SSumM

Sparse Summarization of Massive Graphs

Overview

SSumM is a scalable and effective graph summarization algorithm
that yields a sparse summary graph.

SSumM has the following advantages:

Concise

SSumM yields up to 11.2x smaller summary graphs with similar reconstruction error.

Accurate

SSumM achieves up to 4.2x smaller reconstruction error with similarly concise outputs.

Scalable

SSumM summarizes 26x larger graphs while exhibiting linear scalability.

Paper

SSumM: Sparse Summarization of Massive Graphs

{Kyuhan Lee*, Hyeonsoo Jo*}, Jihoon Ko, Sungsu Lim, and Kijung Shin
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’20)

[PDF] [BIBTEX]

Code

The source used in the paper is available. [Github Repository]

Datasets
Name #Nodes #Edges Source Download
Ego-Facebook 4k 88k SNAP Link
Caida 26k 106k SNAP Link
Email-Enron 36k 183k SNAP Link
Amazon-0302 262k 899k SNAP Link
DBLP 317k 1.0M SNAP Link
Amazon-0601 403k 2.4M SNAP Link
Skitter 1.7M 11M SNAP Link
LiveJournal 3.9M 34M SNAP Link
Web-UK-02 18M 262M LAW Link
Web-UK-05 39M 783M LAW Link
Social
Internet
Email
Co-purchase
Collaboration
Hyperlinks
Teams