Semi-supervised Domain Adaptation in Graph Transfer Learning

TL;DR

To deal with the domain shift, a method named Semi-supervised Graph Domain Adaptation (SGDA) is proposed, which adds adaptive shift parameters to each of the source nodes, which are trained in an adversarial manner to align the cross-domain distributions of node embedding.

Abstract

As a specific case of graph transfer learning, unsupervised domain adaptation on graphs aims for knowledge transfer from label-rich source graphs to unlabeled target graphs. However, graphs with topology and attributes usually have considerable cross-domain disparity and there are numerous real-world scenarios where merely a subset of nodes are labeled in the source graph. This imposes critical challenges on graph transfer learning due to serious domain shifts and label scarcity. To address these challenges, we propose a method named Semi-supervised Graph Domain Adaptation (SGDA). To deal with the domain shift, we add adaptive shift parameters to each of the source nodes, which are trained in an adversarial manner to align the cross-domain distributions of node embedding, thus the node classifier trained on labeled source nodes can be transferred to the target nodes. Moreover, to address the label scarcity, we propose pseudo-labeling on unlabeled nodes, which improves classification on the target graph via measuring the posterior influence of nodes based on their relative position to the class centroids. Finally, extensive experiments on a range of publicly accessible datasets validate the effectiveness of our proposed SGDA in different experimental settings.

Figures (6)

• Figure 1: The semi-supervised domain adaptation on graphs.
• Figure 2: The framework of SGDA. The source graph and target graph with reconstructed high-order topologies $P^s$ and $P^t$ are fed into a two-layer graph convolutional network to generate generalized node embeddings, where source graph are added with shift parameters $\bm{\xi}$ to promote distribution alignment. Three losses $\mathcal{L}_{Sup}$, $\mathcal{L}_{AT}$, and $\mathcal{L}_{PL}$ perform supervised learning, domain adversarial transformation via shifting, and pseudo-labeling with posterior scores, respectively.
• Figure 3: The results of ablation study on the A $\Rightarrow$ C task (left) and the A $\Rightarrow$ D task (right).
• Figure 4: Visualization of node embeddings learned by UDAGCN and SGDA on the A $\Rightarrow$ C task.
• Figure 5: The model performance with different label rates on the A $\Rightarrow$ C task (left) and the A $\Rightarrow$ D task (right).