新的正则化方法解决了知识图谱嵌入中的候选集干扰问题

作者 PulseAugur 编辑部 · [1 个来源] · 2026-08-26 04:00

研究人员推出了一种名为Matched Excess-Outranker Regularization (MEOR) 的新方法，以解决持续知识图谱嵌入中的候选集干扰问题。这种干扰发生在当新实体的引入改变现有答案的排名时，即使它们的得分保持稳定。MEOR引入了一个主机级目标，仅当新来者的压力超过结构匹配的旧参考时才对其进行惩罚，从而保留了合法新实体的学习信号。实验表明，MEOR提高了历史当前宇宙平均倒数排名 (MRR)，并减少了候选集干扰，其性能优于重放和持久校准等现有方法。 AI

影响这项研究通过更好地处理动态数据，有望提高依赖知识图谱的信息检索系统的准确性和相关性。

排序理由该集群包含一篇详细介绍知识图谱嵌入新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 46 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍知识图谱嵌入新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Hao Ren, Junbin Gao, Jiaojiao Jiang · 2026-08-26 04:00

用于持续知识图谱嵌入中候选集干扰的匹配超排序正则化

arXiv:2608.24273v1 Announce Type: new Abstract: Continual knowledge graph embedding updates entity and relation representations as a graph grows. Existing methods primarily address catastrophic forgetting, but entity admission also changes the candidate universe of every compatib…

本文如何被排名

报道来源 [1]

用于持续知识图谱嵌入中候选集干扰的匹配超排序正则化

相关话题