GEMSEC: Graph Embedding with Self Clustering

ASONAM 2019 Benedek RozemberczkiRyan DaviesRik SarkarCharles Sutton

Network embedding procedures localize nodes of a graph in a low dimensional feature space, which enables machine learning on graph data. In this paper we propose GEMSEC - a graph embedding algorithm which learns a clustering of the nodes simultaneously with the embedding... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK BENCHMARK
Node Classification Deezer Croatia GEMSEC 2 Micro-F1 0.381 # 1
Node Classification Deezer Hungary Smooth GEMSEC 2 Micro-F1 0.409 # 1
Node Classification Deezer Romania GEMSEC 2 Micro-F1 0.378 # 1
Community Detection Facebook Artists Smooth GEMSEC 2 Modularity 0.562 # 1
Community Detection Facebook Athletes Smooth GEMSEC 2 Modularity 0.692 # 1
Community Detection Facebook Celebrities Smooth GEMSEC 2 Modularity 0.649 # 1
Community Detection Facebook Companies Smooth GEMSEC 2 Modularity 0.684 # 1
Community Detection Facebook Government Smooth GEMSEC 2 Modularity 0.712 # 1
Community Detection Facebook Media Smooth GEMSEC 2 Modularity 0.571 # 1
Community Detection Facebook Politicians Smooth GEMSEC 2 Modularity 0.859 # 1
Community Detection Facebook TV Show Smooth GEMSEC 2 Modularity 0.847 # 1

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet