Concept-free Causal Disentanglement with Variational Graph Auto-Encoder
This addresses the challenge of interpretable graph representation learning for domains with rapidly growing graph data, offering a novel unsupervised approach that avoids reliance on concept labels.
The paper tackles the problem of learning disentangled representations for graphs without concept labels, proposing an unsupervised method based on a theoretically provable upper bound and incorporating it into Variational Graph Auto-Encoders and meta-learning frameworks, resulting in up to 29% and 11% absolute improvements in AUC over baselines.
In disentangled representation learning, the goal is to achieve a compact representation that consists of all interpretable generative factors in the observational data. Learning disentangled representations for graphs becomes increasingly important as graph data rapidly grows. Existing approaches often rely on Variational Auto-Encoder (VAE) or its causal structure learning-based refinement, which suffer from sub-optimality in VAEs due to the independence factor assumption and unavailability of concept labels, respectively. In this paper, we propose an unsupervised solution, dubbed concept-free causal disentanglement, built on a theoretically provable tight upper bound approximating the optimal factor. This results in an SCM-like causal structure modeling that directly learns concept structures from data. Based on this idea, we propose Concept-free Causal VGAE (CCVGAE) by incorporating a novel causal disentanglement layer into Variational Graph Auto-Encoder. Furthermore, we prove concept consistency under our concept-free causal disentanglement framework, hence employing it to enhance the meta-learning framework, called concept-free causal Meta-Graph (CC-Meta-Graph). We conduct extensive experiments to demonstrate the superiority of the proposed models: CCVGAE and CC-Meta-Graph, reaching up to $29\%$ and $11\%$ absolute improvements over baselines in terms of AUC, respectively.