DeepWalk: Online Learning of Social Representations

26 Mar 2014Bryan Perozzi • Rami Al-Rfou • Steven Skiena

We present DeepWalk, a novel approach for learning latent representations of vertices in a network. These latent representations encode social relations in a continuous vector space, which is easily exploited by statistical models. DeepWalk generalizes recent advancements in language modeling and unsupervised feature learning (or deep learning) from sequences of words to graphs.

Full paper

Evaluation


Task Dataset Model Metric name Metric value Global rank Compare
Node Classification BlogCatalog DeepWalk Accuracy 22.50% # 3
Node Classification BlogCatalog DeepWalk Macro-F1 0.214 # 3
Node Classification Citeseer DeepWalk Accuracy 43.2% # 9
Node Classification Cora DeepWalk Accuracy 67.2% # 9
Node Classification NELL DeepWalk Accuracy 58.1% # 3
Node Classification Pubmed DeepWalk Accuracy 65.3% # 9
Node Classification Wikipedia DeepWalk Accuracy 19.40% # 3
Node Classification Wikipedia DeepWalk Macro-F1 0.183 # 3