A knowledge graph-based power marketing business alignment method and system

By using entity type embedding vector clustering and knowledge distillation verification, a power marketing business alignment rule base and training set are generated, which solves the problem of inaccurate alignment of power marketing businesses in existing technologies and achieves more accurate alignment results.

CN122366426APending Publication Date: 2026-07-10GUANGDONG POWER GRID CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-14
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing alignment methods for electricity marketing operations fail to deeply understand the semantics of electricity marketing operations, resulting in inaccurate alignment, especially when dealing with synonyms, abbreviations, and noise.

Method used

By clustering and aligning entity type embedding vectors, combined with knowledge distillation and inference verification, an alignment rule base and training set for power marketing business are generated. The alignment model is then trained using a pre-trained large model to improve alignment accuracy.

Benefits of technology

This approach ensures that the alignment results for electricity marketing business are semantically reasonable and logically correct, improving the accuracy and reliability of the alignment while reducing background noise and the risk of mismatches.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122366426A_ABST
    Figure CN122366426A_ABST
Patent Text Reader

Abstract

The application discloses a kind of based on knowledge graph's electric power marketing business alignment method and system, belong to electric power data processing technical field.The application obtains several entities and several entity type embedding vectors by the knowledge graph of electric power marketing business, then the entity type embedding vector is clustered to obtain several entity sets, then the entity in entity set is aligned and processed, determines several to be reasoned entity pairs;Afterwards, the knowledge distillation and reasoning verification are carried out by the to-be-reasoned entity pair to obtain the electric power marketing business alignment rule base and the electric power marketing business alignment training set;Finally, the electric power marketing business alignment model is obtained by training the electric power marketing big model through the electric power marketing business alignment training set, and the electric power marketing business alignment model and the electric power marketing business alignment rule base are used to align the to-be-processed electric power marketing business and the knowledge graph, which can solve the technical problem of inaccurate electric power marketing business alignment in the prior art.
Need to check novelty before this filing date? Find Prior Art