让AI自己发现知识漏洞并补全，提升回答准确性
Enhancing Metacognitive AI: Knowledge-Graph Population with Graph-Theoretic LLM Enrichment
- 用图论指标检测知识图谱中的薄弱区域
- 自动生成问题并检索网页证据，补全知识
- 适合追求高精度问答的AI系统开发者
现代AI普遍缺乏元认知能力，即自我监控、识别知识缺口并自主填补的能力。本文提出MetaKGEnrich，一个全自动流程，赋予大语言模型（LLM）自我修复知识的能力。该系统首先从初始查询构建知识图谱，利用七种图论指标检测知识稀疏区域，再由GPT-4o生成针对性问题，通过Tavily检索网络证据并存入Neo4j，最后使用GraphRAG让GPT-4重新回答并评估改进效果。在三个主流数据集（Google Research Natural Questions、MS MARCO、HotpotQA）各30个查询上测试，分别在87%、83%和80%的问题中提升了答案质量，同时保持已有可靠知识不变。这一概念验证表明，基于拓扑自诊断与定向检索的机制可推动AI向类人元认知学习迈进。
原文摘要 · Abstract (English)
Metacognition-the ability to monitor one's own knowledge state, spot gaps, and autonomously fill them--remains largely absent from modern AI. Here, we present MetaKGEnrich, a fully automated pipeline that endows large language model (LLM) applications with self-directed knowledge repair. The system (i) builds knowledge graphs from a seed query, (ii) detects sparse regions via seven graph metrics, (iii) has GPT-4o generate targeted questions, (iv) retrieves web evidence with Tavily and ingests it into Neo4j, and (v) re-answers the query with GraphRAG for GPT-4 to evaluate improvement. Tested on 30 queries from each of three widely-used datasets: Google Research Natural Questions, MS MARCO, and Hot-potQA. MetaKGEnrich improved answer quality in 80% of HotpotQA questions, 87% of Google Research Natural Questions and 83% of MS MARCO questions, while preserving well-supported regions. This proof of concept demonstrates how topological self-diagnosis plus targeted retrieval can advance AI toward humanlike metacognitive learning.
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。