Exploring In-Context Learning Capabilities of Foundation Models for Generating Knowledge Graphs from Text
Hanieh Khorashadizadeh Nandana Mihindukulasooriya Sanju Tiwari Jinghua Groppe Sven Groppe University of Lübeck, Germany IBM Research, Ireland Universidad Autonoma de Tamaulipas, Mexico
Abstract
Knowledge graphs can represent information about the real-world using entities and their relations in a structured and semantically rich manner and they enable a variety of downstream applications such as question-answering, recommendation systems, semantic search, and advanced analytics. However, at the moment, building a knowledge graph involves a lot of manual effort and thus hinders their application in some situations and the automation of this process might benefit especially for small organizations. Automatically generating structured knowledge graphs from a large volume of natural language is still a challenging task and the research on sub-tasks such as named entity extraction, relation extraction, entity and relation linking, and knowledge graph construction aims to improve the state of the art of automatic construction and completion of knowledge graphs from text.
中文速览
大规模语言模型（如ChatGPT）在预训练过程中积累了海量世界知识，但这些知识以非结构化的方式"封存"在模型参数里，难以被直接检验和利用。这篇论文尝试用"上下文学习"（in-context learning）的方式——即通过精心设计的提示词（prompt）而非重新训练——让ChatGPT等基础模型直接完成知识图谱的构建与补全任务，涵盖实体关系抽取、缺失事实填充、错误事实核查等多个子任务。作者围绕七个具体研究问题设计实验，对模型输出进行定性分析，结果表明基础模型在上述场景中展现出相当可观的能力，能够根据给定本体和少量示例生成结构化三元组。这项工作的意义在于：既为资源有限的小型机构提供了一条低成本构建知识图谱的路径，也揭示了幻觉、偏见、无法验证等关键局限，为后续将基础模型与知识图谱互补融合的研究指明了方向。
原文 arXiv:2305.08804;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2305.08804v1