预训练模型驱动的猪病知识图谱构建与应用

文纪威; 杨喆; 王成; 靳哲; 邓兵; 李国亮; 方亚平; 刘峰; 杜小勇

doi:10.13300/j.cnki.hnlkxb.2026.03.011

您当前的位置：

首页 >

文章列表页 >

预训练模型驱动的猪病知识图谱构建与应用

农业人工智能专题 | 更新时间：2026-11-05

- 预训练模型驱动的猪病知识图谱构建与应用
- Pre-trained model-driven construction and application of a swine disease knowledge graph
- 华中农业大学学报 2026年45卷第3期页码：127-136
- 作者机构：
  
  1.华中农业大学信息学院/农业农村部智慧养殖技术重点实验室/农业智能技术教育部工程研究中心/ 湖北省农业大数据工程技术研究中心，武汉430070
  2.农业动物遗传育种与繁殖教育部重点实验室，武汉430070
  3.武汉市农业科学院畜牧兽医研究所，武汉430200
- 作者简介：
  
  文纪威，E-mail：wenjiwei@webmail.hzau.edu.cn
  杜小勇，E-mail：duxiaoyong@mail.hzau.edu.cn
- 基金信息：
  
  湖北省重点研发计划项目(2022BBA0015);中央高校基本科研业务费专项(2662023XXPY005;2662025PY018;2662025DKPY005)
- DOI：10.13300/j.cnki.hnlkxb.2026.03.011
  中图分类号： TP391.4;S858.28
- 收稿：2025-09-30，
  
  纸质出版：2026-05-25
- 稿件说明：
移动端阅览
文纪威，杨喆，王成，等.预训练模型驱动的猪病知识图谱构建与应用［J］.华中农业大学学报，2026，45（3）：127‐136.

WEN Jiwei,YANG Zhe,WANG Cheng,et al.Pre-trained model-driven construction and application of a swine disease knowledge graph[J].Journal of Huazhong Agricultural University,2026,45(03):127-136.
文纪威，杨喆，王成，等.预训练模型驱动的猪病知识图谱构建与应用［J］.华中农业大学学报，2026，45（3）：127‐136. DOI： 10.13300/j.cnki.hnlkxb.2026.03.011.

WEN Jiwei,YANG Zhe,WANG Cheng,et al.Pre-trained model-driven construction and application of a swine disease knowledge graph[J].Journal of Huazhong Agricultural University,2026,45(03):127-136. DOI： 10.13300/j.cnki.hnlkxb.2026.03.011.

摘要

为提高猪病防治知识的获取效率，缓解基层畜牧兽医服务资源匮乏的问题，本研究构建了面向猪病防治的垂直领域知识图谱及智能问答系统。针对多源异构数据描述存在冲突的问题，研究依据兽医病理学建立领域本体，引入BERT语义相似度融合策略，构建专业标注语料库。针对兽医术语复合性强且长尾分布的特点，提出ERNIE-Bi-LSTM-CRF实体识别模型。命名实体识别试验及消融实验结果显示，ERNIE独特的实体级掩码机制能有效克服传统字级掩码对长难术语的语义割裂问题，其

值达89.63%，显著优于Bi-LSTM+CRF及BERT等基线模型。基于此，集成Neo4j图数据库与意图识别技术开发了微信端问答机器人，测试结果显示该机器人能够准确理解专业性的兽医诊疗问题，并能提供全面、科学的检测指导，有效辅助养殖人员进行疾病确诊。结果表明，所构建系统能够实现从自然语言问题到图谱知识检索的闭环处理，可为猪病辅助诊断与防控知识服务提供技术支持。

Abstract

To enhance the efficiency of acquiring knowledge on swine disease prevention and control and to alleviate the shortage of veterinary service resources at the grassroots level，this study develops a vertic

al domain knowledge graph and an intelligent question-answering system specifically for swine disease prevention and control.To resolve descriptive conflicts in multi-source heterogeneous data，the study establishes a domain ontology based on veterinary pathology and introduces a BERT-based semantic similarity fusion strategy to construct a professionally annotated corpus.Given the complex and long-tail distribution of veterinary terms，we propose an ERNIE-Bi-LSTM-CRF entity recognition model.Named entity recognition experiments and ablation tests show that ERNIE’s unique entity-level masking mechanism effectively overcomes the semantic fragmentation problem caused by traditional character-level masking of complex terms.The

score reaches 89.63%，significantly outperforming baseline models such as Bi-LSTM+CRF and BERT.Based on this，the study integrates the Neo4j graph database and intent recognition technology to develop a WeChat-based question-answering chatbot.Tests show that the chatbot can accurately understand professional veterinary diagnosis and treatment issues，providing comprehensive，and scientifically sound diagnostic guidance.This effectively assists farmers in disease diagnosis.The results indicate that the developed system can achieve a closed-loop process，from natural language queries to knowledge retrieval within the knowledge graph，thereby offering technical support for swine disease diagnosis and prevention knowledge services.

