Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 29, 2026IEEJ Transactions on Electrical and Electronic Engineering

Enhancing Graph Query Generation for Medical Text by Aligning Domain Knowledge Graph with Large Language Models

View Full Paper
Ask AI
Bookmark
Share

Authors

ZXZhen XuShanghai University of Engineering ScienceWLWei LiaoShanghai University of Engineering ScienceLWLei WangShanghai University of Engineering Science

Discussion

Loading...

Member takes

Implication

Randomized trial improves Cypher generation accuracy in medical domains, suggesting better query responses.

Key Points

  • The central aim is to enhance the capability of large language models in generating accurate Cypher queries for graph databases using medical domain knowledge.
  • Constructed a medical NL2Cypher dataset with single-hop and multi-hop question types.
  • Performed named entity recognition and classified recognized entities for contextual relevance.
  • Utilized a fine-tuning approach that integrated code-structure graph schema details.
  • Achieved 95.65% accuracy in Cypher execution for single-hop questions.
  • Achieved 96.69% accuracy in Cypher execution for multi-hop questions.
  • Demonstrated effectiveness with a smaller training set compared to traditional methods.

Cite This Study

Xu et al. (2026) studied this question.

synapsesocial.com/papers/69f19ff5edf4b46824806a06https://doi.org/10.1002/tee.70313
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1GLaM: Fine-Tuning Large Language Models for Domain Knowledge Graph Alignment via Neighborhood Partitioning and Generative Subgraph Encoding2024 · 12 citations
  2. 2Natural language question answering over RDF data2013 · 10 citations
  3. 3PAROT: Translating natural language to SPARQL2020 · 23 citations
  4. 4A survey on deep learning approaches for text-to-SQL2023 · 171 citations
  5. 5Natural Language Question/Answering2017 · 66 citations