# cgSpan: Closed Graph-Based Substructure Pattern Mining

arXiv: https://arxiv.org/abs/2112.09573 · HTML: https://mlsanity.com/paper/2112.09573/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 0 of its 14 references that are in the map.
- First posted on arXiv 17 Dec 2021.

## Built on by (most cited first, 5 of 5)

- [Pattern Mining for Anomaly Detection in Graphs: Application to Fraud in Public Procurement](https://mlsanity.com/paper/2306.10857/) (2023)
- [MDL-based Compressing Sequential Rules](https://mlsanity.com/paper/2212.10252/) (2022)
- [In-Context Learning of Physical Properties: Few-Shot Adaptation to Out-of-Distribution Molecular Graphs](https://mlsanity.com/paper/2406.01808/) (2024)
- [Farspredict: A benchmark dataset for link prediction](https://mlsanity.com/paper/2303.14647/) (2023)
- [Frequent subgraph-based persistent homology for graph classification](https://mlsanity.com/paper/2512.24917/) (2025)

## Abstract (arXiv)

gSpan is a popular algorithm for mining frequent subgraphs. cgSpan (closed graph-based substructure pattern mining) is a gSpan extension that only mines closed subgraphs. A subgraph g is closed in the graphs database if there is no proper frequent supergraph of g that has equivalent occurrence with g. cgSpan adds the Early Termination pruning method to the gSpan pruning methods, while leaving the original gSpan steps unchanged. cgSpan also detects and handles cases in which Early Termination should not be applied. To the best of our knowledge, cgSpan is the first publicly available implementation for closed graphs mining