Article(id=1281692414578504180, tenantId=1146029695717560320, journalId=1281212937352253451, issueId=1281692318004646631, articleNumber=null, orderNo=null, doi=10.12133/j.smartag.SA202509011, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1756742400000, receivedDateStr=2025-09-02, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1783508884329, onlineDateStr=2026-07-08, pubDate=1774800000000, pubDateStr=2026-03-30, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1783508884329, onlineIssueDateStr=2026-07-08, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1783508884329, creator=13701087609, updateTime=1783508884329, updator=13701087609, issue=Issue{id=1281692318004646631, tenantId=1146029695717560320, journalId=1281212937352253451, year='2026', volume='8', issue='2', pageStart='1', pageEnd='278', issueExtLink='null', onlineDate='null', pubDate='1774800000000', pubDateStr='2026-03-30', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1783508861304, creator='13701087609', updateTime=1783509039471, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1281693065375101617, tenantId=1146029695717560320, journalId=1281212937352253451, issueId=1281692318004646631, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1281693065375101618, tenantId=1146029695717560320, journalId=1281212937352253451, issueId=1281692318004646631, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=265, endPage=278, ext={EN=ArticleExt(id=1281692416314946038, articleId=1281692414578504180, tenantId=1146029695717560320, journalId=1281212937352253451, language=EN, title=Key Factor Extraction Method of Agricultural User Demand Based on Large Language Models, columnId=1281692416189116917, journalTitle=Smart Agriculture, columnName=Digital Economy, runingTitle=null, highlight=null, articleAbstract=

[Objective] In the agricultural domain, user demand texts serve as essential primary sources for agricultural extension, production management, and policy services. However, these texts typically contain highly specialized terminology, exhibit non-standard, colloquial, and diverse linguistic expressions, present fragmented semantics, and rely heavily on contextual reasoning. Such characteristics make them difficult to parse accurately using traditional rule-based approaches or shallow machine learning models. Consequently, these limitations often lead to biased demand classification and incomplete extraction of key factors, thereby constraining the quality of data available for intelligent agricultural decision-making. To address these challenges, the aim of this research is to develop a robust, domain-adapted, and highly interpretable structured analysis method for agricultural user demands. [Methods] Agri-NeedAgent, an agricultural user demand analysis framework, was proposed based on a "three-stage training + multi-agent collaboration" paradigm. First, during the domain knowledge pretraining stage, 80 000 agriculture-related texts, including crop cultivation manuals, pest and disease control guides, agricultural policy documents, and farmer consultation records, were used to construct domain-specific semantic understanding, thereby enhancing the model's capability to interpret agricultural terminology, dialectal expressions, contextual logic, and implicit semantics. Second, in the instruction fine-tuning stage, 6 320 annotated samples in an "instruction-input-output" format were employed to establish an explicit mapping from raw demand texts to structured outputs. Third, in the agricultural knowledge low-rank adaptation stage, Low-rank Adaptation (LoRA) was applied to perform lightweight parameter tuning on task-specific agents, enabling targeted adaptation for demand classification and key-factor extraction tasks. Built upon the above training process, a multi-agent collaborative framework was constructed, in which the manager agent was responsible for task scheduling and quality control, while task agents were designed to perform demand classification, key-factor extraction, and explanation generation, respectively. Through this division of labor and collaborative mechanism, the framework achieved efficient and structured analysis of agricultural user demands. [Results and Discussions] Experimental results demonstrate that the proposed Agri-NeedAgent achieved a demand classification accuracy of 84.6%, a key-factor extraction F1-Score of 85.2%, a structured interface compliance rate of 94.2%, and an interpretability score of 90.2.These results showed clear improvements over traditional deep learning models such as Bidirectional Encoder Representations from Transformers (BERT) as well as general-purpose large language models (LLMs) without domain adaptation. The findings confirmed the critical role of domain knowledge injection, explicit task alignment, and multi-agent specialization in enhancing semantic understanding and structured analysis of agricultural texts. Ablation experiments further validated the effectiveness of each component. Removing domain pretraining or LoRA fine-tuning resulted in substantial performance degradation in classification and key-factor extraction, indicating the necessity of domain adaptation and task-specific optimization for handling non-standard agricultural expressions. Moreover, eliminating the manager agent or the Reasoning and Acting (ReAct) mechanism significantly reduced structured interface compliance and interpretability, highlighting the importance of task coordination, intermediate verification, and multi-step reasoning for ensuring logical consistency and output completeness. Additionally, removing the external knowledge base reduced the interpretability score from 90.2 to 77.6, underscoring its essential role in providing theoretical grounding, reasoning support, and professional explanations. Although the multi-agent collaboration introduced an additional inference overhead of approximately 140 ms, the overall per-sample inference time remained within 225 ms, meeting the real-time requirements of agricultural consultation scenarios. [Conclusions] Supported by a "three-stage training + multi-agent collaboration" framework, LLMs can effectively address challenges posed by non-standard expressions, semantic fragmentation, and multi-factor reasoning in agricultural user demand texts. The proposed method demonstrated significant improvements in demand classification, key-factor extraction, structured output compliance, and interpretability, providing high-quality and traceable structured data for intelligent agricultural decision-making. After domain adaptation and task-specific tuning, the model not only gains enhanced capability for deep semantic analysis of agricultural user demands but also ensures the completeness and interpretability of outputs through multi-agent coordination. Although the current workflow still requires optimization in terms of data preparation, staged training, and knowledge-base updating, future work will focus on expanding region-specific and emerging-technology-related demand data, developing a dynamically updated agricultural knowledge system, improving multi-agent coordination efficiency, and exploring cross-lingual agricultural demand analysis to further promote the application and deployment of agricultural large models across broader scenarios.

, authors=Runteng LI1, 3, 4, Yiqun WANG1, Hongda LI2, 3, 4, Jingchen LI3, 4, Wenbai CHEN1, authorsList=Runteng LI, Yiqun WANG, Hongda LI, Jingchen LI, Wenbai CHEN, authorCompany=null, correspAuthors=Wenbai CHEN, authorNote=

LI Runteng, E-mail:

, correspAuthorsNote=
CHEN Wenbai, E-mail:
, copyrightStatement=copyright©2025 by the authors, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1281692418114302472, articleId=1281692414578504180, tenantId=1146029695717560320, journalId=1281212937352253451, language=CN, title=基于大语言模型的农业用户主体需求关键因子提取方法, columnId=1281692416403026423, journalTitle=智慧农业(中英文), columnName=数字经济, runingTitle=null, highlight=null, articleAbstract=

【目的/意义】 在农业领域,用户需求文本是农技精准推广、政策靶向服务的核心依据,但其包含的农业专业术语、多场景差异、动态更新特征,导致传统方法存在“专业术语识别误差大、多场景需求分类模糊、动态需求响应滞后”的问题,难以支撑高效的需求分析。因此,本研究构建基于大语言模型的农业需求智能分析方法,针对性解决传统方法的上述痛点,实现农业用户需求的精准识别、结构化提取与高效解析,为农技推广、政策服务提供高质量的需求数据支撑,推动农业数字化转型的精准落地。 【方法】 在农业需求分析中引入大语言模型,利用其语义理解与推理能力实现对用户需求的精准解析。提出了“3阶段训练+多智能体协同”的技术路径:首先收集8万条农业文本与2.28万条标注样本,用于领域预训练与指令微调,以增强模型的领域适应性;随后设计多智能体协同框架,由管理型智能体负责任务调度与质量控制,任务型智能体分别承担需求分类、关键因子提取与解释生成,从而实现对农业用户需求的结构化分析。 【结果和讨论】 农业需求分析智能体框架(Agri-NeedAgent)在需求类型匹配的准确率达到84.6%,关键因子提取的F1值达到85.2%,接口合规率94.2%,可解释性90.2分,均显著优于未经过领域微调的通用大语言模型与不含多智能体协同机制的对照方法。 【结论】 大语言模型经过领域适配与微调后,能够实现对农业用户需求的深度解析与关键因子精准提取,且能通过多智能体协同机制提升分析的完整性与可解释性,为智慧农业决策支持系统提供高质量的数据支撑。

, authors=栗润腾1, 3, 4, 王一群1, 李宏达2, 3, 4, 李静晨3, 4, 陈雯柏1, authorsList=栗润腾, 王一群, 李宏达, 李静晨, 陈雯柏, authorCompany=null, correspAuthors=陈雯柏, authorNote=

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

, correspAuthorsNote=
陈雯柏,博士,教授,研究方向为模式识别与智能系统、智能科学与技术。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=bikqdzP8a40fqqI1YLydOw==, magXml=rPe6TTEeq+OhsIjRu6ctLA==, pdfUrl=null, pdf=HR+0eeDWqQivZub3hnN+wQ==, pdfFileSize=3973449, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=2QlSVRQX15lC3CwKWwQ6ow==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=dsgrAQhRLz/6HjLmlTOvSA==, mapNumber=null, fund=null)}, authors=[Author(id=1282336207627596768, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=369805210@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336207732454372, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207627596768, language=EN, stringName=Runteng LI, firstName=Runteng, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, 4, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China
3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio={"content":"

LI Runteng, E-mail:

"}, bioImg=null, bioContent=

LI Runteng, E-mail:

, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336207795368933, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207627596768, language=CN, stringName=栗润腾, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, 4, address=1.北京信息科技大学自动化学院,北京 100192,中国
3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio={"content":"

