Table 6 Effectiveness analysis of the adaptive structural community extraction Module.

From: Simple yet effective heuristic community detection with graph convolution network

  

Min DBI

Max Q

Max NMI

Max ACC

Max F1

Max ARI

Max FMI

Max SC

Cora

With global structure

0.458282

0.765320

0.561161

0.673929

0.663000

0.472555

0.562167

0.858014

Without global structure

0.676832

0.681079

0.528015

0.673929

0.657000

0.451359

0.566008

0.708492

Acm

With global structure

0.684481

0.745318

0.620495

0.844298

0.670900

0.692292

0.788906

0.798375

Without global structure

2.697249

0.470766

0.460258

0.639339

0.740500

0.428861

0.678737

0.269921

Amap

With global structure

0.377126

0.670459

0.646764

0.685882

0.684800

0.506579

0.607000

0.899968

Without global structure

0.242789

0.461736

0.347316

0.452810

0.451600

0.233878

0.453156

0.834500

Uat

With global structure

0.774019

0.280535

0.248141

0.547059

0.575400

0.243589

0.476756

0.648977

Without global structure

0.723423

0.241941

0.219666

0.512605

0.500600

0.196388

0.499371

0.655903

Back to article page