HiTS-CL：面向长时域知识图谱外推的持续学习框架

中文摘要 AI 辅助

外推式时域知识图谱推理（TKGR）旨在从历史快照中预测未来事实。现有方法大多在时间线的早期前缀上训练一次，然后对所有未来时间戳使用冻结模型。我们认为这种固定前缀协议与外推目标不一致：它从静态前缀中学习，而目标流是非平稳的——新实体和事实不断涌现，时域依赖关系在不同阶段发生转移，且反复出现的历史信号需要在线刷新。因此，仅基于早期快照训练的模型会变得过时，并在长时域上性能退化。为解决这一错配问题，我们将外推式TKGR形式化为对流式快照的持续学习。在此视角下，有效的外推必须同时处理当前动态、稳定知识和反复出现的历史证据。基于这些需求，我们提出历史增强的两步持续学习（HiTS-CL），一种与骨干网络无关的持续学习框架，用于外推式TKGR。HiTS-CL通过持续微调追踪当前动态，通过多教师自适应蒸馏保留稳定知识，并通过选择性记忆近期和频繁事实来保留反复出现的历史证据。我们将HiTS-CL集成到五个代表性TKGR骨干网络中，并在四个基准数据集上评估。HiTS-CL持续提升外推准确性，减少长时域退化，并优于强持续学习基线，包括一种近期针对时域知识图谱的方法。源代码和数据可在该HTTPS URL获取。

英文摘要

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 As a result, models trained only on early snapshots become outdated and degrade over long horizons.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.

发表机构

• Beihang University（北京航空航天大学）
• Beijing University of Posts and Telecommunications（北京邮电大学）
• Hubei Engineering University（湖北工程学院）

机构由 AI 辅助整理，请以论文原文为准。