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

"}, bioImg=null, bioContent=

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)]), AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336207870866407, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336207937975273, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207870866407, language=EN, stringName=Yiqun WANG, firstName=Yiqun, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336208000889834, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207870866407, language=CN, stringName=王一群, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.北京信息科技大学自动化学院,北京 100192,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)])]), Author(id=1282336208059610092, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336208160273392, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208059610092, language=EN, stringName=Hongda LI, firstName=Hongda, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, 3, 4, address=2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China
3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336208248353777, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208059610092, language=CN, stringName=李宏达, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, 3, 4, address=2.中国农业大学信息与电气工程学院,北京 100083,中国
3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207346578390, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=2., ext=[AuthorCompanyExt(id=1282336207354966999, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China), AuthorCompanyExt(id=1282336207363355608, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中国农业大学信息与电气工程学院,北京 100083,中国)]), AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336208332239859, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336209963824118, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208332239859, language=EN, stringName=Jingchen LI, firstName=Jingchen, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336210035127287, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208332239859, language=CN, stringName=李静晨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336210110624761, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chenwb@bistu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1282336210286785531, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336210110624761, language=EN, stringName=Wenbai CHEN, firstName=Wenbai, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336210358088700, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336210110624761, language=CN, stringName=陈雯柏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.北京信息科技大学自动化学院,北京 100192,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)])])], keywords=[Keyword(id=1282336210597164029, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, orderNo=1, keyword=agricultural demand analysis), Keyword(id=1282336210681050110, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, orderNo=2, keyword=multi-agent), Keyword(id=1282336210769130495, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, orderNo=3, keyword=large language model), Keyword(id=1282336210836239360, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, orderNo=4, keyword=key factor extraction), Keyword(id=1282336210903347200, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, orderNo=5, keyword=domain fine-tuning), Keyword(id=1282336210970456065, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, orderNo=1, keyword=农业需求分析), Keyword(id=1282336211054342146, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, orderNo=2, keyword=多智能体), Keyword(id=1282336211117256707, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, orderNo=3, keyword=大语言模型), Keyword(id=1282336211205337092, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, orderNo=4, keyword=关键因子提取), Keyword(id=1282336211318583301, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, orderNo=5, keyword=领域微调)], refs=[Reference(id=1282336215693242407, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=8, pageStart=47, pageEnd=53, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=刘文, 王昊, 李啸林, journalName=技术经济与管理研究, refType=null, unstructuredReference=刘文, 王昊, 李啸林. 农业数字化转型: 研究现状、热点与趋势[J]. 技术经济与管理研究, 2025(8): 47-53., articleTitle=农业数字化转型: 研究现状、热点与趋势, refAbstract=null), Reference(id=1282336215772934184, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=8, pageStart=47, pageEnd=53, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=LIU W, WANG H, LI X L, journalName=Journal of Technical Economics & Management, refType=null, unstructuredReference=LIU W, WANG H, LI X L. Digital transformation of agriculture: Research status, hotspots and trends[J]. Journal of Technical Economics & Management, 2025(8): 47-53., articleTitle=null, refAbstract=null), Reference(id=1282336215865208873, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2021, volume=27, issue=4, pageStart=373, pageEnd=399, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=DAYIOĞLU MALI, TURKER U, journalName=Tarım Bilimleri Dergisi, refType=null, unstructuredReference=DAYIOĞLU MALI, TURKER U. Digital transformation for sustainable future - agriculture 4.0: A review[J]. Tarım Bilimleri Dergisi, 2021, 27(4): 373-399., articleTitle=Digital transformation for sustainable future - agriculture 4.0: A review, refAbstract=null), Reference(id=1282336215940706346, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2021, volume=2, issue=5, pageStart=377, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=SARKER I H, journalName=SN Computer Science, refType=null, unstructuredReference=SARKER I H. Data science and analytics: An overview from data-driven smart computing, decision-making and applications perspective[J]. SN Computer Science, 2021, 2(5): 377., articleTitle=Data science and analytics: An overview from data-driven smart computing, decision-making and applications perspective, refAbstract=null), Reference(id=1282336216007815211, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2021, volume=2021, issue=1, pageStart=5584754, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=AYED RBEN, HANANA M, journalName=Journal of Food Quality, refType=null, unstructuredReference=AYED RBEN, HANANA M. Artificial intelligence to improve the food and agriculture sector[J]. Journal of Food Quality, 2021, 2021(1): 5584754., articleTitle=Artificial intelligence to improve the food and agriculture sector, refAbstract=null), Reference(id=1282336216079118380, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=2023, issue=1, pageStart=6916213, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=SHARMA S, SHARMA C, ASENSO E, journalName=Journal of Sensors, refType=null, unstructuredReference=SHARMA S, SHARMA C, ASENSO E, et al. Research constituents and trends in smart farming: An analytical retrospection from the lens of text mining[J]. Journal of Sensors, 2023, 2023(1): 6916213., articleTitle=Research constituents and trends in smart farming: An analytical retrospection from the lens of text mining, refAbstract=null), Reference(id=1282336216146227245, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=56, issue=10, pageStart=200, pageEnd=222, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=李孝鹏, 向玉云, 张培君, journalName=农业机械学报, refType=null, unstructuredReference=李孝鹏, 向玉云, 张培君, . 农业领域自然语言理解技术应用综述[J]. 农业机械学报, 2025, 56(10): 200-222., articleTitle=农业领域自然语言理解技术应用综述, refAbstract=null), Reference(id=1282336216209141806, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=56, issue=10, pageStart=200, pageEnd=222, url=null, language=null, rfNumber=[6], rfOrder=7, authorNames=LI X P, XIANG Y Y, ZHANG P J, journalName=Transactions of the Chinese Society for Agricultural Machinery, refType=null, unstructuredReference=LI X P, XIANG Y Y, ZHANG P J, et al. Natural language understanding in agriculture: A comprehensive review of technologies and applications[J]. Transactions of the Chinese Society for Agricultural Machinery, 2025, 56(10): 200-222., articleTitle=null, refAbstract=null), Reference(id=1282336216301416495, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=11, issue=null, pageStart=111255, pageEnd=111264, url=null, language=null, rfNumber=[7], rfOrder=8, authorNames=SHARMA P, DADHEECH P, ANEJA N, journalName=IEEE Access, refType=null, unstructuredReference=SHARMA P, DADHEECH P, ANEJA N, et al. Predicting agriculture yields based on machine learning using regression and deep learning[J]. IEEE Access, 2023, 11: 111255-111264., articleTitle=Predicting agriculture yields based on machine learning using regression and deep learning, refAbstract=null), Reference(id=1282336216376913968, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=56, issue=9, pageStart=240, pageEnd=256, url=null, language=null, rfNumber=[8], rfOrder=9, authorNames=王耀君, 徐国威, 朱建军, journalName=农业机械学报, refType=null, unstructuredReference=王耀君, 徐国威, 朱建军, . 农业领域大语言模型研究进展[J]. 农业机械学报, 2025, 56(9): 240-256., articleTitle=农业领域大语言模型研究进展, refAbstract=null), Reference(id=1282336216452411441, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=56, issue=9, pageStart=240, pageEnd=256, url=null, language=null, rfNumber=[8], rfOrder=10, authorNames=WANG Y J, XU G W, ZHU J J, journalName=Transactions of the Chinese Society for Agricultural Machinery, refType=null, unstructuredReference=WANG Y J, XU G W, ZHU J J, et al. Survey of research on large language models in agriculture[J]. Transactions of the Chinese Society for Agricultural Machinery, 2025, 56(9): 240-256., articleTitle=null, refAbstract=null), Reference(id=1282336216544686130, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=56, issue=5, pageStart=4403, pageEnd=4445, url=null, language=null, rfNumber=[9], rfOrder=11, authorNames=NISMI MOL E A, SANTOSH KUMAR M B, journalName=Artificial Intelligence Review, refType=null, unstructuredReference=NISMI MOL E A, SANTOSH KUMAR M B. Review on knowledge extraction from text and scope in agriculture domain[J]. Artificial Intelligence Review, 2023, 56(5): 4403-4445., articleTitle=Review on knowledge extraction from text and scope in agriculture domain, refAbstract=null), Reference(id=1282336216615989299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=12, authorNames=CHATTERJEE N, KAUSHIK N, journalName=null, refType=null, unstructuredReference=CHATTERJEE N, KAUSHIK N. Automatic extraction of agriculture terms from domain text: A survey of tools and techniques[EB/OL]. arXiv: 2020., articleTitle=Automatic extraction of agriculture terms from domain text: A survey of tools and techniques, refAbstract=null), Reference(id=1282336216699875380, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2021, volume=191, issue=null, pageStart=106546, pageEnd=null, url=null, language=null, rfNumber=[11], rfOrder=13, authorNames=KOK Z H, MOHAMED SHARIFF A R, ALFATNI M S M, journalName=Computers and Electronics in Agriculture, refType=null, unstructuredReference=KOK Z H, MOHAMED SHARIFF A R, ALFATNI M S M, et al. Support vector machine in precision agriculture: A review[J]. Computers and Electronics in Agriculture, 2021, 191: 106546., articleTitle=Support vector machine in precision agriculture: A review, refAbstract=null), Reference(id=1282336216783761461, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2022, volume=15, issue=1, pageStart=1, pageEnd=17, url=null, language=null, rfNumber=[12], rfOrder=14, authorNames=SHI L, QIN Y Q, ZHANG J J, journalName=Journal of Information Technology Research, refType=null, unstructuredReference=SHI L, QIN Y Q, ZHANG J J, et al. Multi-class classification of agricultural data based on random forest and feature selection[J]. Journal of Information Technology Research, 2022, 15(1): 1-17., articleTitle=Multi-class classification of agricultural data based on random forest and feature selection, refAbstract=null), Reference(id=1282336216846676022, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=4171, pageEnd=4186, url=null, language=null, rfNumber=[13], rfOrder=15, authorNames=DEVLIN J, CHANG M W, LEE K, journalName=null, refType=null, unstructuredReference=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, USA: Association for Computational Linguistics, 2019: 4171-4186., articleTitle=Bert: Pre-training of deep bidirectional transformers for language understanding, refAbstract=null), Reference(id=1282336216909590583, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=234, issue=null, pageStart=121103, pageEnd=null, url=null, language=null, rfNumber=[14], rfOrder=16, authorNames=LIU Y F, WEI S Q, HUANG H J, journalName=Expert Systems with Applications, refType=null, unstructuredReference=LIU Y F, WEI S Q, HUANG H J, et al. Naming entity recognition of citrus pests and diseases based on the BERT-BiLSTM-CRF model[J]. Expert Systems with Applications, 2023, 234: 121103., articleTitle=Naming entity recognition of citrus pests and diseases based on the BERT-BiLSTM-CRF model, refAbstract=null), Reference(id=1282336217022836792, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=5, issue=4, pageStart=105, pageEnd=116, url=null, language=null, rfNumber=[15], rfOrder=17, authorNames=王婷, 王娜, 崔运鹏, journalName=智慧农业(中英文), refType=null, unstructuredReference=王婷, 王娜, 崔运鹏, . 基于人工智能大模型技术的果蔬农技知识智能问答系统[J]. 智慧农业(中英文), 2023, 5(4): 105-116., articleTitle=基于人工智能大模型技术的果蔬农技知识智能问答系统, refAbstract=null), Reference(id=1282336217098334265, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=5, issue=4, pageStart=105, pageEnd=116, url=null, language=null, rfNumber=[15], rfOrder=18, authorNames=WANG T, WANG N, CUI Y P, journalName=Smart Agriculture, refType=null, unstructuredReference=WANG T, WANG N, CUI Y P, et al. Agricultural technology knowledge intelligent question-answering system based on large language model[J]. Smart Agriculture, 2023, 5(4): 105-116., articleTitle=null, refAbstract=null), Reference(id=1282336217186414650, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2024, volume=6, issue=2, pageStart=1, pageEnd=13, url=null, language=null, rfNumber=[16], rfOrder=19, authorNames=郭旺, 杨雨森, 吴华瑞, journalName=智慧农业(中英文), refType=null, unstructuredReference=郭旺, 杨雨森, 吴华瑞, . 农业大模型:关键技术、应用分析与发展方向[J]. 智慧农业(中英文), 2024, 6(2): 1-13., articleTitle=农业大模型:关键技术、应用分析与发展方向, refAbstract=null), Reference(id=1282336217249329211, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2024, volume=6, issue=2, pageStart=1, pageEnd=13, url=null, language=null, rfNumber=[16], rfOrder=20, authorNames=GUO W, YANG Y S, WU H R, journalName=Smart Agriculture, refType=null, unstructuredReference=GUO W, YANG Y S, WU H R, et al. Big models in agriculture: Key technologies, application and future directions[J]. Smart Agriculture, 2024, 6(2): 1-13., articleTitle=null, refAbstract=null), Reference(id=1282336218876719164, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=6, pageStart=81, pageEnd=83, url=null, language=null, rfNumber=[17], rfOrder=21, authorNames=李小玲, journalName=中国农机装备, refType=null, unstructuredReference=李小玲. 智慧农业领域人工智能大模型的应用研究[J]. 中国农机装备, 2025(6): 81-83., articleTitle=智慧农业领域人工智能大模型的应用研究, refAbstract=null), Reference(id=1282336218985771069, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=6, pageStart=81, pageEnd=83, url=null, language=null, rfNumber=[17], rfOrder=22, authorNames=LI X L, journalName=China Agricultural Machinery Equipment, refType=null, unstructuredReference=LI X L. Research on application of large language models in smart agriculture[J]. China Agricultural Machinery Equipment, 2025(6): 81-83., articleTitle=null, refAbstract=null), Reference(id=1282336219069657150, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=2018, url=https://gwern.net/doc/www/s3-us-west-2.amazonaws.com/d73fdc5ffa8627bce44dcda2fc012da638ffb158.pdf, language=null, rfNumber=[18], rfOrder=23, authorNames=RADFORD A, NARASIMHAN K, SALIMANS T, journalName=null, refType=null, unstructuredReference=RADFORD A, NARASIMHAN K, SALIMANS T, et al. Improving language understanding by generative pre-training[J/OL]. 2018.[2025-08-26]., articleTitle=Improving language understanding by generative pre-training, refAbstract=null), Reference(id=1282336219157737535, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=7, issue=1, pageStart=20, pageEnd=32, url=null, language=null, rfNumber=[19], rfOrder=24, authorNames=姜京池, 闫莲, 刘劼, journalName=智慧农业(中英文), refType=null, unstructuredReference=姜京池, 闫莲, 刘劼. 基于精准知识筛选及知识协同生成的农业大语言模型[J]. 智慧农业(中英文), 2025,7(1): 20-32., articleTitle=基于精准知识筛选及知识协同生成的农业大语言模型, refAbstract=null), Reference(id=1282336219233235008, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=7, issue=1, pageStart=20, pageEnd=32, url=null, language=null, rfNumber=[19], rfOrder=25, authorNames=JIANG J C, YAN L, LIU J, journalName=Smart Agriculture, refType=null, unstructuredReference=JIANG J C, YAN L, LIU J. Agricultural large language model based on precise knowledge retrieval and knowledge collaborative generation[J]. Smart Agriculture, 2025,7(1): 20-32., articleTitle=null, refAbstract=null), Reference(id=1282336219296149569, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2025, volume=13, issue=null, pageStart=18912, pageEnd=18936, url=null, language=null, rfNumber=[20], rfOrder=26, authorNames=ACHARYA D B, KUPPAN K, DIVYA B, journalName=IEEE Access, refType=null, unstructuredReference=ACHARYA D B, KUPPAN K, DIVYA B. Agentic AI: Autonomous intelligence for complex goals: A comprehensive survey[J]. IEEE Access, 2025, 13: 18912-18936., articleTitle=Agentic AI: Autonomous intelligence for complex goals: A comprehensive survey, refAbstract=null), Reference(id=1282336219388424258, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2024, volume=20, issue=6, pageStart=92, pageEnd=106, url=null, language=null, rfNumber=[21], rfOrder=27, authorNames=任荣荣, 胡崇宇, 吴国龙, journalName=农业展望, refType=null, unstructuredReference=任荣荣, 胡崇宇, 吴国龙, . 农业种植智能体(Agri-agent)的构建与应用展望[J]. 农业展望, 2024, 20(6): 92-106., articleTitle=农业种植智能体(Agri-agent)的构建与应用展望, refAbstract=null), Reference(id=1282336219463921731, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2024, volume=20, issue=6, pageStart=92, pageEnd=106, url=null, language=null, rfNumber=[21], rfOrder=28, authorNames=REN R R, HU C Y, WU G L, journalName=Agricultural Outlook, refType=null, unstructuredReference=REN R R, HU C Y, WU G L, et al. Construction and application prospect of agricultural planting agent (Agri-agent)[J]. Agricultural Outlook, 2024, 20(6): 92-106., articleTitle=null, refAbstract=null), Reference(id=1282336219547807812, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[22], rfOrder=29, authorNames=HONG S R, ZHUGE M C, CHEN J Q, journalName=null, refType=null, unstructuredReference=HONG S R, ZHUGE M C, CHEN J Q, et al. MetaGPT: Meta programming for a multi-agent collaborative framework[EB/OL]. arXiv: 2023., articleTitle=MetaGPT: Meta programming for a multi-agent collaborative framework, refAbstract=null), Reference(id=1282336219627499589, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[23], rfOrder=30, authorNames=WU Q Y, BANSAL G, ZHANG J Y, journalName=null, refType=null, unstructuredReference=WU Q Y, BANSAL G, ZHANG J Y, et al. AutoGen: Enabling next-gen LLM applications via multi-agent conversation[EB/OL]. arXiv: 2023., articleTitle=AutoGen: Enabling next-gen LLM applications via multi-agent conversation, refAbstract=null), Reference(id=1282336219707191366, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=24824, pageEnd=24837, url=null, language=null, rfNumber=[24], rfOrder=31, authorNames=WEI J, WANG X Z, SCHUURMANS D, journalName=null, refType=null, unstructuredReference=WEI J, WANG X Z, SCHUURMANS D, et al. Chain-of-thought prompting elicits reasoning in large language models[C]// Proceedings of the 36th International Conference on Neural Information Processing Systems. New York, USA: ACM, 2022: 24824-24837., articleTitle=Chain-of-thought prompting elicits reasoning in large language models, refAbstract=null)], funds=[Fund(id=1282336215185731617, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=2021ZD0113603, language=EN, fundingSource=Major Project of Scientific and Technological Innovation 2030(2021ZD0113603), fundOrder=null, country=null), Fund(id=1282336215252840482, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=2021ZD0113603, language=CN, fundingSource=科技部2030新一代人工智能重大专项(2021ZD0113603), fundOrder=null, country=null), Fund(id=1282336215311560739, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=62276028, language=EN, fundingSource=National Natural Science Foundation of China(62276028), fundOrder=null, country=null), Fund(id=1282336215378669604, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=62276028, language=CN, fundingSource=国家自然科学基金(62276028), fundOrder=null, country=null), Fund(id=1282336215445778469, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=92267110, language=EN, fundingSource=Major Research Plan of National Natural Science Foundation of China(92267110), fundOrder=null, country=null), Fund(id=1282336215521275942, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, awardId=92267110, language=CN, fundingSource=国家自然科学基金重大研究计划(92267110), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)]), AuthorCompany(id=1282336207346578390, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=2., ext=[AuthorCompanyExt(id=1282336207354966999, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China), AuthorCompanyExt(id=1282336207363355608, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中国农业大学信息与电气工程学院,北京 100083,中国)]), AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])], figs=[ArticleFig(id=1282336211507326982, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 1, caption=Diagram of the analysis system for key factor extraction of agricultural user subject demand, figureFileSmall=cYa+22SkiXhHbyl/oS9JTg==, figureFileBig=vfaEV6KA+24MMO0R6dVo9g==, tableContent=null), ArticleFig(id=1282336211587018759, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图1, caption=农业用户主体需求关键因子提取分析体系图, figureFileSmall=cYa+22SkiXhHbyl/oS9JTg==, figureFileBig=vfaEV6KA+24MMO0R6dVo9g==, tableContent=null), ArticleFig(id=1282336211679293448, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 2, caption=Diagram of data distribution proportion for seven major categories of agricultural user demand labels, figureFileSmall=LIeEpIiB9IN6wzye7duirQ==, figureFileBig=HxBsIMzKXPdFa8qmZ0rT/g==, tableContent=null), ArticleFig(id=1282336211742208009, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图2, caption=农业用户需求七大类标签数据分布占比图, figureFileSmall=LIeEpIiB9IN6wzye7duirQ==, figureFileBig=HxBsIMzKXPdFa8qmZ0rT/g==, tableContent=null), ArticleFig(id=1282336211813511178, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 3, caption=Overall framework design of Agri-NeedAgent for key factor extraction of agricultural user demand, figureFileSmall=vkml4yLnjW/21SKRYGTvEQ==, figureFileBig=I1fZX0CiyFENWFq06syMGw==, tableContent=null), ArticleFig(id=1282336211876425739, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图3, caption=Agri-NeedAgent农业用户需求关键因子提取整体框架设计, figureFileSmall=vkml4yLnjW/21SKRYGTvEQ==, figureFileBig=I1fZX0CiyFENWFq06syMGw==, tableContent=null), ArticleFig(id=1282336211947728908, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 4, caption=Training workflow of the LLM-based method for key-factor extraction of agricultural user demands, figureFileSmall=S4a8KaZqqWy8YloYy7wP6Q==, figureFileBig=7QC5nZSwS4v4pxOjQ+Wl/A==, tableContent=null), ArticleFig(id=1282336212044197901, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图4, caption=基于LLMs的农业用户需求关键因子提取方法训练流程图, figureFileSmall=S4a8KaZqqWy8YloYy7wP6Q==, figureFileBig=7QC5nZSwS4v4pxOjQ+Wl/A==, tableContent=null), ArticleFig(id=1282336212111306766, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 5, caption=Schematic diagram of LoRA model training, figureFileSmall=f6KsgQEC12qmkGYk4bO6Gg==, figureFileBig=+Rt0U6NErhULWVzJEGvISw==, tableContent=null), ArticleFig(id=1282336212182609935, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图5, caption=LoRA模型训练示意图, figureFileSmall=f6KsgQEC12qmkGYk4bO6Gg==, figureFileBig=+Rt0U6NErhULWVzJEGvISw==, tableContent=null), ArticleFig(id=1282336212249718800, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 6, caption=Multi-agent collaboration workflow and MA decision logic diagram, figureFileSmall=GbR5jpVfWBoDmHDeJqb/fA==, figureFileBig=/Tq7DAPU2+DJKmAZ6w1Nmw==, tableContent=null), ArticleFig(id=1282336212337799185, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图6, caption=多智能体协作流程与MA决策逻辑图, figureFileSmall=GbR5jpVfWBoDmHDeJqb/fA==, figureFileBig=/Tq7DAPU2+DJKmAZ6w1Nmw==, tableContent=null), ArticleFig(id=1282336212400713746, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 7, caption=Training and validation loss curves of Agri-NeedAgent in the second and third stages, figureFileSmall=pAX1YTgIG2psSs4Ef1eucg==, figureFileBig=fA1PLwIVhllOe+12yA7eIw==, tableContent=null), ArticleFig(id=1282336212476211219, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图7, caption=Agri-NeedAgent第二、三阶段训练与验证损失曲线

