Inductive Logic Programming : 9th International Workshop, ILP-99, Bled, Slovenia, June 1999 : Proceedings (Lecture Notes in Computer Science, 1634)

✍ Scribed by Peter Flach, Saso Dzeroski (editor)

• Publisher: Springer
• Year: 1999
• Tongue: English
• Leaves: 308
• Category: Library

✦ Synopsis

This book constitutes the refereed proceedings of the 9th International Conference on Inductive Logic Programming, ILP-99, held in Bled, Slovenia, in June 1999.
The 24 revised papers presented were carefully reviewed and selected from 40 submissions. Also included are abstracts of three invited contributions. The papers address all current issues in inductive logic programming and inductive learning, from foundational and methodological issues to applications, e.g. in natural language processing, knowledge discovery, and data mining.

✦ Table of Contents

Lecture Notes in Artificial Intelligence
Foreword
Committee
Table of Contents
Probabilistic Relational Models
Relational Logic
Probabilistic Models
Probabilistic Relational Models
Basic Language
PRM Semantics
Inference
Learning
Structural Uncertainty
Conclusions and Further Directions
Inductive Databases (Abstract)
Some Elements of Machine Learning (Extended Abstract)
Introduction
Structured Models
Right-Sizing Models
Computation-Intensive Learning
Boosting
Looking Ahead
Refinement Operators Can Be (Weakly) Perfect
Introduction
Refinement Operators
Ideal Versus Optimal Refinement Operators
Refinement Operators for Hypotheses Spaces Bounded below by a MSC
From Ideal to Optimal Refinement Operators
Refinement Operators for Clauses with Variable Dependencies
Combining Divide-and-Conquer and Separate-and-Conquer for Ecient and E ective Rule Induction
Introduction
Preliminaries
Reconsider-and-Conquer
Empirical Evaluation
Experimental Setting
Experimental Results
Related Work
Concluding Remarks
Refining Complete Hypotheses in ILP
Introduction
Mechanisms of HYPER
Search
Hypothesis Refinement
Interpreter for Hypotheses
Example
Mechanisms that Did Not Help
Experiments
Conclusions
Acknowledgements
References
Acquiring Graphic Design Knowledge with Nonmonotonic Inductive Learning
1 Introduction
2 Target
3 Page Model and Its Representation
4 Positive and Negative Examples
5 Background Knowledge
6 Learning Methods and Experimental Results
6.1 Normal ILP
6.2 Nonmonotonic Learning
6.3 Another Repeat Learning
7 Conclusions
References
Morphosyntactic Tagging of Slovene Using Progol
Introduction
Morphosyntactic Descriptions
Method
Examples
Background Knowledge
Splitting the Data
P-Progol Parameters and Constraints
Results
The Induction Process
Structure of Induced Theories
Consistency Checking and Ambiguity Reduction
Error Detection
Tagging Accuracy
Conclusions and Future Work
Experiments in Predicting Biodegradability
Introduction
The Dataset
Experiments
Representations
Systems
Evaluation
Results
Discussion
Related Work
Conclusions and Further Work
1BC: a First-Order Bayesian Classier
Sorted Downward Refinement: Building Background Knowledge into a Renement Operator for Inductive Logic Programming
A Strong Complete Schema for Inductive Functional Logic Programming
Application of Different Learning Methods to Hungarian Part-of-speech Tagging
Combining Lapis and WordNet for the Learning of LR Parsers with Optimal Semantic Constraints
Introduction
LR Parsers
Related Work
Lapis
WordNet as a Source of Semantic Tags
Dataset
Results and Evaluation
Conclusions
Acknowledgements
Learning Word Segmentation Rules for Tag Prediction
Introduction
Overview of GA&ILP Learning of Segmentation Rules
Na$mathaccent "707Frelax {i }$ve Theory of Morphology as Word Segmentation Bias
Genetic Algorithms
GA Search for Best NTM
Segmentation Rule Learning Using unhbox voidb @x hbox {textsc {Clog}}
Dataset
Tag Prediction
Method
Evaluation
Results
Conclusions
Acknowledgements
Approximate ILP Rules by Backpropagation Neural Network: A Result on Thai Character Recognition
Introduction
Feature Extraction
Applying Progol to Thai Character Recognition
Approximate ILP rules by BNN
Experimental Results
Related Works and Limitations
Conclusions
Reference
Rule Evaluation Measures: A Unifying View
Introduction
Terminology and Notation
Rules
Contingency Table
Selected Rule Evaluation Measures
A Unifying View
Rule Evaluation Measures in Practice
An Experiment
Rule Filtering
Summary and Discussion
Improving Part of Speech Disambiguation Rules by Adding Linguistic Knowledge
Introduction
Background
Constraint Grammar
The Stockholm-Ume{{accent 23 a}} Corpus
Previous Work on the Induction of Constraint Grammars
Current Work
Training Data
Background Knowledge
The Rule Types: Remove, Barrier and Select
Under-Specification and Feature Unification
Results
Discussion
Future Work
On Sufficient Conditions for Learnability of Logic Programs from Positive Data*
Introduction
General Notation and Terminology
Frames
Admissible Programs
Strongly Admissible Programs
Safe Programs
Examples
Conclusion
A Bounded Search Space of Clausal Theories
Discovering New Knowledge from Graph Data Using Inductive Logic Programming
Analogical Prediction
Introduction
De nitions
Implementation
Experiments
Discussion and related work
Conclusions and further work
Generalizing Refinement Operators to Learn Prenex Conjunctive Normal Forms
Introduction
Prenex Conjunctive Normal Forms
Preliminaries
Some Properties of PCNF
Adding Literals
Substitutions and Specializations
Elementary Substitutions for PCNF
Substitutions and Specializations
Downward Refinement Operator
The Completeness of the Refinement Operator $rho $
Learning PCNF in Practice
Testing Interpretation in a Search Space
The Downward Refinement Operator
Experiments
Conclusion and Future Work
Theory Recovery
Introduction
A Theory Recovery Error Bound
Known Prior
Unknown Prior
Experiments
The Experimental Domain
Method
Theoretical Results for the Domain
Results
Conclusions
Experimental Results
Progol 5.0: Theory Recovery by Logical Back-Propagation
Proof of Theorem T @ref {thm:theorem2}
Instance based function learning
Some Properties of Inverse Resolution in Normal Logic Programs
Introduction
Normal Logic Programs
Declarative Properties
Semantic Properties
Syntactic Properties
Procedural Properties
Discussion
Summary
An assessment of ILP-assisted models for toxicology and the PTE-3 experiment *
Author Index