الملخص
Extrapolative temporal knowledge graph reasoning (TKGR) predicts future facts from historical snapshots. Most existing methods train once on an early prefix of the timeline and then use a frozen model for all future timestamps. We argue that this fixed-prefix protocol is misaligned with extrapolation. It learns from a static prefix, whereas the target stream is non-stationary: new entities and facts emerge, temporal dependencies shift across regimes, and recurring historical signals must be refreshed online. As a result, models trained only on early snapshots become outdated and degrade over long horizons. We address this mismatch by formulating extrapolative TKGR as continual learning over streaming snapshots. Under this view, effective extrapolation must jointly handle current dynamics, stable knowledge, and recurring historical evidence. Based on these requirements, we propose History-enhanced Two-Step Continual Learning (HiTS-CL), a backbone-agnostic continual learning framework for extrapolative TKGR. HiTS-CL tracks current dynamics via continual fine-tuning, preserves stable knowledge via multi-teacher adaptive distillation, and retains recurring historical evidence via a selective memory of recent and frequent facts. We integrate HiTS-CL into five representative TKGR backbones and evaluate it on four benchmark datasets. HiTS-CL consistently improves extrapolation accuracy, reduces long-horizon degradation, and outperforms strong continual-learning baselines, including a recent method for temporal knowledge graphs. Source code and data are available at https://github.com/liuyansong98/HiTS-CL.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Liu, Y., Liu, R., Zuo, Y., Zhao, H., Fu, D., Zhang, F., Zhuang, F., Chen, Y., & Li, Z. (2026). HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation. https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation
MLA 9
Liu, Yansong, et al. "HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation." https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation.
شيكاغو (المؤلف–التاريخ)
Liu, Yansong, Rui Liu, Yuan Zuo, Hongwei Zhao, Da Fu, Fuwei Zhang, Fuzhen Zhuang, Yong Chen, and Zhe Li. 2026. "HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation." https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation.
هارفارد
Liu, Y., Liu, R., Zuo, Y., Zhao, H., Fu, D., Zhang, F., Zhuang, F., Chen, Y. and Li, Z. (2026) 'HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation', Available at: https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation.
فانكوفر
Liu Y, Liu R, Zuo Y, Zhao H, Fu D, Zhang F, et al. HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation. https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation
IEEE
Y. Liu, R. Liu, Y. Zuo, H. Zhao, D. Fu, F. Zhang, F. Zhuang, Y. Chen, and Z. Li, "HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation," https://omanscience.com/ar/articles/hits-cl-a-continual-learning-framework-for-long-horizon-temporal-knowledge-graph-extrapolation.