用大模型补全知识图谱缺失实体关系，结合文本描述提升预测准确率。
GenIC: An LLM-Based Framework for Instance Completion in Knowledge Graphs
- 分两步：先分类预测可能关系，再生成尾实体序列。
- 在三个数据集上优于现有方法，尤其在实体描述丰富时效果更佳。
- 适合需要利用文本信息补全知识图谱的场景，如智能问答、推荐系统。
知识图谱补全旨在通过添加新三元组来填补知识库中的空白。该任务的复杂性取决于已知三元组部分的数量。实例补全（Instance Completion）是在仅给出头实体（h, ?, ?）的情况下预测关系-尾实体对。值得注意的是，现代知识库常包含实体描述和类型信息，这些文本可为推断缺失事实提供重要上下文。本文提出一种基于大语言模型（LLM）的端到端实例补全框架GenIC。该框架分为两步：第一步将属性预测视为多标签分类任务；第二步将链接预测建模为序列到序列的生成任务。利用大模型从实体描述中提取事实并识别知识图谱模式的能力，实验在三个数据集上验证了其优越性。代码已开源。
原文摘要 · Abstract (English)
Knowledge graph completion aims to address the gaps of knowledge bases by adding new triples that represent facts. The complexity of this task depends on how many parts of a triple are already known. Instance completion involves predicting the relation-tail pair when only the head is given (h, ?, ?). Notably, modern knowledge bases often contain entity descriptions and types, which can provide valuable context for inferring missing facts. By leveraging these textual descriptions and the ability of large language models to extract facts from them and recognize patterns within the knowledge graph schema, we propose an LLM-powered, end-to-end instance completion approach. Specifically, we introduce GenIC: a two-step Generative Instance Completion framework. The first step focuses on property prediction, treated as a multi-label classification task. The second step is link prediction, framed as a generative sequence-to-sequence task. Experimental results on three datasets show that our method outperforms existing baselines. Our code is available at https://github.com/amal-gader/genic.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。