Using application conditions to rank graph transformations for graph repair

2024-05-14

International Conference on Graph Transformation

AI Summary

This work addresses the quantification and optimization of consistency repair in graph models, moving beyond traditional binary consistency paradigms by modeling consistency as a gradable, quantitative property. Methodologically, it establishes a theoretical theorem that characterizes constraint violation changes via application condition differences, enabling the first predictable and rankable assessment of repair gains induced by graph transformation rules. It further designs a constraint-driven rule derivation and look-ahead ranking algorithm, yielding a formally verifiable framework for evaluating repair gains. Experimental results demonstrate significant improvements in both accuracy and efficiency of graph repair, establishing a novel paradigm for error detection and automated correction in graph-based systems.

Technology Category

• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

When using graphs and graph transformations to model systems, consistency is an important concern. While consistency has primarily been viewed as a binary property, i.e., a graph is consistent or inconsistent with respect to a set of constraints, recent work has presented an approach to consistency as a graduated property. This allows living with inconsistencies for a while and repairing them when necessary. When repairing inconsistencies in a graph, we use graph transformation rules with so-called impairment- and repair-indicating application conditions to understand how much repair gain certain rule applications would bring. Both types of conditions can be derived from given graph constraints. Our main theorem shows that the difference between the number of actual constraint violations before and after a graph transformation step can be characterized by the difference between the numbers of violated impairment-indicating and repair-indicating application conditions. This theory forms the basis for algorithms with look-ahead that rank graph transformations according to their potential for graph repair. An initial evaluation shows that graph repair can be well supported by rules with these new types of application conditions.

Problem

• Graph Representation
• Correctness Quantification
• Error Correction

Innovation

• Degree of Graph Correctness
• Adaptive Error Correction
• Efficient Algorithm Design