a. 第二阶段(指令微调)训练与验证损失曲线 b. 第三阶段(LoRA 低秩适配)训练与验证损失曲线

, figureFileSmall=pAX1YTgIG2psSs4Ef1eucg==, figureFileBig=fA1PLwIVhllOe+12yA7eIw==, tableContent=null), ArticleFig(id=1282336212543320084, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Fig. 8, caption=Extraction results of Agri-NeedAgent and baseline models on the agricultural user demand analysis task, figureFileSmall=HMJb2e1qWLewTtRxqnX1eg==, figureFileBig=ZMzHazJSGdHsTbqCB6Oxwg==, tableContent=null), ArticleFig(id=1282336212606234645, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=图8, caption=Agri-NeedAgent与基线模型在农业用户需求分析任务中的提取结果对比, figureFileSmall=HMJb2e1qWLewTtRxqnX1eg==, figureFileBig=ZMzHazJSGdHsTbqCB6Oxwg==, tableContent=null), ArticleFig(id=1282336212681732118, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Table 1, caption=

Statistics of fine-tuning datasets for the second and third stages in agricultural user demand analysis

, figureFileSmall=null, figureFileBig=null, tableContent=
来源数据类型文本数量格式时间
合计22 800
农业论坛文本14 000JSON2022—2025
用户咨询记录文本6 500JSON2021—2025
专家访谈文本文本2 300JSON2022—2024
), ArticleFig(id=1282336212765618199, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=表1, caption=

农业用户需求分析二、三阶段微调数据集统计

, figureFileSmall=null, figureFileBig=null, tableContent=
来源数据类型文本数量格式时间
合计22 800
农业论坛文本14 000JSON2022—2025
用户咨询记录文本6 500JSON2021—2025
专家访谈文本文本2 300JSON2022—2024
), ArticleFig(id=1282336212853698584, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Table 2, caption=

Statistics of data in the multi-source heterogeneous knowledge base for agricultural user demand analysis

, figureFileSmall=null, figureFileBig=null, tableContent=
数据源数据类型实体/记录数量格式
农业书籍文本300本JSON
农技问答文本5 000条问答对JSON
研究论文文本150篇论文JSON
农业政策文件文本220份PDF
), ArticleFig(id=1282336214447534106, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=表2, caption=

农业用户需求分析多源异构知识库数据统计

, figureFileSmall=null, figureFileBig=null, tableContent=
数据源数据类型实体/记录数量格式
农业书籍文本300本JSON
农技问答文本5 000条问答对JSON
研究论文文本150篇论文JSON
农业政策文件文本220份PDF
), ArticleFig(id=1282336214577557531, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Table 3, caption=

Five-level scoring rubric for the“semantic consistency”dimension in the interpretability analysis of agricultural user demand analysis

, figureFileSmall=null, figureFileBig=null, tableContent=
评分档位分数范围农业场景示例(对应本研究需求类型)
优秀90~100完全覆盖用户需求核心要素,如“黄土高原种植节水苹果”需求中,解释文本明确包含:黄土高原(区域)、节水(诉求)、苹果(作物)所有关键信息
良好70~89覆盖核心要素但表述略有偏差,如遗漏“节水”的具体表述,仅提及“黄土高原地区苹果品种选择”
合格50~69部分覆盖核心要素,如仅提及“苹果品种选择”,未包含“黄土高原”区域信息
不足30~49核心要素偏差,如将“苹果”误表述为“梨”,需求指向错误
缺失0~29无需求语义关联,如解释文本与“黄土高原种植节水苹果”完全无关,仅重复通用农业术语
), ArticleFig(id=1282336214657249308, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=表3, caption=

农业用户主体需求分析可解释性“语义一致性”维度5档评分锚定表

, figureFileSmall=null, figureFileBig=null, tableContent=
评分档位分数范围农业场景示例(对应本研究需求类型)
优秀90~100完全覆盖用户需求核心要素,如“黄土高原种植节水苹果”需求中,解释文本明确包含:黄土高原(区域)、节水(诉求)、苹果(作物)所有关键信息
良好70~89覆盖核心要素但表述略有偏差,如遗漏“节水”的具体表述,仅提及“黄土高原地区苹果品种选择”
合格50~69部分覆盖核心要素,如仅提及“苹果品种选择”,未包含“黄土高原”区域信息
不足30~49核心要素偏差,如将“苹果”误表述为“梨”,需求指向错误
缺失0~29无需求语义关联,如解释文本与“黄土高原种植节水苹果”完全无关,仅重复通用农业术语
), ArticleFig(id=1282336214724358173, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Table 4, caption=

Extraction results of various models/methods in the agricultural user demand analysis task

, figureFileSmall=null, figureFileBig=null, tableContent=
模型/方法需求类型分析准确率/%关键因子提取F1值/%合规率/%可解释性
BERT59.8±0.565.2±0.450.2±0.666.6±1.4
Qwen3-1.7B56.5±0.461.1±0.578.7±0.377.9±0.8
Qwen3-4B72.5±0.371.8±0.486.6±0.283.3±0.7
DeepSeek-R1:1.5B55.2±0.560.7±0.477.6±0.376.1±0.9
DeepSeek-R1:7B75.6±0.376.4±0.390.3±0.282.7±0.6
ChatGLM3-6B70.5±0.371.2±0.483.8±0.281.5±0.6
Baichuan2-7B-Base74.2±0.374.1±0.487.5±0.282.2±0.6
InternLM2.5-7B77.8±0.476.0±0.390.8±0.383.7±0.6
Agri-NeedAgent84.6±0.285.2±0.294.2±0.190.2±0.5
), ArticleFig(id=1282336214804049950, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=表4, caption=

各模型/方法在农业用户需求分析任务中的提取结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型/方法需求类型分析准确率/%关键因子提取F1值/%合规率/%可解释性
BERT59.8±0.565.2±0.450.2±0.666.6±1.4
Qwen3-1.7B56.5±0.461.1±0.578.7±0.377.9±0.8
Qwen3-4B72.5±0.371.8±0.486.6±0.283.3±0.7
DeepSeek-R1:1.5B55.2±0.560.7±0.477.6±0.376.1±0.9
DeepSeek-R1:7B75.6±0.376.4±0.390.3±0.282.7±0.6
ChatGLM3-6B70.5±0.371.2±0.483.8±0.281.5±0.6
Baichuan2-7B-Base74.2±0.374.1±0.487.5±0.282.2±0.6
InternLM2.5-7B77.8±0.476.0±0.390.8±0.383.7±0.6
Agri-NeedAgent84.6±0.285.2±0.294.2±0.190.2±0.5
), ArticleFig(id=1282336214929879071, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=EN, label=Table 5, caption=

Ablation experiment results of Agri-NeedAgent

, figureFileSmall=null, figureFileBig=null, tableContent=
系统组件需求类型分析准确率/%关键因子提取F1值/%合规率/%可解释性
全模块(Agri-NeedAgent)84.6±0.285.2±0.294.2±0.190.2±0.5
移除ReAct思维链82.3±0.382.6±0.487.1±0.487.2±0.5
移除少样本提示81.5±0.281.7±0.286.6±0.287.5±0.6
移除MA82.0±0.481.9±0.581.4±0.786.4±0.7
没有领域预训练78.8±0.379.5±0.491.3±0.288.6±0.6
没有指令微调80.4±0.280.6±0.290.8±0.287.4±0.5
没有LoRA微调77.5±0.276.2±0.292.6±0.288.9±0.5
移除外部知识库83.7±0.284.4±0.293.9±0.177.6±0.9
), ArticleFig(id=1282336215005376544, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, language=CN, label=表5, caption=

Agri-NeedAgent的消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
系统组件需求类型分析准确率/%关键因子提取F1值/%合规率/%可解释性
全模块(Agri-NeedAgent)84.6±0.285.2±0.294.2±0.190.2±0.5
移除ReAct思维链82.3±0.382.6±0.487.1±0.487.2±0.5
移除少样本提示81.5±0.281.7±0.286.6±0.287.5±0.6
移除MA82.0±0.481.9±0.581.4±0.786.4±0.7
没有领域预训练78.8±0.379.5±0.491.3±0.288.6±0.6
没有指令微调80.4±0.280.6±0.290.8±0.287.4±0.5
没有LoRA微调77.5±0.276.2±0.292.6±0.288.9±0.5
移除外部知识库83.7±0.284.4±0.293.9±0.177.6±0.9
)], attaches=null, journal=Journal(id=1281211802411339782, delFlag=0, nameCn=智慧农业(中英文), nameEn=Smart Agriculture, nameHistory1=null, nameHistory2=null, issn=2096-8094, eissn=2097-485X, cn=10-1681/S, coden=ZNZHD7, periodic=1, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=3ySu/Wank9ycKptCQ9bxHg==, journalPrice=null, startedYear=null, abbrevIsoEn=Smart Agriculture, journalRemark=null, publicationField=null, createdTime=1783394297461, updatedTime=1784022346804, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=Z, firstLetterEn=Z, subjectCode=Life Sciences, subjectName=null, subjectCodeEn=Life Sciences, subjectNameEn=null, picCn=3ySu/Wank9ycKptCQ9bxHg==, picEn=6aR2ETcq7WHPaEneIXN2aw==, jcr=null, cjcr=null, exts=[JournalExt(id=1283846032387719244, language=CN, name=智慧农业(中英文), nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1784022346827, updatedTime=1784022346827, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.smartag.net.cn/Journalx_zhny/authorLogOn.action, submissionEditorUrl=http://www.smartag.net.cn/Journalx_zhny/editorLogOn.action, submissionReviewUrl=http://www.smartag.net.cn/Journalx_zhny/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1283846032433856589, language=EN, name=Smart Agriculture, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1784022346838, updatedTime=1784022346838, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.smartag.net.cn/Journalx_zhny/authorLogOn.action, submissionEditorUrl=http://www.smartag.net.cn/Journalx_zhny/editorLogOn.action, submissionReviewUrl=http://www.smartag.net.cn/Journalx_zhny/expertLogOn.action, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1281212937352253451, websiteList=[Website(id=1281216327134188517, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1281212937352253451, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zhny/CN, language=CN, createTime=1783395376237, createBy=18614031015, updateTime=1783395541424, updateBy=18614031015, name=智慧农业(中英文)-中文, tplId=1146099689490845704, title=智慧农业(中英文), delFlag=0, indexPage=/home, props=[WebsiteProps(id=1281217159418319048, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=articleTextType, value=kx, createTime=1783395574669, updateTime=1783395574669, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159388958917, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=banner, value=null, createTime=1783395574662, updateTime=1783395574662, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159443484875, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=grayFlag, value=0, createTime=1783395574675, updateTime=1783395574675, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159355404484, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=logo, value=https://castjournals.cast.org.cn/joweb/zhny/CN/file/pic?fileId=RbfWW55ASXpJGAr1dytAUA==, createTime=1783395574654, updateTime=1783395574654, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159456067789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=minRunFlag, value=0, createTime=1783395574678, updateTime=1783395574678, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159414124743, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zhny/CN/file/pic, createTime=1783395574668, updateTime=1783395574668, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159451873484, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=silenceFlag, value=0, createTime=1783395574677, updateTime=1783395574677, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159401541830, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1783395574665, updateTime=1783395574665, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159426707657, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=themeColor, value=null, createTime=1783395574671, updateTime=1783395574671, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217159435096266, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327134188517, code=themeStyle, value=null, createTime=1783395574673, updateTime=1783395574673, creator=18614031015, updator=18614031015)]), Website(id=1281216327880774635, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1281212937352253451, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zhny/EN, language=EN, createTime=1783395376415, createBy=18614031015, updateTime=1783395537123, updateBy=18614031015, name=智慧农业(中英文)-英文, tplId=1146101810881728533, title=Smart Agriculture, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1281217117114568888, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=articleTextType, value=kx, createTime=1783395564583, updateTime=1783395564583, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117089403061, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=banner, value=null, createTime=1783395564577, updateTime=1783395564577, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117135540411, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=grayFlag, value=0, createTime=1783395564588, updateTime=1783395564588, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117064237236, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=logo, value=https://castjournals.cast.org.cn/joweb/zhny/EN/file/pic?fileId=RbfWW55ASXpJGAr1dytAUA==, createTime=1783395564571, updateTime=1783395564571, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117148123325, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=minRunFlag, value=0, createTime=1783395564591, updateTime=1783395564591, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117106180279, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zhny/EN/file/pic, createTime=1783395564581, updateTime=1783395564581, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117139734716, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=silenceFlag, value=0, createTime=1783395564589, updateTime=1783395564589, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117097791670, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1783395564579, updateTime=1783395564579, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117122957497, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=themeColor, value=null, createTime=1783395564585, updateTime=1783395564585, creator=18614031015, updator=18614031015), WebsiteProps(id=1281217117127151802, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1281216327880774635, code=themeStyle, value=null, createTime=1783395564586, updateTime=1783395564586, creator=18614031015, updator=18614031015)])], journalTitle=智慧农业(中英文), weixinUrl=null, journalUrl=https://www.smartag.net.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Smart Agriculture, journalPhotoCn=3ySu/Wank9ycKptCQ9bxHg==, journalPhotoEn=6aR2ETcq7WHPaEneIXN2aw==, journalFirstLetter=Z, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/zhny/CN/10.12133/j.smartag.SA202509011, detailUrlEn=https://castjournals.cast.org.cn/joweb/zhny/EN/10.12133/j.smartag.SA202509011, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zhny/CN/PDF/10.12133/j.smartag.SA202509011, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zhny/EN/PDF/10.12133/j.smartag.SA202509011, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1774800000000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于大语言模型的农业用户主体需求关键因子提取方法
收藏切换
PDF下载
栗润腾 1, 3, 4 , 王一群 1 , 李宏达 2, 3, 4 , 李静晨 3, 4 , 陈雯柏 1
智慧农业(中英文) | 数字经济 2026,8(2): 265-278
收起
收藏切换
智慧农业(中英文) |数字经济 2026 , 8 (2) : 265 -278
基于大语言模型的农业用户主体需求关键因子提取方法
全屏
[Author(id=1282336207627596768, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=369805210@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336207732454372, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207627596768, language=EN, stringName=Runteng LI, firstName=Runteng, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, 4, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China
3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio={"content":"

LI Runteng, E-mail:

"}, bioImg=null, bioContent=

LI Runteng, E-mail:

, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336207795368933, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207627596768, language=CN, stringName=栗润腾, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, 4, address=1.北京信息科技大学自动化学院,北京 100192,中国
3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio={"content":"

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

"}, bioImg=null, bioContent=

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)]), AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336207870866407, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336207937975273, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207870866407, language=EN, stringName=Yiqun WANG, firstName=Yiqun, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336208000889834, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336207870866407, language=CN, stringName=王一群, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.北京信息科技大学自动化学院,北京 100192,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)])]), Author(id=1282336208059610092, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336208160273392, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208059610092, language=EN, stringName=Hongda LI, firstName=Hongda, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, 3, 4, address=2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China
3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336208248353777, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208059610092, language=CN, stringName=李宏达, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, 3, 4, address=2.中国农业大学信息与电气工程学院,北京 100083,中国
3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207346578390, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=2., ext=[AuthorCompanyExt(id=1282336207354966999, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China), AuthorCompanyExt(id=1282336207363355608, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207346578390, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中国农业大学信息与电气工程学院,北京 100083,中国)]), AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336208332239859, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1282336209963824118, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208332239859, language=EN, stringName=Jingchen LI, firstName=Jingchen, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336210035127287, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336208332239859, language=CN, stringName=李静晨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.国家农业信息化工程技术研究中心,北京 100097,中国
4.北京市农林科学院信息技术研究中心,北京 100079,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207443047385, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=3., ext=[AuthorCompanyExt(id=1282336207447241690, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China), AuthorCompanyExt(id=1282336207455630299, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207443047385, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.国家农业信息化工程技术研究中心,北京 100097,中国)]), AuthorCompany(id=1282336207535322076, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=4., ext=[AuthorCompanyExt(id=1282336207539516381, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China), AuthorCompanyExt(id=1282336207547904990, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207535322076, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.北京市农林科学院信息技术研究中心,北京 100079,中国)])]), Author(id=1282336210110624761, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chenwb@bistu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1282336210286785531, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336210110624761, language=EN, stringName=Wenbai CHEN, firstName=Wenbai, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1282336210358088700, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, authorId=1282336210110624761, language=CN, stringName=陈雯柏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.北京信息科技大学自动化学院,北京 100192,中国, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1282336207262692307, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, xref=1., ext=[AuthorCompanyExt(id=1282336207275275220, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China), AuthorCompanyExt(id=1282336207283663829, tenantId=1146029695717560320, journalId=1281212937352253451, articleId=1281692414578504180, companyId=1282336207262692307, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.北京信息科技大学自动化学院,北京 100192,中国)])])]
栗润腾1, 3, 4 , 王一群1, 李宏达2, 3, 4, 李静晨3, 4, 陈雯柏1
作者信息
  • 1.北京信息科技大学自动化学院,北京 100192,中国
  • 2.中国农业大学信息与电气工程学院,北京 100083,中国
  • 3.国家农业信息化工程技术研究中心,北京 100097,中国
  • 4.北京市农林科学院信息技术研究中心,北京 100079,中国
通讯作者:
陈雯柏,博士,教授,研究方向为模式识别与智能系统、智能科学与技术。E-mail:
作者简介:

栗润腾,硕士研究生,研究方向为大语言模型与多智能体协作。E-mail:

Key Factor Extraction Method of Agricultural User Demand Based on Large Language Models
Runteng LI1, 3, 4 , Yiqun WANG1, Hongda LI2, 3, 4, Jingchen LI3, 4, Wenbai CHEN1
Affiliations
  • 1.School of Automation, Beijing Information Science & Technology University, Beijing 100192, China
  • 2.School of Electronic Information and Electrical Engineering, China Agricultural University, Beijing 100083, China
  • 3.National Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China
  • 4.Information Technology Research Center, Beijing Academy of Agriculture and Forestry Sciences, Beijing 100079, China
  • LI Runteng, E-mail:

出版时间: 2026-03-30 doi: 10.12133/j.smartag.SA202509011
文章导航
收藏切换

【目的/意义】 在农业领域,用户需求文本是农技精准推广、政策靶向服务的核心依据,但其包含的农业专业术语、多场景差异、动态更新特征,导致传统方法存在“专业术语识别误差大、多场景需求分类模糊、动态需求响应滞后”的问题,难以支撑高效的需求分析。因此,本研究构建基于大语言模型的农业需求智能分析方法,针对性解决传统方法的上述痛点,实现农业用户需求的精准识别、结构化提取与高效解析,为农技推广、政策服务提供高质量的需求数据支撑,推动农业数字化转型的精准落地。 【方法】 在农业需求分析中引入大语言模型,利用其语义理解与推理能力实现对用户需求的精准解析。提出了“3阶段训练+多智能体协同”的技术路径:首先收集8万条农业文本与2.28万条标注样本,用于领域预训练与指令微调,以增强模型的领域适应性;随后设计多智能体协同框架,由管理型智能体负责任务调度与质量控制,任务型智能体分别承担需求分类、关键因子提取与解释生成,从而实现对农业用户需求的结构化分析。 【结果和讨论】 农业需求分析智能体框架(Agri-NeedAgent)在需求类型匹配的准确率达到84.6%,关键因子提取的F1值达到85.2%,接口合规率94.2%,可解释性90.2分,均显著优于未经过领域微调的通用大语言模型与不含多智能体协同机制的对照方法。 【结论】 大语言模型经过领域适配与微调后,能够实现对农业用户需求的深度解析与关键因子精准提取,且能通过多智能体协同机制提升分析的完整性与可解释性,为智慧农业决策支持系统提供高质量的数据支撑。

农业需求分析  /  多智能体  /  大语言模型  /  关键因子提取  /  领域微调

[Objective] In the agricultural domain, user demand texts serve as essential primary sources for agricultural extension, production management, and policy services. However, these texts typically contain highly specialized terminology, exhibit non-standard, colloquial, and diverse linguistic expressions, present fragmented semantics, and rely heavily on contextual reasoning. Such characteristics make them difficult to parse accurately using traditional rule-based approaches or shallow machine learning models. Consequently, these limitations often lead to biased demand classification and incomplete extraction of key factors, thereby constraining the quality of data available for intelligent agricultural decision-making. To address these challenges, the aim of this research is to develop a robust, domain-adapted, and highly interpretable structured analysis method for agricultural user demands. [Methods] Agri-NeedAgent, an agricultural user demand analysis framework, was proposed based on a "three-stage training + multi-agent collaboration" paradigm. First, during the domain knowledge pretraining stage, 80 000 agriculture-related texts, including crop cultivation manuals, pest and disease control guides, agricultural policy documents, and farmer consultation records, were used to construct domain-specific semantic understanding, thereby enhancing the model's capability to interpret agricultural terminology, dialectal expressions, contextual logic, and implicit semantics. Second, in the instruction fine-tuning stage, 6 320 annotated samples in an "instruction-input-output" format were employed to establish an explicit mapping from raw demand texts to structured outputs. Third, in the agricultural knowledge low-rank adaptation stage, Low-rank Adaptation (LoRA) was applied to perform lightweight parameter tuning on task-specific agents, enabling targeted adaptation for demand classification and key-factor extraction tasks. Built upon the above training process, a multi-agent collaborative framework was constructed, in which the manager agent was responsible for task scheduling and quality control, while task agents were designed to perform demand classification, key-factor extraction, and explanation generation, respectively. Through this division of labor and collaborative mechanism, the framework achieved efficient and structured analysis of agricultural user demands. [Results and Discussions] Experimental results demonstrate that the proposed Agri-NeedAgent achieved a demand classification accuracy of 84.6%, a key-factor extraction F1-Score of 85.2%, a structured interface compliance rate of 94.2%, and an interpretability score of 90.2.These results showed clear improvements over traditional deep learning models such as Bidirectional Encoder Representations from Transformers (BERT) as well as general-purpose large language models (LLMs) without domain adaptation. The findings confirmed the critical role of domain knowledge injection, explicit task alignment, and multi-agent specialization in enhancing semantic understanding and structured analysis of agricultural texts. Ablation experiments further validated the effectiveness of each component. Removing domain pretraining or LoRA fine-tuning resulted in substantial performance degradation in classification and key-factor extraction, indicating the necessity of domain adaptation and task-specific optimization for handling non-standard agricultural expressions. Moreover, eliminating the manager agent or the Reasoning and Acting (ReAct) mechanism significantly reduced structured interface compliance and interpretability, highlighting the importance of task coordination, intermediate verification, and multi-step reasoning for ensuring logical consistency and output completeness. Additionally, removing the external knowledge base reduced the interpretability score from 90.2 to 77.6, underscoring its essential role in providing theoretical grounding, reasoning support, and professional explanations. Although the multi-agent collaboration introduced an additional inference overhead of approximately 140 ms, the overall per-sample inference time remained within 225 ms, meeting the real-time requirements of agricultural consultation scenarios. [Conclusions] Supported by a "three-stage training + multi-agent collaboration" framework, LLMs can effectively address challenges posed by non-standard expressions, semantic fragmentation, and multi-factor reasoning in agricultural user demand texts. The proposed method demonstrated significant improvements in demand classification, key-factor extraction, structured output compliance, and interpretability, providing high-quality and traceable structured data for intelligent agricultural decision-making. After domain adaptation and task-specific tuning, the model not only gains enhanced capability for deep semantic analysis of agricultural user demands but also ensures the completeness and interpretability of outputs through multi-agent coordination. Although the current workflow still requires optimization in terms of data preparation, staged training, and knowledge-base updating, future work will focus on expanding region-specific and emerging-technology-related demand data, developing a dynamically updated agricultural knowledge system, improving multi-agent coordination efficiency, and exploring cross-lingual agricultural demand analysis to further promote the application and deployment of agricultural large models across broader scenarios.

