Abstract:Graph-structured data is ubiquitous in the world which models complexrelationships between objects, enabling various Web applications. Dailyinfluxes of unlabeled graph data on the Web offer immense potential for theseapplications. Graph self-supervised algorithms have achieved significantsuccess in acquiring generic knowledge from abundant unlabeled graph data.These pre-trained models can be applied to various downstream Web applications,saving training time and improving downstream (target) performance. However,different graphs, even across seemingly similar domains, can differsignificantly in terms of attribute semantics, posing difficulties, if notinfeasibility, for transferring the pre-trained models to downstream tasks.Concretely speaking, for example, the additional task-specific node informationin downstream tasks (specificity) is usually deliberately omitted so that thepre-trained representation (transferability) can be leveraged. The trade-off assuch is termed as "transferability-specificity dilemma" in this work. Toaddress this challenge, we introduce an innovative deployment module coined asGraphControl, motivated by ControlNet, to realize better graph domain transferlearning. Specifically, by leveraging universal structural pre-trained modelsand GraphControl, we align the input space across various graphs andincorporate unique characteristics of target data as conditional inputs. Theseconditions will be progressively integrated into the model during fine-tuningor prompt tuning through ControlNet, facilitating personalized deployment.Extensive experiments show that our method significantly enhances theadaptability of pre-trained models on target attributed datasets, achieving1.4-3x performance gain. Furthermore, it outperforms training-from-scratchmethods on target data with a comparable margin and exhibits fasterconvergence.

Graph-based fine-grained model selection for multi-source domain

Target-Discriminability-Induced Multi-Source-Free Domain Adaptation

Transformer-Based Multi-Source Domain Adaptation Without Source Data.

Multi-source Unsupervised Domain Adaptation on Graphs with Transferability Modeling

Self-Supervised Graph Neural Network for Multi-Source Domain Adaptation

Gradual Fine-Tuning with Graph Routing for Multi-Source Unsupervised Domain Adaptation

Source Free Unsupervised Graph Domain Adaptation

GraphControl: Adding Conditional Control to Universal Graph Pre-trained Models for Graph Domain Transfer Learning

Multiple Graphs and Low-Rank Embedding for Multi-Source Heterogeneous Domain Adaptation

Multi-source unsupervised domain adaptation for object detection

Source Free Graph Unsupervised Domain Adaptation

Adaptive multi-source domain collaborative fine-tuning for transfer learning

Domain-Aware Graph Network for Bridging Multi-Source Domain Adaptation

MSDS: A Novel Framework for Multi-Source Data Selection Based Cross-Network Node Classification

Multi-source Fully Test-Time Adaptation

One Model for One Graph: A New Perspective for Pretraining with Cross-domain Graphs

Scalable Multi-Source Pre-training for Graph Neural Networks

Model Selection with Model Zoo via Graph Learning

Heterogeneous Graph Attention Network for Unsupervised Multiple-Target Domain Adaptation

Multi- View Teacher with Curriculum Data Fusion for Robust Unsupervised Domain Adaptation

Multi-Source Collaborative Gradient Discrepancy Minimization for Federated Domain Generalization