One-shot Learning for Temporal Knowledge Graphs
This addresses the challenge of handling new, unseen relations in temporal knowledge graphs, which is an incremental improvement over static graph methods.
The paper tackles the problem of link prediction in temporal knowledge graphs under data scarcity, particularly for sparse relations, by proposing a one-shot learning framework that outperforms state-of-the-art baselines on two benchmarks.
Most real-world knowledge graphs are characterized by a long-tail relation frequency distribution where a significant fraction of relations occurs only a handful of times. This observation has given rise to recent interest in low-shot learning methods that are able to generalize from only a few examples. The existing approaches, however, are tailored to static knowledge graphs and not easily generalized to temporal settings, where data scarcity poses even bigger problems, e.g., due to occurrence of new, previously unseen relations. We address this shortcoming by proposing a one-shot learning framework for link prediction in temporal knowledge graphs. Our proposed method employs a self-attention mechanism to effectively encode temporal interactions between entities, and a network to compute a similarity score between a given query and a (one-shot) example. Our experiments show that the proposed algorithm outperforms the state of the art baselines for two well-studied benchmarks while achieving significantly better performance for sparse relations.