Graph Edit Distance

Graph Edit Distance (GED) quantifies the similarity between two graphs by measuring the minimum cost of transforming one into the other using node and edge insertions, deletions, and substitutions. Current research focuses on developing efficient approximate GED computation methods, employing techniques like neural networks (often incorporating attention mechanisms and set divergences), A* search algorithms enhanced with learned node matching, and even quantum computing approaches. These advancements are crucial for various applications, including graph-based data analysis, machine learning on graph-structured data, and improving the explainability of graph-based models.