一种结合语言模型的电力知识图谱构建方法及设备
By automatically extracting knowledge from power documents using language models and ontology extraction instructions, and combining data sampling and quality inspection methods, the problem of uneven data quality in the power knowledge graph was solved, thus achieving the accuracy and completeness of the knowledge graph.
CN117708344BActive Publication Date: 2026-07-17STATE GRID FUJIAN ELECTRIC POWER CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID FUJIAN ELECTRIC POWER CO LTD
- Filing Date
- 2023-12-05
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
The existing power knowledge graph has complex relationships between entities and inconsistent data quality, making it difficult to guarantee the accuracy and completeness of the knowledge graph.
Method used
The system automatically extracts knowledge from power documents using language models, ontology extraction instructions, and knowledge extraction instructions. It also samples data based on user access volume, node degree, and information source confidence, and performs quality checks on the sampling results using language models to complete entities or relationships with missing attribute values.
Benefits of technology
This ensured the accuracy and completeness of the power knowledge graph and improved the data quality of the knowledge graph.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN117708344B_ABST
Abstract
本发明涉及一种结合语言模型的电力知识图谱构建方法，包括：构建电力知识图谱：获取若干信息源文档；从信息源文档中抽取三元组；将三元组存入数据库，得到电力知识图谱；数据质检：从电力知识图谱中抽取k条知识作为采样结果；构建第一任务文本，第一任务文本包括采样结果、信息源文档以及质检指示；将第一任务文本输入至语言模型，语言模型根据信息源文档按质检指示对采样结果进行数据质检，得到质检结果；数据补全：遍历知识图谱中所有实体，根据所述知识图谱的本体，查找属性值缺失的实体；获取该实体的属性值并加入知识图谱。
Need to check novelty before this filing date? Find Prior Art