Nota: Instead of using the algorithms proposed here (Energetics and Excess), we recommend to use their improved version (Cenergetics and Cexpress). These more recent algorithm correspond to the approaches described in [2], their code source is available on GitHub.
This repository provides the code and the data used in the paper Unsupervised Exceptional Attributed Sub-graph Mining in Urban Data. It is organized as follows :
- Algorithms: This folder contains the algorithm implementations of the approaches defined in the paper. For each approach, we provide the code and the release.
- Data: This folder contains the real world data and the synthetic graph generator used in the experiments.
All the source codes are implemented in Java.
1 - Anes Bendimerad, Marc Plantevit, Céline Robardet. Unsupervised Exceptional Attributed Sub-graph Mining in Urban Data. ICDM 2016.
2 - Anes Bendimerad, Marc Plantevit, Céline Robardet. Mining exceptional closed patterns in attributed graphs. KAIS 2017.