Bending the Future: Autoregressive Modeling of Temporal Knowledge Graphs in Curvature-Variable Hyperbolic Spaces
This work addresses the challenge of capturing hierarchies in temporal knowledge graphs for applications like knowledge base completion, though it appears incremental as it builds on existing hyperbolic embedding methods.
The paper tackled the problem of modeling hierarchical characteristics in temporal knowledge graphs (TKGs) by proposing HyperVC, which uses hyperbolic space with variable curvature to encode hierarchies between graphs at different timestamps and within individual graphs. Experiments on four benchmark datasets showed substantial improvements, particularly on datasets with higher hierarchical levels.
Recently there is an increasing scholarly interest in time-varying knowledge graphs, or temporal knowledge graphs (TKG). Previous research suggests diverse approaches to TKG reasoning that uses historical information. However, less attention has been given to the hierarchies within such information at different timestamps. Given that TKG is a sequence of knowledge graphs based on time, the chronology in the sequence derives hierarchies between the graphs. Furthermore, each knowledge graph has its hierarchical level which may differ from one another. To address these hierarchical characteristics in TKG, we propose HyperVC, which utilizes hyperbolic space that better encodes the hierarchies than Euclidean space. The chronological hierarchies between knowledge graphs at different timestamps are represented by embedding the knowledge graphs as vectors in a common hyperbolic space. Additionally, diverse hierarchical levels of knowledge graphs are represented by adjusting the curvatures of hyperbolic embeddings of their entities and relations. Experiments on four benchmark datasets show substantial improvements, especially on the datasets with higher hierarchical levels.