "SubdueCL" with the ILP systems FOIL and Progol. In order to be fair in the comparison, we use the conceptual graphs representation. Conceptual graphs have a standard translation from graphs into logic. In this way, we introduce less bias during the translation process. We experiment with different types of domains. First, we show our experiments with an artificial domain to describe how SubdueCL performs with the conceptual graphs representation. Second, we experiment with several flat and relational domains. The results of the comparison show that the SubdueCL system is competitive with ILP systems in both flat and relational domains.