Combinatorial Learning of Graph Edit Distance via Dynamic Embedding

Runzhong Wang,Tianqi Zhang,Tianshu Yu,Junchi Yan,Xiaokang Yang

DOI: https://doi.org/10.48550/arXiv.2011.15039

2020-12-01

Abstract:Graph Edit Distance (GED) is a popular similarity measurement for pairwise graphs and it also refers to the recovery of the edit path from the source graph to the target graph. Traditional A* algorithm suffers scalability issues due to its exhaustive nature, whose search heuristics heavily rely on human prior knowledge. This paper presents a hybrid approach by combing the interpretability of traditional search-based techniques for producing the edit path, as well as the efficiency and adaptivity of deep embedding models to achieve a cost-effective GED solver. Inspired by dynamic programming, node-level embedding is designated in a dynamic reuse fashion and suboptimal branches are encouraged to be pruned. To this end, our method can be readily integrated into A* procedure in a dynamic fashion, as well as significantly reduce the computational burden with a learned heuristic. Experimental results on different graph datasets show that our approach can remarkably ease the search process of A* without sacrificing much accuracy. To our best knowledge, this work is also the first deep learning-based GED method for recovering the edit path.

Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to simultaneously achieve high efficiency and the ability of path recovery in Graph Edit Distance (GED) calculation. Traditional methods have scalability problems when dealing with large - scale graphs, while deep - learning - based methods can improve efficiency but usually cannot recover the edit path. This paper proposes a hybrid method that combines traditional search techniques and deep learning, aiming to overcome the limitations of existing methods. The specific objectives are as follows:

1. **Improve computational efficiency**: By learning graph embeddings to predict heuristic functions in the search process, unnecessary search branches are reduced, thereby significantly reducing the computational burden.
2. **Recover the edit path**: While ensuring computational efficiency, it is able to recover the optimal edit path from the source graph to the target graph.
3. **Combine the idea of dynamic programming**: Utilize the idea of dynamic programming, reuse previous graph embedding information, avoid repeated calculations, and further improve the efficiency of the algorithm.

The specific contributions of the paper include:
- Proposing the first GED solution based on a deep network, learning heuristic functions for search tree state selection through dynamic graph embeddings.
- Designing a specific graph embedding method that can reuse previous calculation results after each graph modification, thus naturally integrating with the A* algorithm to achieve lower complexity.
- Experimental results show that this method outperforms existing manually - designed approximate solutions on real - world graph datasets, and is faster than traditional exact solutions while maintaining high accuracy.

In summary, the main purpose of this paper is to develop a new method that can both calculate GED efficiently and recover the edit path to meet the challenges of large - scale graph data.