Unverified60% confidenceFactTime unknown
GraphRAG通过构建知识图谱，将函数、类、模块作为节点，将调用关系、继承关系、依赖关系作为边来组织代码知识
2
Sources
60%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
CodeGraph实测：代码知识图谱如何将Token消耗降低90%
bilibiliAGI_Ananas
Related Entities
Related Claims
UnverifiedGraphRAG架构通过知识图谱显式建模角色关系与事件因果链来弥补纯向量检索的局限83% similarUnverified在本体论作为模式层的 GraphRAG 流程中，本体论定义了图谱中允许存在的实体和关系类型，直接决定知识图谱质量与推理能力上限80% similarUnverifiedGraphRAG 先用 LLM 从代码中抽取实体和依赖关系构建知识图谱，检索时通过图遍历而非向量相似度定位答案，适合多跳推理场景但放弃了对非结构化文本的模糊语义匹配80% similarUnverifiedKnowledge Graphs and decision trees both organize scattered information into logically connected network structures, enabling AI to reason and retrieve along relationship chains.77% similarUnverifiedGraphRAG范式先用向量检索定位候选实体节点，再通过图遍历沿关系链条扩展上下文，形成语义召回加结构推理的两阶段管道75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/9786API
curl https://kongchang.com/api/v1/knowledge/claims/9786MCP
get_claim(id=9786)