关键词

Keywords

references

佟丹丹，曲艳鹏，管鹏宇，等.我国畜牧兽医类企业人才需求现状调研分析［J］.黑龙江畜牧兽医，2021（14）：15-22. TONG D D，QU Y P，GUAN P Y，et al.Investigation and analysis on the current situation of talent demand in animal husbandry and veterinary industry enterprises in China［J］.Heilongjiang animal science and veterinary medicine，2021（14）：15-22 （in Chinese with English abstract） .

PUJARA J，MIAO H，GETOOR L，et al.Knowledge graph identification［C］//The Semantic Web-ISWC 2013.Berlin，Heidelberg：Springer，2013：542-557.

漆桂林，高桓，吴天星.知识图谱研究进展［J］.情报工程，2017，3（1）：4-25. QI G L，GAO H，WU T X.The research advances of knowledge graph［J］.Technology intelligence engineering，2017，3（1）：4-25 （in Chinese with English abstract） .

李岩.基于知识图谱的禽畜疾病问答系统分析与设计［D］.石家庄：河北经贸大学，2022. LI Y.Analysis and design of livestock disease question answering system based on knowledge graph［D］.Shijiazhuang：Hebei University of Economics and Business，2022 （in Chinese with English abstract） .

王雅童.基于知识图谱的兽药知识问答系统研究与实现［D］.泰安：山东农业大学，2022. WANG Y T.Research and implementation of veterinary drug knowledge question answering system based on knowledge graph［D］.Tai'an：Shandong Agricultural University，2022 （in Chinese with English abstract） .

杨喆，许甜，靳哲，等.基于知识图谱的羊群疾病问答系统的构建与实现［J］.华中农业大学学报，2023，42（3）：63-70. YANG Z，XU T，JIN Z，et al.Construction and application of knowledge graph of sheep & goat disease［J］.Journal of Huazhong Agricultural University，2023，42（3）：63-70 （in Chinese with English abstract） .

陈海辉. 基于Albert的猪病知识图谱构建研究与实现［D］. 合肥：安徽农业大学，2022. CHEN H H. Research and implementation of swine disease knowledge graph construction based on Albert［D］. Hefei： Anhui Agricultural University，2022 （in Chinese with English abstract） .

朱思强.面向猪病的命名实体识别及知识图谱构建研究与应用［D］.合肥：安徽农业大学，2023. ZHU S Q.Research and application of named entity recognition and knowledge graph construction for pig disease［D］.Hefei：Anhui Agricultural University，2023 （in Chinese with English abstract） .

PENG C，WANG X，LI Q，et al.A new Chinese named entity recognition method for pig disease domain based on lexicon-enhanced BERT and contrastive learning ［J/OL］.Applied sciences，2024，14（16）： 6944［2025-09-30］.https：//doi.org/10.3390/app14166944.

YANG S T，LIU Z H，MAYER W，et al.ShizishanGPT：an agricultural large language model integrating tools andResources［C］//Web Information Systems Engineering-WISE 2024.Singapore：Springer，2025：284-298.

沈宙锋，苏前敏，郭晶磊.基于XLNet-BiLSTM的中文电子病历命名实体识别方法［J］.智能计算机与应用，2021，11（8）：97-102.SHEN Z F，SU Q M，GUO J L.Named entity recognition model of Chinese clinical electronic medical record based on XLNet-BiLSTM［J］.Intelligent computer and applications，2021，11（8）：97-102 （in Chinese with English abstract） .

王鑫雷，李帅驰，杨志豪，等.基于预训练语言模型的中文知识图谱问答系统［J］.山西大学学报（自然科学版），2020，43（4）：955-962.WANG X L，LI S C，YANG Z H，et al.Chinese knowledge base question answering system based on pre-trained language model［J］.Journal of Shanxi University （natural science edition），2020，43（4）：955-962 （in Chinese with English abstract） .

