Representation Learning over Dynamic Graphs
This work addresses the challenge of learning representations for dynamic graphs, which is crucial for applications like social network analysis and recommendation systems, but it appears incremental as it builds on existing representation learning methods with specific enhancements.
The paper tackles the problem of encoding evolving information in dynamic graphs into low-dimensional node embeddings, proposing DyRep, which uses a time-scale dependent multivariate point process model and significantly outperforms baselines on real-world datasets for dynamic link and event time prediction.
How can we effectively encode evolving information over dynamic graphs into low-dimensional representations? In this paper, we propose DyRep, an inductive deep representation learning framework that learns a set of functions to efficiently produce low-dimensional node embeddings that evolves over time. The learned embeddings drive the dynamics of two key processes namely, communication and association between nodes in dynamic graphs. These processes exhibit complex nonlinear dynamics that evolve at different time scales and subsequently contribute to the update of node embeddings. We employ a time-scale dependent multivariate point process model to capture these dynamics. We devise an efficient unsupervised learning procedure and demonstrate that our approach significantly outperforms representative baselines on two real-world datasets for the problem of dynamic link prediction and event time prediction.