agricultural demand analysis  /  multi-agent  /  large language model  /  key factor extraction  /  domain fine-tuning
栗润腾, 王一群, 李宏达, 李静晨, 陈雯柏. 基于大语言模型的农业用户主体需求关键因子提取方法. 智慧农业(中英文), 2026 , 8 (2) : 265 -278 . DOI: 10.12133/j.smartag.SA202509011
Runteng LI, Yiqun WANG, Hongda LI, Jingchen LI, Wenbai CHEN. Key Factor Extraction Method of Agricultural User Demand Based on Large Language Models[J]. Smart Agriculture, 2026 , 8 (2) : 265 -278 . DOI: 10.12133/j.smartag.SA202509011
随着信息技术与农业领域的深度融合,农业正加速向智能化、精准化的数字化转型阶段迈进1。人工智能、大数据、物联网等前沿技术的广泛应用,正在推动农业生产方式向智能化、精准化与可持续方向迈进2。特别是在农业信息服务、生产决策支持、环境调控及市场供需匹配等方面,对智能系统的依赖程度不断加深,传统人工经验驱动的决策模式正逐步被数据驱动的智能方法所替代3
用户需求文本是农技推广与政策服务的重要原始信息,其精准识别与结构化处理对农业数字化转型具有关键作用。与通用领域文本相比,农业用户需求文本具有鲜明的领域特征:内容上高度专业化,涵盖作物种类、病虫草害、施肥技术、气象灾害等多个子领域4。表达方式则呈现出多样性与非标准化特征,常包含口语化、模糊化和区域化的表述,语义理解高度依赖上下文推理能力5, 6。同时,农业需求还具有显著的动态性和多因交互性,受到季节、气候、政策、市场等多重因素影响,其演变规律呈现非线性和非平稳性特征,从而加剧了需求解析的复杂性7, 8。在此背景下,如何从海量、非规范的农业需求文本中精准提取核心因子,已成为农业数字化转型中亟待解决的关键问题,但其相关研究却相对稀缺,现有技术难以满足实际应用需求。
国内外学者围绕农业文本处理展开多方面探索,但针对农业用户需求关键因子提取的专项研究仍存在明显短板。在传统方法层面,早期研究多采用规则模板与关键词匹配等传统方法,对农业用户需求进行模式化解析,其优势在于实现简单、计算成本低,但在面对复杂语义时适应性不足9。CHATTERJEE和KAUSHIK10对农业领域术语的自动抽取进行了综述,指出现有工具在精度和召回之间存在较大波动,并强调选择合适工具的重要性。随后,机器学习方法如支持向量机(Support Vector Machine, SVM)和随机森林(Random Forest, RF)等,这类方法依赖人工设计特征,通过统计学习实现文本分类与因子抽取。在农业数据少样本场景中,SVM表现较稳健,但对标注语料质量敏感,且难以处理复杂语义和上下文依赖。已有研究指出,SVM在精准农业领域具有优势,但亦需借助空间、遥感等特征提升效果11。另一方面,将RF应用于农业数据的多分类任务,也取得了显著改善12。然而,这类方法在应对语义歧义与上下文依赖时依旧存在局限,且对高质量标注语料依赖较大,泛化能力有限。
随着深度学习的发展,基于Transformer的双向编码器表征模型(Bidirectional Encoder Representations from Transformers, BERT)13、双向长短期记忆网络(Bidirectional Long Short-Term Memory, BiLSTM)等模型逐渐应用于农业需求分析任务。例如,LIU等14用BERT-BiLSTM-CRF实现柑橘病虫害实体识别,虽提升了实体识别精度,但仍局限于表层语义建模,无法深度解析多维度因子。这类模型缺乏可解释性,结果难以追溯,限制了其在高风险农业决策中的应用。
近年来,大语言模型(Large Language Models, LLMs)凭借强大的语义理解与跨句推理能力,为农业需求关键因子提取提供了新的技术路径,但现有研究仍未聚焦该核心任务。例如,王婷等15设计的果蔬农技知识智能问答系统,通过LLMs实现用户提问-知识回答的交互,但未将“需求分类-因子提取-解释生成”形成闭环,可能导致回答缺乏针对性。郭旺等16指出,LLMs在农业领域的应用多集中于问答、生产预测等场景,针对需求关键因子提取的专项适配不足,且存在领域术语识别精度低、多因子关联建模缺失等瓶颈。这些研究清晰展示了LLMs在农业领域的应用前景与实践价值,但同时也暴露出LLMs在农业需求分析任务中仍存在显著缺陷:一是通用LLMs对方言、农业专业术语的识别精度不足,易导致关键因子遗漏;二是现有研究多聚焦单一任务,未涉及需求分类-因子提取-解释生成的全流程结构化分析;三是难以处理多步骤、多维度的复杂需求解析,且可解释性支撑不足,无法满足农业决策对结果透明性的要求。
针对上述研究缺口与实际需求,本研究提出基于LLMs的“3阶段训练+多智能体协同”方法,专项解决农业用户需求关键因子提取的核心痛点。首先,在领域知识预训练阶段,通过引入大规模农业专业语料,增强模型对方言、口语化表述及专业术语的识别能力,从而提升对非规范化表达的理解水平。其次,在需求因子提取导向优化阶段,结合标注样本与思维链推理机制,使模型能够建模跨句、跨维度的复杂上下文关系,精准捕捉分散在不同语境中的关键因子。最后,在农业知识低秩适配阶段,通过低秩适应LoRA轻量化微调方式注入结构化知识与专家规则,不仅提高了模型对环境条件与多因子交互的敏感性,还增强了输出结果的可解释性,实现了分析过程的透明与可追溯。通过这一系列设计,本研究方法实现了从通用语言理解到农业需求解析的有效迁移,突破了现有方法在非规范化表达处理、上下文依赖建模及可解释性方面的瓶颈,为农业用户需求关键因子提取提供了一条可行且可扩展的技术路径。
LLMs是近年来自然语言处理(Natural Language Processing, NLP)领域取得突破性进展的核心技术之一。在农业领域,LLMs的研究和应用正伴随智慧农业的深化而持续深入,凭借其对复杂语言模式的深度捕捉与长上下文建模能力,为破解农业用户需求文本的解析难题提供了核心技术支撑17。以生成式预训练Transformer(Generative Pre-trained Transformer, GPT)18系列模型、BERT13等为代表的经典模型,通过预训练阶段的因果语言建模或掩码语言建模目标,构建了强大的通用语言理解与生成基础,能够在一定程度上应对农业文本中的口语化表达、专业术语交织等问题。
此外,LLMs强大的文本理解与生成能力,使其能够突破自身参数化知识存储的局限,与知识图谱等外部专业知识系统深度融合,通过外源知识的精准注入弥补通用模型在垂直领域的知识短板。例如,姜京池等19在构建知识图谱引导的农业大语言模型(Knowledge Graph-Guided Agricultural LLM, KGLLM)时进一步验证并拓展LLMs的农业领域适配潜力。该研究在百川大模型(Baichuan)、通用语言模型对话系统(Chat General Language Model, ChatGLM)、通义千问大模型(Qwen)等主流通用LLMs上的实验表明,优化后的模型在农业问答任务的内容流畅性、准确性、真实性及领域忠诚度上均有突出表现。
从应用前景来看,LLMs在农业领域的价值潜力显著,其不仅能为农业用户需求解析提供智能化工具,更可支撑农业生产全链条的精准化服务。通过理解农户的模糊需求表述,快速定位核心诉求并关联对应农技资源,或为基层农技推广人员提供高效的信息筛选与匹配能力,减少农业信息传递中的不对称问题。这种技术赋能不仅能提升农业信息服务的效率,更能为农业数字化转型提供底层语言理解支撑,推动传统农业服务模式向数据驱动、智能响应的方向升级。
基于LLMs的智能体技术,使人工智能从被动响应工具向主动任务执行者演进,为处理农业需求分析中的多环节复杂任务提供了新范式。这类智能体以LLMs为认知中枢,通过自然语言接口解析需求,自主调用外部工具与知识资源完成推理决策,相较于传统规则驱动的智能体,展现出更强的环境适应性与任务泛化能力20。但单一智能体在应对农业需求分析中“需求分类-关键因子提取-解释生成”的多维度任务时,易陷入认知负荷过载,导致结果完整性与可解释性下降,难以满足农业决策对可靠性的高要求21
近年来,多智能体协作的元编程框架(Meta Programming Framework for Multi-Agent Collaboration, MetaGPT)22、多智能体大语言模型应用对话框架(Auto-Generation: Enabling Next-Gen LLM Applications via Multi-Agent Conversation, AutoGen)23等多智能体协作框架的发展为解决这一问题提供了关键思路。MetaGPT借鉴软件工程团队的分工模式,通过项目经理、工程师等专业化角色划分与标准化操作程序的编码化设计,实现复杂任务的系统化拆解与协同执行,其多角色复核机制有效降低了单一智能体的决策偏差。AutoGen则以高度模块化的设计支持多智能体通过自然语言协议进行灵活对话与任务编排,通过管理型智能体(Manager Agent)与任务型智能体(Assistant Agent)的协同,可动态适配不同场景的任务需求。本研究充分借鉴上述框架的核心优势,构建适配农业需求分析的多智能体协同体系:设置管理型智能体负责任务调度与质量控制,参照MetaGPT的标准操作流程(Standard Operating Procedure, SOP)机制建立标准化流程,确保需求解析全链路的一致性;划分需求分类、关键因子提取、解释生成3类任务型智能体,依托AutoGen的专业化分工理念,使各智能体聚焦单一子任务以提升处理精度。同时,引入思维链(Chain-of-Thought, CoT)24技术引导智能体显式呈现推理过程,显著降低跨智能体信息传递中的语义歧义,最终形成“分工明确-协同高效-可追溯”的处理机制,为后续农业领域专用框架的设计奠定基础。
但从农业需求分析的实际场景来看,当前仍面临文本表达碎片化、方言与专业术语交织、多源信息耦合等独特挑战。同时,传统需求解析方法普遍存在对人工标注依赖度高、模型推理可解释性不足及知识更新滞后等问题。为此,本研究提出一种基于大语言模型的农业需求分析多智能体协同框架:农业需求分析智能体系统(Agri-NeedAgent)。该框架充分借鉴前述LLMs的预训练知识基础与指令遵循能力,结合多智能体系统的任务分解与协同机制,通过引入领域知识增强模块、动态任务分配策略与结果校验机制,最终实现对农业用户需求的精准解析与结构化输出,以期为后续智能决策提供可靠数据支撑。
本研究第一阶段农业领域知识预训练所用的农业领域文本数据集约8万条,涵盖2022—2025年发布的农业政策文件、农技问答等最新技术与政策动态,确保模型掌握前沿农业实践与趋势。数据集涵盖水稻、小麦、玉米、苹果、柑橘等作物,也涵盖猪、鸡、牛等畜禽品类及鱼、虾、蟹等水产养殖对象。全面覆盖种植管理、养殖技术、病虫害诊断、施肥灌溉等核心场景,形成兼具专业性与时效性的领域语料库。
第二阶段农业需求因子提取导向优化及第三阶段农业知识低秩适配所用实验数据集为经标注的农业领域多源文本,覆盖用户多样且复杂的需求。数据来源包括农业论坛用户发帖、农业服务热线咨询记录及农业技术专家访谈,涉及多种作物及养殖品类,充分体现实际应用场景,见表1
为强化模型推理的语义支撑与结果溯源能力,研究同步构建了结构化的多源异构知识库,形成了1个“知识注入-推理验证-结果溯源”的闭环机制。该知识库整合300本农业专著、5 000条农技问答对、150篇领域研究论文及220份农业政策文件,经标准化处理与关联整合后,构建起覆盖“理论-实践-科研-政策”的农业知识网络。在需求分析场景中,知识库可提供专业术语阐释、技术流程规范、政策边界界定等语义锚点,为多智能体协作中的知识推理与需求匹配提供可解释性依据,从理论溯源、实践验证、科研创新、政策合规4个维度保障需求关键因子提取的准确性与决策参考价值。知识库数据的具体设置如表2所示。
将数据划分为作物品种与选育、种植与栽培管理、病虫草害防治、畜牧养殖与水产、农资与农业设施、农业技术应用与咨询、农业政策与经营管理七大类需求标签体系,并细化至30个子标签。例如,在种植与栽培管理类别下,进一步区分水肥管理、田间管理、设施农业技术等子标签。标注标准围绕农业用户需求的核心要素展开,确保模型能够准确进行需求分类并识别关键因子。具体包括以下5类标注维度:(1)核心实体识别。作物名称、病虫害名称等。(2)重要属性/特征。抗倒伏、有机种植等描述性特征。(3)关键措施/环节。药物防治、配方施肥、夏季修剪等农事操作。(4)环境/条件/范围。南方多雨、冬季低温等其他相关要素。(5)量化指标。每公顷产1 800 kg、用药浓度等具体数值参数。图1更直观呈现了上述七大类30个子标签的需求分类体系,以及核心实体识别、重要属性/特征等5类标注维度的具体对应关系。
标注流程共分为3个环节。首先,由3名农业领域的研究生独立开展“双盲”标注,以尽可能降低主观偏差。标注结果的一致性通过Kappa系数进行评估,设定阈值为0.85,实际平均Kappa值达到0.88,表明标注具有较高的一致性。其次,在此基础上对最初的22 800条标注样本进行一致性检验与人工复核,最终筛选出15 800条高质量样本用于模型训练。其中,6 320条(40%)用于面向需求因子提取的优化,4 740条(30%)用于农业知识的低秩适配,1 580条(10%)作为验证集,3 160条(20%)作为测试集。最后,根据标注过程中收集的反馈信息,不断迭代完善标签体系,并新增“新型农业技术应用”子标签,以更好地覆盖用户对物联网、人工智能等新兴技术应用方面的需求表达。为直观呈现数据集的类型分布特征,图2展示了七大类需求标签的占比情况。数据显示,种植与栽培管理类(27.0%)、病虫草害防治类(20.0%)样本占比相对较高,与农户生产中田间管理、病虫害应对的高频需求场景一致;农业技术应用与咨询、农业政策与经营管理类样本占比均为8.0%,符合当前农业数字化转型中新兴需求的规模特征,整体数据分布虽非完全均衡,但无极端类别失衡,具备良好的代表性与实用性。
本研究提出的基于大语言模型的农业用户关键因子提取方法,以“领域知识注入-多智能体专业化协同-结构化解析输出”为核心设计思路,构建了系统性的技术框架。图3展示了该方法的整体架构,主要包括以下3个部分。
1)通过农业领域知识预训练、农业需求因子提取导向优化和农业知识低秩适配“3阶段训练”策略实现模型语义增强。
2)设定管理型智能体和任务型智能体,通过严格的提示词设计赋予管理型智能体推理与行动(Reasoning and Acting, ReAct)的思维模式,通过多个智能体之间的动态调度,实现需求分类、关键因子提取与结构化可解释性输出的专业化分工。
3)构建多源异构农业知识库,将其作为模型推理的语义锚点与结果溯源的证据库,为推理过程提供语义支撑与证据溯源。
本研究基于LLMs构建农业用户需求关键因子提取系统,并采用三阶段训练策略以实现模型的领域适应与任务专业化。具体流程包括:农业领域知识预训练、农业需求因子提取导向优化,以及农业知识低秩适配。通过逐层递进的训练设计,模型逐步从通用语言能力过渡到农业领域理解,最终完成对需求分类、关键因子提取子任务的精准建模。整体的训练流程如图4所示。
为增强LLMs对农业领域的语言理解能力,本研究首先对基础模型Qwen3-1.7B进行领域知识语义构建。训练数据来源于农业领域的非结构化文本,包括农技问答、政策文件、作物栽培手册、病虫害防治指南等,共计约8万条文本样本。通过该阶段训练,模型可系统掌握抗倒伏、节水灌溉等农业专业术语、行业表达习惯及上下文逻辑关系,从而更精准地识别农业场景中的隐含语义,为后续任务微调奠定基础。
领域预训练完成后,模型进入指令微调阶段,旨在建立对农业用户需求分析任务的映射能力。本阶段基于标注数据集总样本的40%,采用“指令+输入+输出”的三元组格式进行训练。
训练过程中,模型需从自然语言中识别出“设施农业技术”这类农业用户的需求类型、提取温室大棚番茄、熊蜂授粉优势等关键因子。为提升模型的泛化能力,本研究引入少样本提示与CoT策略。通过构造“推理步骤+输出”的样本,引导模型逐步分析问题结构,增强其逻辑推理能力。
为实现各智能体对专业任务的适配性,本研究采用LoRA技术分别对模型开展子任务微调。LoRA的核心思想是在预训练模型的权重更新过程中引入低秩分解,仅训练少量新增参数,即可在显著降低计算成本的同时,有效保留模型原有的通用语言能力,从而实现对预训练模型的高效参数微调。
具体而言,对于模型中某一层的原始权重矩阵W0d×k,传统的全参数微调会直接更新W0,涉及d×k量级的可训练参数。而LoRA假设权重更新W具有低内在秩,因此将其分解为两个低秩矩阵的乘积:W=BA。此时,模型前向传播的计算如公式(1)所示。