JAFFE A，KLUGER Y，LINDENBAUM O，et al.The spectral underpinning of word2vec［J/OL］.Frontiers in applied mathematics and statistics，2020，6［2025-09-30］.https：//doi.org/10.3389/fams.2020.593406.

KAUSHIK S，CHOUDHURY A，SHERON P K，et al.AI in healthcare：time-series forecasting using statistical，neural，and ensemble architectures［J/OL］.Frontiers in big data，2020，3：4［2025-09-30］.https：// doi： 10.3389/fdata.2020.00004

ABSAR N，UDDIN N，KHANDAKER M U，et al.The efficacy of deep learning based LSTM model in forecasting the outbreak of contagious diseases［J］.Infectious disease modelling，2022，7（1）：170-183.

SKARAMAGKAS V，PENTARI A，KEFALOPOULOU Z，et al.Multi-modal deep learning diagnosis of Parkinson’s disease：a systematic review［J］.IEEE transactions on neural systems and rehabilitation engineering，2023，31：2399-2423.

CATELLI R，CASOLA V，DE PIETRO G，et al.Combining contextualized word representation and sub-document level analysis through Bi-LSTM+CRF architecture for clinical de-identification［J/OL］.Knowledge-based systems，2021，213：106649［2025-09-30］.https：//doi.org/10.1016/j.knosys.2020.106649.

DERNONCOURT F，LEE J Y，UZUNER O，et al.De-identification of patient notes with recurrent neural networks［J］.Journal of the American medical informatics association，2017，24（3）：596-606.

HUANG Z H，XU W，YU K.Bidirectional LSTM-CRF models for sequence tagging［PP/OL］.arXiv， 2015 ：1508.01991［2025-09-30］.https：//arxiv.org/abs/1508.01991.

LAMPLE G，BALLESTEROS M，SUBRAMANIAN S，et al.Neural Architectures for Named Entity Recognition ［C］ // KNIGHT K，NENKOVA A，RAMBOW O.Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics： Human Language Technologies.San Diego： Association for Computational Linguistics，2016： 260-270.

ZHAO Y S，ZHANG K L，MA H C，et al.Leveraging text skeleton for de-identification of electronic medical records［J/OL］.BMC medical informatics and decision making，2018，18（1）：18［2025-09-30］. https：//doi.org/10.1186/s12911-018-0598-6.

BAHDANAU D，CHO K，BENGIO Y.Neural machine translation by jointly learning to align and translate［PP/OL］.arXiv，2014：1409.0473［2025-09-30］.https：//arxiv.org/abs/1409.0473.

MNIH V，HEESS N，GRAVES A，et al.Recurrent models of visual attention ［C］ // Proceedings of the 28th International Conference on Neural Information Processing Systems-Volume 2.Red Hook， NY， USA： Curran Associates， Inc.，2014： 2204-2212.

DEVLIN J，CHANG M W，LEE K，et al.BERT：pre-training of deep bidirectional transformers for language understanding［C］//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics：Human Language Technologies，Volume 1 （Long and Short Papers）.Minneapolis，Minnesota： Association for Computational Linguistics，2019：4171-4186.

YANG Z，DAI Z，YANG Y，et al.XLNet： generalized autoregressive pretraining for language understanding ［C］ // Proceedings of the 33rd International Conference on Neural Information Processing Systems.Vancouver，Canada： Curran Associates，Inc.，2019： 5753-5763.

SUN Y，WANG S H，LI Y K，et al.ERNIE：enhanced representation through knowledge integration［PP/OL］.arXiv， 2019 ：1904.09223［2025-09-30］.https：//arxiv.org/abs/1904.09223.

NEO4J INC.Neo4j graph data platform ［CP/OL］.San Mateo， California： Neo4j， Inc.， 2022［2025-09-30］. https：//neo4j.com.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

渔光一体与普通池塘养殖模式下马口鱼肌肉品质的代谢组学比较

山桐子IRAP标记开发及种质的分子鉴定

3个不同倍性油茶CoSAD家族的全基因组鉴定及表达分析

复合菌剂对猪粪水重金属的钝化效果及沙地燕麦的吸收特征

荸荠秧苗U型吊杯式移栽机设计与试验