Draw a Portrait of Your Graph Data: An Instance-Level Profiling Framework for Graph-Structured Data
This addresses the need for better diagnostic tools in graph ML to understand model reliability at the node level, though it is incremental as it builds on existing evaluation methods.
The paper tackles the problem of diagnosing fine-grained differences in graph machine learning models that achieve similar overall performance but fail on different subsets of nodes, by introducing NodePro, a node profiling framework that assigns interpretable profile scores to individual nodes, revealing systematic model differences and identifying inconsistent nodes in knowledge graphs.
Graph machine learning models often achieve similar overall performance yet behave differently at the node level, failing on different subsets of nodes with varying reliability. Standard evaluation metrics such as accuracy obscure these fine grained differences, making it difficult to diagnose when and where models fail. We introduce NodePro, a node profiling framework that enables fine-grained diagnosis of model behavior by assigning interpretable profile scores to individual nodes. These scores combine data-centric signals, such as feature dissimilarity, label uncertainty, and structural ambiguity, with model-centric measures of prediction confidence and consistency during training. By aligning model behavior with these profiles, NodePro reveals systematic differences between models, even when aggregate metrics are indistinguishable. We show that node profiles generalize to unseen nodes, supporting prediction reliability without ground-truth labels. Finally, we demonstrate the utility of NodePro in identifying semantically inconsistent or corrupted nodes in a structured knowledge graph, illustrating its effectiveness in real-world settings.