Inductive Logic Programming

The 11th international conference on Inductive Logic Programming, ILP2001, was held in Strasbourg, France, September 9 11, 2001. ILP2001 was co-located withthe3rdinternationalworkshoponLogic,Learning,andLanguage(LLL2001), and nearly co-located with the joint 12th European Conference on Machine Learning (ECML2001) and 5th European conference on Principles and Practice of Knowledge Discovery in Databases (PKDD2001). Continuing a series of international conferences devoted to Inductive Logic Programming and Relational Learning, ILP2001 is the central annual event for researchersinterestedinlearningstructuredknowledgefromstructuredexamples and background knowledge. One recent one major challenge for ILP has been to contribute to the ex- nentialemergenceofDataMining,andtoaddressthehandlingofmulti-relational databases. On the one hand, ILP has developed a body of theoretical results and algorithmicstrategiesforexploringrelationaldata,essentiallybutnotexclusively from a supervised learning viewpoint. These results are directly relevant to an e?cient exploration of multi-relational databases. Ontheotherhand,DataMiningmightrequirespeci?crelationalstrategiesto be developed, especially with regard to the scalability issue. The near-colocation of ILP2001 with ECML2001-PKDD2001 was an incentive to increase cro- fertilization between the ILP relational savoir-faire and the new problems and learning goals addressed and to be addressed in Data Mining. Thirty-seven papers were submitted to ILP, among which twenty-one were selected and appear in these proceedings. Several non-disjoint trends can be observed, along an admittedly subjective clustering. On the theoretical side, a new mode of inference is proposed by K. Inoue, analog to the open-ended mode of Bayesian reasoning (where the frontier - tween induction and abduction wanes). New learning re?nement operators are proposed by L. Badea, while R. Otero investigates negation-handling settings.

Includes supplementary material: sn.pub/extras

Klappentext

(where the frontier - tween induction and abduction wanes). New learning re?nement operators are proposed by L. Badea, while R. Otero investigates negation-handling settings.

Inhalt
A Refinement Operator for Theories.- Learning Logic Programs with Neural Networks.- A Genetic Algorithm for Propositionalization.- Classifying Uncovered Examples by Rule Stretching.- Relational Learning Using Constrained Confidence-Rated Boosting.- Induction, Abduction, and Consequence-Finding.- From Shell Logs to Shell Scripts.- An Automated ILP Server in the Field of Bioinformatics.- Adaptive Bayesian Logic Programs.- Towards Combining Inductive Logic Programming with Bayesian Networks.- Demand-Driven Construction of Structural Features in ILP.- Transformation-Based Learning Using Multirelational Aggregation.- Discovering Associations between Spatial Objects: An ILP Application.- ?-Subsumption in a Constraint Satisfaction Perspective.- Learning to Parse from a Treebank: Combining TBL and ILP.- Induction of Stable Models.- Application of Pruning Techniques for Propositional Learning to Progol.- Application of ILP to Cardiac Arrhythmia Characterization for Chronicle Recognition.- Efficient Cross-Validation in ILP.- Modelling Semi-structured Documents with Hedges for Deduction and Induction.- Learning Functions from Imperfect Positive Data.