Learning disentangled representations in signed directed graphs without social assumptions

Signed graphs can represent complex systems of positive and negative relationships such as trust or preference in various domains. Learning node representations is indispensable because they serve as pivotal features for downstream tasks on signed graphs. However, most existing methods often oversimplify the modeling of signed relationships by relying on social theories, while real-world relationships can be influenced by multiple latent factors. This hinders those methods from effectively capturing the diverse factors, thereby limiting the expressiveness of node representations. In this paper, we propose DINES, a novel method for learning disentangled node representations in signed directed graphs without social assumptions. We adopt a disentangled framework that separates each embedding into distinct factors, allowing for capturing multiple latent factors, and uses signed directed graph convolutions that focus solely on sign and direction, without depending on social theories. Additionally, we propose a new decoder that effectively classifies an edge's sign by considering correlations between the factors. To further enhance disentanglement, we jointly train a self-supervised factor discriminator with our encoder and decoder. Throughout extensive experiments on real-world signed directed graphs, we show that DINES effectively learns disentangled node representations, and significantly outperforms its competitors in predicting link signs.