h=W0x+Wx=W0x+BAx
式中: h 为输出向量;x为输入向量;W0为原始权重矩;W为权重更新矩阵Ar×dBk×r为低秩矩阵;r为低秩矩阵的秩,且rd,k
训练过程中,原始权重矩阵W0被冻结,不参与梯度更新,仅优化低秩矩阵AB。因此,可训练参数量由d×k大幅减少至r×d+k。因rdk,所以该方法以极小的参数开销实现高效微调。LoRA的权重矩阵AB通过高斯分布进行初始化,缩放系数α用于调控LoRA模块对最终输出的影响强度,更新后的权重表示为W=αrBA。模型反向传播时,仅计算损失函数对AB的梯度,并通过优化器进行更新,而W0保持冻结状态。该过程如公式(2)公式(3)所示。
Anew=Aold-ηA
Bnew=Bold-ηB
式中:η为学习率,为0到1的实数;为损失函数;AB分别为损失函数对矩阵AB的梯度。通过该策略,本研究分别对需求分类智能体(Demand Classifier Agent, DCA)和关键因子提取智能体(Key Factor Extractor Agent, KFEA)进行子任务微调,使其在保持通用语言能力的基础上,深度适配农业领域特定任务,提升关键因子提取的精度与鲁棒性。LoRA模型训练示意图如图5所示。
为提升农业用户需求分析的系统化与智能化水平,本研究提出基于LLMs构建的多智能体协同框架Agri-NeedAgent,用于完成农业用户需求类型分类、关键因子提取,以及可解释性输出3项核心任务。该框架由4个功能明确的智能体组成:DCA、KFEA、综合解释生成智能体(Structured Explanation Generation Agent, SEGA),以及1个负责整体任务调度与结果整合的管理者智能体(Manager Agent, MA)。通过引入管理者智能体MA,赋予其ReAct的思维模式实现了农业用户主体需求关键因子任务间的有效协作与质量控制。MA具备任务规划、过程监控、反馈审查与结果汇总的能力。具体而言,MA首先解析用户的原始查询,初始化任务流程并协调DCA、KFEA和SEGA的执行顺序。随后,在每一步任务完成后,MA对各智能体的中间输出进行合理性判断与一致性校验,必要时触发修正机制或引导重试。最终,MA将3部分结果进行整合,生成统一格式的输出报告,包括需求类别标签、关键因子及需求分析解释,确保最终结果的完整性与准确性。基于ReAct框架的多智能体任务协调机制如图6所示。
其中,DCA负责对输入的农业用户需求文本进行语义理解与意图识别,依据预定义的七大类30种子标签体系,将其归类至最匹配的需求类型。KFEA基于标注标准中:核心实体识别、重要属性/特征、关键措施/环节、环境/条件/范围、量化指标5类关键因子维度,从文本中抽取出与用户需求密切相关的因素,并进行结构化表示。SEGA则整合原始文本内容、DCA和KFEA的分析提取结果,在此基础上结合知识库的内容,进行可解释的结构化输出。上述3个任务智能体分别专注于单一子任务,以提升各模块的专业性与可解释性。
本研究设计了一系列对比实验,旨在验证所提出方法的有效性,并揭示LLMs参数规模、领域微调策略对模型性能的影响。在数据集划分上,为避免数据分布偏差,采用分层抽样方式将标注数据集划分为:第二阶段指令微调训练集40%、第三阶段LoRA微调数据集30%、验证集10%和测试集20%,确保模型评估的独立性和泛化性。实验过程中将Agri-NeedAgent与BERT、Qwen3-1.7B、DeepSeek-R1:1.5B、Qwen3-4B、DeepSeek-R1:7B、ChatGLM3-6B、Baichuan2-7B-Base、InternLM2.5-7B基线模型进行比较。所有实验均独立重复10次,每次实验采用不同的随机种子,以确保模型初始化、数据划分及训练过程相互独立,最终结果取10次实验的平均值。
为深入分析模型在各训练阶段的收敛行为与学习动态,本研究记录并可视化了第二阶段农业需求因子提取导向优化和第三阶段农业知识低秩适配的训练过程。由图7a可知,第二阶段训练过程中,训练损失从初始1.10稳步下降至0.37,验证损失从初始1.15同步下降至0.42,每轮两者差距均≤ 0.05,且全程保持“训练损失下降时验证损失同步下降”的趋势,证明模型在“需求分类-因子提取”任务映射建立过程中泛化稳定。由图7b可知,第三阶段训练中,训练损失从初始0.75逐步下降至0.28,验证损失从初始0.78同步下降至0.31,每轮差距≤ 0.03,并全程保持“训练损失下降时验证损失同步递减”的稳定趋势,证明模型训练过程稳定且泛化能力良好。
本研究构建了一套多维度的评估体系,涵盖关键因子提取精度、需求类型匹配率等核心指标,以全面衡量模型在农业用户需求分析任务中的性能。
采用F1值作为主要评估指标。每个样本的真实因子集合为TrueFactors=f1,f2,...,fn,,模型预测因子集合为PredFactors。用精确率(Precision)计算模型提取出的关键因子中与人工标注结果匹配的精度,如公式(4)所示。
Precision=TrueFactorsPredFactorsPredFactors
式中:TrueFactors表示人工标注的真实关键因子集合;PredFactors表示模型预测得到的关键因子集合;其中TrueFactorsPredFactors表示模型成功识别并与人工标注一致的关键因子数量。
召回率(Recall)是衡量模型对人工标注因子的覆盖能力,如公式(5)所示。
Recall=TrueFactorsPredFactorsTrueFactors
F1值作为精确率与召回率的调和平均数,用于综合评价模型在处理农业用户需求分析问题上的表现,如公式(6)所示。
F1=2×Precision×RecallPrecision+Recall
需求类型匹配准确率用于衡量模型对用户需求类别的识别能力,如公式(7)所示。判定标准同样基于Jaccard相似度,若模型预测的需求类型文本与人工标注的需求类型文本的相似度≥0.6,则判定匹配成功。
准确=正确预测的需求标签数总样本数×100%
式中:正确预测的需求标签数表示满足相似度判定标准的样本数量,即需求类型识别正确的样本数;总样本数表示参与评估的测试集样本总数,涵盖各类农业用户需求场景。
用于衡量系统生成的农业用户需求分析报告的规范化程度与信息完备性。该指标要求模型不仅需准确识别需求类型与关键因子,还需将其组织为标准化结构JSON格式,完整覆盖预设输出字段,为下游决策支持系统提供即用型数据接口。
这部分指标通过自动化校验脚本直接评估系统输出结果的结构化合规性,核心度量标准为“代码验证通过率”。该指标反映系统生成的农业需求分析报告能否无需人工干预,直接通过预定义的机器校验规则,满足下游决策系统的数据接入要求,如公式(8)所示。
合规=通过检验的样本量总样本量×100%
式中:通过检验的样本量表示满足预设结构化标准并通过自动化校验脚本验证的样本数量。
可解释性用于衡量系统对农业用户需求分析结果的逻辑可追溯性与决策支撑价值,其核心要求是模型输出不仅需明确包含关键因子与需求类型,还需提供清晰的推理链条、农业专业知识依据及语义关联说明,以满足农业决策对过程透明、结果可验证的高阶需求,避免传统方法因解释模糊导致的决策风险。
为实现可解释性的量化评估,同时保障评分标准与农业场景的适配性,本研究采用基于GPT-4的自动化评分机制,并针对“抽象标准难把控、主观偏差易干扰”的问题,构建语义一致性、推理逻辑性、领域知识关联性、表达清晰度四大核心维度的5档评分锚定体系。该体系通过“维度拆解-档位量化-场景锚定”的递进式设计,将抽象的可解释性评估转化为具象、可参照的农业场景行为。以四大维度中的“语义一致性”为例,其5档评分标准与对应农业场景示例如表3所示。该维度的评估核心是:需求核心要素的覆盖完整性。例如,在“黄土高原种植节水苹果”这一典型需求中,“优秀”档位需完全覆盖:黄土高原(区域)、节水(诉求)、苹果(作物)所有关键信息,而“不足”档位则表现为核心要素偏差,各档位均以具体农业场景行为为锚点,确保生成式预训练变换模型4(Generative Pre-trained Transformer 4, GPT-4)评分时有明确参照,减少主观臆断空间。同时,为降低单次评分的随机性,对每条测试样本,独立调用GPT-4完成3次评分并计算均值,最终以四维得分的算术平均值作为该样本的最终结果。
本实验采用的操作系统为Linux,环境配置为:CPU为14核,Xeon(R) Platinum 8362的内存45 G;GPU为NVIDIA GeForce RTX3090显卡;显存为24 GB;Python版本为3.10.11;Cuda版本为12.4。模型训练过程中的超参数配置如下:第一阶段为农业领域知识预训练阶段,采用Batch Size为32,学习率为5e-5,训练迭代轮次为10轮,正则化系数为0.01,梯度裁剪阈值为1.0,以8万条农业领域文本为训练数据,实现模型对农业术语与表达习惯的初步掌握。第二阶段为指令微调阶段,基于6 320条“指令+输入+输出”三元组格式标注样本进行训练,采用Batch Size为16,学习率调整为2e-5,迭代轮次为8轮,引入少样本提示(Few-shot Prompting)与CoT推理策略,其中少样本提示示例数量为3条,思维链推理步骤控制在3~5步,以增强模型对需求分类与关键因子提取任务的映射能力。第三阶段农业知识低秩适配阶段,针对DCA与KFEA分别进行子任务优化,LoRA策略中的秩设置为16,缩放系数为32,Batch Size为8,学习率为1e-4,迭代轮次为12轮,梯度累计步数为2,正则化系数为0.001,通过低秩矩阵适配实现对模型少量参数的精准调整,在降低计算成本的同时提升子任务处理精度。所有训练过程均采用权重衰减自适应矩估计优化器(Adaptive Moment Estimation with Weight Decay, AdamW),学习率调度策略为线性衰减,热身比例为0.1,以确保模型训练的稳定性与收敛效率。
为验证所提出的基于LLMs的农业用户需求关键因子提取方法的有效性,本研究设计并实施了多组对比实验,主要围绕4个核心任务展开:需求类型分析、关键因子提取、接口合规率,以及可解释性。通过对比传统深度学习模型BERT和Qwen3-1.7B、Qwen3-4B、DeepSeek-R1:1.5B、DeepSeek-R1:7B、ChatGLM3-6B、Baichuan2-7B-Base、InternLM2.5-7B不同参数规模的LLMs模型,评估其在农业领域任务中的表现。
实验结果展现了不同模型在农业用户需求分析任务中的性能差异,具体数据如表4所示。从整体表现来看,经过领域适配与多智能体协同优化的Agri-NeedAgent方法在各项指标中均处于领先地位,验证了本研究提出的“3阶段训练+多智能体协同”技术路径的有效性。
在需求类型分析准确率方面,Agri-NeedAgent达到84.6%,显著高于其他模型。传统深度学习模型BERT的准确率为59.8%,未经过领域适配的小参数通用大语言模型Qwen3-1.7B、DeepSeek-R1:1.5 B准确率为55%~57%,即使是参数规模提升至4 B-7 B量级的模型,准确率为72%~78%,也与Agri-NeedAgent存在一定的差距。这表明,单纯依靠模型参数规模提升或通用预训练,难以满足农业领域专业化的需求分类任务,而领域微调与多智能体协同机制能有效增强模型对农业需求语义的精准理解。
在关键因子提取F1值上,Agri-NeedAgent以85.2%的成绩表现最优。对比来看,BERT的F1值为65.2%,通用大语言模型中性能较好的DeepSeek-R1:7 B为76.4%,Agri-NeedAgent与之相比提升了8.8个百分点。这一结果印证了3阶段训练策略的有效性。通过3阶段递进式优化显著提升了模型对核心实体、环境条件等关键因子的提取精度。
接口合规率指标中,Agri-NeedAgent达到94.2%,远高于BERT的50.2%和其他通用模型(77.6%~90.8%)。这得益于多智能体框架中MA的动态校验与格式规范化能力,确保输出结果严格遵循JSON标准格式,满足下游决策系统的直接调用需求,解决了传统模型输出非结构化、难以直接应用的痛点。
可解释性评分指标中,Agri-NeedAgent获得90.2分,高于其他模型。其中,通用大语言模型的可解释性评分为77.9~83.7分,BERT仅为66.6分。这一优势源于SEGA与外部知识库的协同作用。SEGA基于需求分类和关键因子提取结果,结合农业专业知识生成逻辑清晰的推理链条,使分析结果具备语义一致性、推理逻辑性和领域知识关联性,满足农业决策对透明性和可验证性的要求。
结合图8所示的带有误差棒的柱状图,可以更直观地反映不同模型在各项任务中的性能差异及稳定性特征。从误差棒的长度来看,Agri-NeedAgent在4个指标上的波动范围均显著缩短:需求类型分析准确率与关键因子提取F1值的误差均控制在±0.2%,接口合规率为±0.1%,可解释性为±0.5。这表明该方法在多次实验中的结果表现出高度一致性,具有较强的鲁棒性。相较之下,未经过领域适配的通用模型误差范围明显更大,说明其在处理农业领域非标准化文本时稳定性较差,容易受到输入差异的影响。
从性能演进趋势来看,柱状图清晰呈现出两类发展路径:其一,通用大语言模型的性能随参数规模的增加而逐步提升。Qwen系列为1.7 B到4 B,需求类型分析准确率提高了16个百分点,关键因子提取F1值提高了10.7个百分点;DeepSeek-R1系列为1.5 B到7 B,分别提高了20.4个百分点和15.7个百分点。这验证了模型规模在提升语义理解能力方面的积极作用。其二,Agri-NeedAgent在相同参数规模下的性能显著优于通用模型。与Qwen3-1.7B相比,其需求类型分析准确率提升28.1个百分点,关键因子提取F1值提升24.1个百分点,甚至超过了参数规模更大的Qwen3-4B、DeepSeek-R1:7B、ChatGLM3-6B、Baichuan2-7B-Base、InternLM2.5-7B。表明单纯依靠模型规模扩展存在边际效益递减,而领域知识注入与多智能体协同机制能够带来显著的性能提升。
在具体指标对比中,接口合规率和可解释性尤为突出。Agri-NeedAgent的接口合规率由通用模型的最高值90.8%提升至94.2%,得益于管理型智能体在输出过程中的动态校验与格式化能力,从而确保结果能够严格遵循结构化标准并直接满足下游系统调用的要求。可解释性方面,Agri-NeedAgent的得分为90.2分,显著高于InternLM2.5-7B的83.7分。其误差范围亦最小,说明该方法结合外部知识库生成的推理链条更加清晰,有效提升了系统结果的透明性与可追溯性,从而满足农业决策对可靠性与可验证性的高阶需求。
为了深入理解Agri-NeedAgent框架各组件的贡献和重要性,本研究设计了一系列消融实验,通过逐一移除ReAct思维链、少样本提示、MA、领域预训练、指令微调、LoRA微调和外部知识库七个关键组件分别验证其在农业用户主体需求分析任务中的影响与贡献,结果如表5所示。
在保留其他模块不变的情况下,移除MA中与ReAct思维链相关的提示词。实验结果显示,需求类型分析准确率从84.6%降至82.3%,下降2.3个百分点,关键因子提取F1值从85.2%降至82.6%,下降2.6个百分点;接口合规率从94.2%降至87.1%,下降7.1个百分点;可解释性评分从90.2分降至87.2分,下降3分。这表明,ReAct引导的分步推理与动态检验机制,能有效提升多智能体在复杂农业场景中的逻辑规划能力与结果一致性,对维持任务流程的稳定性至关重要。
对于少样本提示消融实验,在只去掉少样本提示保持其他模块不变的情况下进行这部分实验。分析实验结果可知,移除后,需求类型分析准确率降至81.5%,下降3.1个百分点;关键因子提取F1值降至81.7%,下降3.5个百分点;接口合规率降至86.6%,下降7.6个百分点,可解释性评分小幅下降到87.5。这说明小样本提示通过提供任务格式范例与领域表达模式,帮助模型快速适配农业需求分析的特定逻辑,对提升任务适配性与输出规范性作用显著。
完全移除MA,仅保留DCA、KFEA、SEGA这3个任务智能体并采用简单顺序调用。实验结果显示,移除后,接口合规率急剧下降至81.4%,下降12.8个百分点;需求类型分析准确率降至82.0%,下降2.6个百分点;关键因子提取F1值降至81.9%,下降3.3个百分点;可解释性评分降至86.4分,下降3.8分。充分验证MA在任务调度、中间结果校验与跨智能体一致性协调中的核心作用,其缺失会导致各模块输出脱节,显著降低系统整体性能。
在第二阶段指令微调、第三阶段LoRA微调和其他模块全保留的前提下,移除第一阶段农业领域知识预训练环节,直接使用未经过农业领域预训练的基础模型进入第二阶段指令微调及后续流程。结果显示,需求类型分析准确率降至78.8%,关键因子提取F1值降至79.5%,接口合规率降至91.3%。该结果印证了第一阶段农业领域知识预训练的核心价值,通过8万条农业文本预训练,模型掌握专业术语及行业表达逻辑,缺失该阶段会导致模型对农业非规范化文本的理解精度显著下降,直接影响后续因子提取与需求分类。
