HOGDA: Boosting Semi-supervised Graph Domain Adaptation via High-Order Structure-Guided Adaptive Feature Alignmen
Abstract: Semi-supervised graph domain adaptation, as a subfield of graph transfer learning, seeks to precisely annotate unlabeled target graph nodes by leveraging transferable features acquired from the limited labeled source nodes. However, most existing studies often directly utilize GCNs-based feature extractors to capture domain-invariant node features, while neglecting the issue that GCNs are insufficient in collecting complex structure information in graph. Considering the importance of graph structure information in encoding the complex relationship among nodes and edges, this paper aims to utilize such powerful information to assist graph transfer learning. To achieve this goal, we develop an novel framework called HOGDA. Concretely, HOGDA introduces a high-order structure information mixing module to effectively assist the feature extractor in capturing transferable node features.Moreover, to achieve fine-grained feature distributions alignment, the AWDA strategy is proposed to dynamically adjust the node weight during adversarial domain adaptation process, effectively boosting the model's transfer ability.
Furthermore, to mitigate the overfitting phenomenon caused by limited source labeled nodes, we also design a TNC strategy to guide the unlabeled nodes to achieve discriminative clustering. Extensive experimental results show that our HOGDA outperforms the state-of-the-art methods on various transfer tasks.
Primary Subject Area: [Systems] Data Systems Management and Indexing
Relevance To Conference: In multimodal applications, graphs are often used to model the correlation between different modal data. Among them, graph node classification techniques play a crucial role in analyzing nodes from different modalities and enabling precise labeling of multi-modal data.
Supplementary Material: zip
Submission Number: 1815
Loading