Page Classification on Wikipedia (90% train ratio)
83.66Macro-F1 ScoreLINE(1st+2nd)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LINE(1st+2nd)Classifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 83.66 | 83.74 | — | |
| LINE(2nd)Classifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 82.07 | 82.17 | — | |
| SkipGramClassifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 82.01 | 82.09 | — | |
| GFClassifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 81.68 | 81.78 | — | |
| LINE(1st)Classifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 81.59 | 81.67 | — | |
| DeepWalkClassifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 81.32 | 81.42 | — | |
| LINE-SGD(2nd)Classifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 79.46 | 79.57 | — | |
| LINE-SGD(1st)Classifier=one-vs-rest logistic regression, Evaluation Protocol=10-run average2015.03 | 78.36 | 78.49 | — | |
| GraphGANTrain-test ratio=9:1, Classifier=logistic regression2017.11 | 19.4 | — | 21.3 | |
| Struc2vecTrain-test ratio=9:1, Classifier=logistic regression2017.11 | 19 | — | 21.1 | |
| DeepWalkTrain-test ratio=9:1, Classifier=logistic regression2017.11 | 18.3 | — | 19.4 | |
| Node2vecTrain-test ratio=9:1, Classifier=logistic regression2017.11 | 17.9 | — | 19.1 | |
| LINETrain-test ratio=9:1, Classifier=logistic regression2017.11 | 16.4 | — | 17.5 |