移除第二阶段需求因子提取导向指令微调,仅通过第一阶段预训练模型直接进入第三阶段子任务适配并保持其他模块不变的情况下执行所有子任务。实验结果显示,需求类型分析准确率从84.6%降至80.4%,关键因子提取F1值降至80.6%,可解释性评分降至87.4分。说明第二阶段的农业需求因子提取导向优化使模型建立“需求文本→需求类型→关键因子”的任务映射能力,缺失该阶段会导致模型难以理解因子提取的任务目标,出现识别术语但无法关联需求类型的问题。
在第三阶段训练中,移除针对DCA的需求分类、KFEA的关键因子提取子任务的LoRA微调,使用指令微调模型执行所有子任务。结果显示,需求类型分析准确率降至77.5%,下降7.1个百分点;关键因子提取F1值降至76.2%,下降9.0个百分点;接口合规率基本稳定,可解释性评分略降为88.9分,降低1.3分。这表明LoRA通过低秩矩阵适配实现了子任务参数的高效优化,是模型在需求分类与因子提取等专业化任务中达到高精度的关键,其作用在核心分析任务中尤为突出。
对于外部知识库消融实验,在只去掉外部知识库保持其他模块不变的情况下进行实验。结果表明,移除后,模型输出的可解释性评分从90.2分大幅下降至77.6分。这直接证明了外部知识库在提供语义支撑、证据溯源及增强结果可信度方面的核心价值。值得注意的是,移除知识库对需求类型分析准确率为83.7%、关键因子提取F1值为84.4%、合规率为93.9%,结构化指标影响相对较小,说明知识库主要赋能于结果的可解释性这一高阶需求,而对基础因子识别和格式化的直接影响有限。
由此可见,Agri-NeedAgent框架中各核心组件在农业用户主体需求分析任务中呈现出清晰的功能分工与协同效应。正是这些组件的有机联动与功能互补,共同支撑了系统在需求类型识别、关键因子提取、接口合规性及可解释性等维度的优异表现,凸显了“3阶段训练+多智能体协同”这一技术路径的科学性与实践有效性,为农业需求结构化分析提供了可靠的技术范式。
为保障多智能体间中间结果的校验精度与最终输出的结构化完整性,系统在协同调度过程中存在轻微推理延迟。从实测结果来看,该协同过程较单智能体模型增加约140 ms的推理耗时,单样本推理总耗时从85 ms增至225 ms。但这一耗时仍远低于基层农技服务“秒级响应”的核心需求阈值,能够满足农户咨询、农技推广等实际场景的效率要求。
本研究提出的基于大语言模型的农业用户主体需求关键因子提取方法,通过“3阶段训练+多智能体协同”的技术路径,有效解决了传统方法在农业需求文本处理中面临的专业性、多样性和动态性挑战。实验结果显示,该方法在农业需求类型匹配准确率上达到84.6%,关键因子提取F1值达到85.2%,接口合规率达94.2%,输出的可解释性评分达到90.2分,优于BERT等传统模型及未经过领域适配的通用大语言模型,验证了领域知识注入对提升模型农业语义理解能力的关键作用。通过将农业领域知识与LLMs的高阶推理能力深度融合,本研究为农业决策支持系统提供了结构化、可溯源的数据支撑,为智慧农业的数字化转型提供了可行的技术路径。然而,现有“3阶段训练”需分阶段准备数据与调试参数,整体流程的轻量化程度有待提升,且标注样本虽覆盖主要农业场景,但在区域特色需求与新兴技术需求的覆盖深度上仍有拓展空间。此外,知识更新机制欠缺动态性,当前知识库依赖人工整理静态数据,未对接官方实时数据源,对最新农业政策、新兴农技的语义理解存在滞后性。未来的工作将进一步扩展农业知识库的覆盖范围,实现知识库的动态更新,优化多智能体的动态协作策略,并探索模型在实时需求响应与跨语言农业文本处理中的应用。
  • 科技部2030新一代人工智能重大专项(2021ZD0113603)
  • 国家自然科学基金(62276028)
  • 国家自然科学基金重大研究计划(92267110)
参考文献 引证文献
排序方式:
[1]
刘文, 王昊, 李啸林. 农业数字化转型: 研究现状、热点与趋势[J]. 技术经济与管理研究, 2025(8): 47-53.
LIU W, WANG H, LI X L. Digital transformation of agriculture: Research status, hotspots and trends[J]. Journal of Technical Economics & Management, 2025(8): 47-53.
[2]
DAYIOĞLU MALI, TURKER U. Digital transformation for sustainable future - agriculture 4.0: A review[J]. Tarım Bilimleri Dergisi, 2021, 27(4): 373-399.
[3]
SARKER I H. Data science and analytics: An overview from data-driven smart computing, decision-making and applications perspective[J]. SN Computer Science, 2021, 2(5): 377.
[4]
AYED RBEN, HANANA M. Artificial intelligence to improve the food and agriculture sector[J]. Journal of Food Quality, 2021, 2021(1): 5584754.
[5]
SHARMA S, SHARMA C, ASENSO E, et al. Research constituents and trends in smart farming: An analytical retrospection from the lens of text mining[J]. Journal of Sensors, 2023, 2023(1): 6916213.
[6]
李孝鹏, 向玉云, 张培君, . 农业领域自然语言理解技术应用综述[J]. 农业机械学报, 2025, 56(10): 200-222.
LI X P, XIANG Y Y, ZHANG P J, et al. Natural language understanding in agriculture: A comprehensive review of technologies and applications[J]. Transactions of the Chinese Society for Agricultural Machinery, 2025, 56(10): 200-222.
[7]
SHARMA P, DADHEECH P, ANEJA N, et al. Predicting agriculture yields based on machine learning using regression and deep learning[J]. IEEE Access, 2023, 11: 111255-111264.
[8]
王耀君, 徐国威, 朱建军, . 农业领域大语言模型研究进展[J]. 农业机械学报, 2025, 56(9): 240-256.
WANG Y J, XU G W, ZHU J J, et al. Survey of research on large language models in agriculture[J]. Transactions of the Chinese Society for Agricultural Machinery, 2025, 56(9): 240-256.
[9]
NISMI MOL E A, SANTOSH KUMAR M B. Review on knowledge extraction from text and scope in agriculture domain[J]. Artificial Intelligence Review, 2023, 56(5): 4403-4445.
[10]
CHATTERJEE N, KAUSHIK N. Automatic extraction of agriculture terms from domain text: A survey of tools and techniques[EB/OL]. arXiv: 2020.
[11]
KOK Z H, MOHAMED SHARIFF A R, ALFATNI M S M, et al. Support vector machine in precision agriculture: A review[J]. Computers and Electronics in Agriculture, 2021, 191: 106546.
[12]
SHI L, QIN Y Q, ZHANG J J, et al. Multi-class classification of agricultural data based on random forest and feature selection[J]. Journal of Information Technology Research, 2022, 15(1): 1-17.
[13]
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, USA: Association for Computational Linguistics, 2019: 4171-4186.
[14]
LIU Y F, WEI S Q, HUANG H J, et al. Naming entity recognition of citrus pests and diseases based on the BERT-BiLSTM-CRF model[J]. Expert Systems with Applications, 2023, 234: 121103.
[15]
王婷, 王娜, 崔运鹏, . 基于人工智能大模型技术的果蔬农技知识智能问答系统[J]. 智慧农业(中英文), 2023, 5(4): 105-116.
WANG T, WANG N, CUI Y P, et al. Agricultural technology knowledge intelligent question-answering system based on large language model[J]. Smart Agriculture, 2023, 5(4): 105-116.
[16]
郭旺, 杨雨森, 吴华瑞, . 农业大模型:关键技术、应用分析与发展方向[J]. 智慧农业(中英文), 2024, 6(2): 1-13.
GUO W, YANG Y S, WU H R, et al. Big models in agriculture: Key technologies, application and future directions[J]. Smart Agriculture, 2024, 6(2): 1-13.
[17]
李小玲. 智慧农业领域人工智能大模型的应用研究[J]. 中国农机装备, 2025(6): 81-83.
LI X L. Research on application of large language models in smart agriculture[J]. China Agricultural Machinery Equipment, 2025(6): 81-83.
[18]
RADFORD A, NARASIMHAN K, SALIMANS T, et al. Improving language understanding by generative pre-training[J/OL]. 2018.[2025-08-26]. https://gwern.net/doc/www/s3-us-west-2.amazonaws.com/d73fdc5ffa8627bce44dcda2fc012da638ffb158.pdf
[19]
姜京池, 闫莲, 刘劼. 基于精准知识筛选及知识协同生成的农业大语言模型[J]. 智慧农业(中英文), 2025,7(1): 20-32.
JIANG J C, YAN L, LIU J. Agricultural large language model based on precise knowledge retrieval and knowledge collaborative generation[J]. Smart Agriculture, 2025,7(1): 20-32.
[20]
ACHARYA D B, KUPPAN K, DIVYA B. Agentic AI: Autonomous intelligence for complex goals: A comprehensive survey[J]. IEEE Access, 2025, 13: 18912-18936.
[21]
任荣荣, 胡崇宇, 吴国龙, . 农业种植智能体(Agri-agent)的构建与应用展望[J]. 农业展望, 2024, 20(6): 92-106.
REN R R, HU C Y, WU G L, et al. Construction and application prospect of agricultural planting agent (Agri-agent)[J]. Agricultural Outlook, 2024, 20(6): 92-106.
[22]
HONG S R, ZHUGE M C, CHEN J Q, et al. MetaGPT: Meta programming for a multi-agent collaborative framework[EB/OL]. arXiv: 2023.
[23]
WU Q Y, BANSAL G, ZHANG J Y, et al. AutoGen: Enabling next-gen LLM applications via multi-agent conversation[EB/OL]. arXiv: 2023.
[24]
WEI J, WANG X Z, SCHUURMANS D, et al. Chain-of-thought prompting elicits reasoning in large language models[C]// Proceedings of the 36th International Conference on Neural Information Processing Systems. New York, USA: ACM, 2022: 24824-24837.
2026年第8卷第2期
PDF下载
95
39
引用本文
BibTeX
文章信息
doi: 10.12133/j.smartag.SA202509011
  • 接收时间:2025-09-02
  • 首发时间:2026-07-08
  • 出版时间:2026-03-30
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-02
基金
Major Project of Scientific and Technological Innovation 2030(2021ZD0113603)
科技部2030新一代人工智能重大专项(2021ZD0113603)
National Natural Science Foundation of China(62276028)
国家自然科学基金(62276028)
Major Research Plan of National Natural Science Foundation of China(92267110)
国家自然科学基金重大研究计划(92267110)
作者信息
    1.北京信息科技大学自动化学院,北京 100192,中国
    2.中国农业大学信息与电气工程学院,北京 100083,中国
    3.国家农业信息化工程技术研究中心,北京 100097,中国
    4.北京市农林科学院信息技术研究中心,北京 100079,中国

通讯作者:

陈雯柏,博士,教授,研究方向为模式识别与智能系统、智能科学与技术。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zhny/CN/10.12133/j.smartag.SA202509011
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