LichAmnesia / MBSCM

Motif-based spectral clustering method

MBSCM

Python implementation for Motif-based spectral clustering method for single cluster.

Motif-based spectral clustering method is an algorithm posted in Higher-order Organization of Complex Networks by Austin R. Benson, David F. Gleich, and Jure Leskovec.

The implemented Code is here.

Algorithm

The algorithm performs the following steps:

1. input Graph G, unweighted and directed.
2. Using G form a new weighted graph W.
3. Apply spectral clustering on W.
4. Output the clusters for each 7 motifs.
5. Examine each output to see which motif has thebest cluster.
6. Return best cluster and the motif used to get that cluster.

Reference

Higher-order Organization of Complex Networks. Austin R. Benson, David F. Gleich, and Jure Leskovec. Science, vol. 353, no. 6295, pp. 163-166, 2016.