[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3088":3,"related-3088":50},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":18,"tags":20,"search_phrases":25,"slug":28,"view_count":15,"doi":29,"paper":30,"created_at":49},3088,"A Hybrid COBIT 2019–RAG Framework to Enhance Consumer Protection Compliance in Digital Finance","https:\u002F\u002Fdoi.org\u002F10.26877\u002Fasset.v8i4.2088","The rapid growth of digital financial services has increased the need for effective regulatory compliance mechanisms, particularly for consumer protection. This study develops and evaluates a hybrid COBIT 2019–Retrieval-Augmented Generation (RAG) framework for compliance with Indonesia’s OJK Regulation No. 22\u002F2023 in a digital multifinance context. Using a design science research methodology, COBIT 2019 design factors were applied to tailor governance objectives, while a RAG pipeline was evaluated on 100 synthetic compliance queries using RAGAS metrics and expert-validated reference answers. The governance assessment prioritized EDM03, APO12, and MEA03, each assessed at capability Level 2 against a target of Level 4, indicating gaps in risk optimization, risk management, and external compliance. The optimal RAG configuration achieved an overall score of 0.8257, with context recall of 0.9217, faithfulness of 0.8502, and semantic similarity of 0.8629. It also outperformed baseline keyword search in retrieving semantically relevant regulatory passages, although broader benchmarking remains limited. The findings show that integrating structured IT governance with AI-assisted regulatory retrieval can strengthen accountability, regulatory interpretation, and compliance decision support. The framework offers a scalable approach for regulated industries, subject to further validation using real operational data and broader institutional settings.","数字金融服务的快速增长，提升了对有效监管合规机制的需求，尤其是在消费者保护方面。本研究开发并评估了一个融合COBIT 2019与检索增强生成（Retrieval-Augmented Generation, RAG）的混合框架，用于在数字多金融（multifinance）场景下遵守印度尼西亚OJK第22\u002F2023号条例。采用设计科学研究方法，应用COBIT 2019设计因子来定制治理目标，同时使用RAGAS指标和专家验证的参考答案，在100条合成合规查询上评估了一条RAG流水线。治理评估优先考虑EDM03、APO12和MEA03，三者均被评定为能力等级2，而目标为等级4，表明在风险优化、风险管理和外部合规方面存在差距。最优RAG配置的总体得分为0.8257，其中上下文召回率为0.9217，忠实度为0.8502，语义相似度为0.8629。在检索语义相关的监管段落方面，它也优于基线关键词搜索，尽管更广泛的基准测试仍然有限。研究结果表明，将结构化IT治理与AI辅助的监管检索相结合，可以加强问责制、监管解释和合规决策支持。该框架为受监管行业提供了一种可扩展的方法，但仍有待使用真实运营数据和更广泛的制度环境进行进一步验证。",null,"Advance Sustainable Science Engineering and Technology","2026-09-18T00:00:00Z","论文",10,false,0,{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":17},"研究数字金融消费者保护合规的IT治理与RAG框架，与三农、农业信息化无关，不予入选。",[19],{"name":10,"url":6},[21,22,23,24],"RAG","数字金融","合规科技","IT治理",[26,27],"COBIT 2019 RAG 合规","OJK 22 2023 消费者保护","COBIT2019RAG合规-3088","10.26877\u002Fasset.v8i4.2088",{"doi":29,"openalex_id":31,"authors":32,"venue":10,"cited_by_count":15,"oa_url":41,"card":42,"direction":46,"ingested_from":48},"W7213555882",[33,36,38],{"name":34,"orcid":35},"Edy Salim","https:\u002F\u002Forcid.org\u002F0009-0007-6264-0117",{"name":37,"orcid":9},"Mohammad Achmad Amin Soetomo",{"name":39,"orcid":40},"Eka Budiarto","https:\u002F\u002Forcid.org\u002F0009-0000-6642-260X","https:\u002F\u002Fjournal2.upgris.ac.id\u002Findex.php\u002Fasset\u002Farticle\u002Fdownload\u002F2088\u002F1736",{"tldr":43,"method":44,"finding":45,"direction":46,"opportunity":47},"构建COBIT 2019与RAG混合框架，提升数字金融消费者保护合规能力。","设计科学研究法，COBIT 2019设计因子与RAG管道，100条合成合规查询及","治理目标存在能力差距，最优RAG配置得分0.8257，优于关键词检索。","农业人工智能与决策模型","可将该混合治理与RAG框架迁移至农业数字金融或农业数据合规场景，探索领域适配与真实数据验证。","openalex","2026-09-21T23:30:45.181565Z",{"total":51,"page":52,"page_size":51,"items":53},5,1,[54,82,106,131,156],{"id":55,"title":56,"url":57,"summary":58,"summary_zh":9,"content":59,"source_name":60,"source_url":9,"published_at":61,"category":62,"cover_url":9,"hotness":13,"is_selected":14,"score":63,"score_detail":64,"sources":70,"tags":72,"search_phrases":77,"slug":80,"view_count":15,"doi":9,"paper":9,"created_at":81},2552,"数智助农成服贸会金融专题展亮点 农业银行展示AI数字人农小耘","https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7685207598819279360\u002F","9月14日服贸会报道。中国农业银行以数智助农乡村振兴为主题展示智能化金融服务体系，AI能力嵌入信贷评估、风险核查等环节；农户在村里线上提交申请后系统可辅助完成信息核验与初步授信。北京农商银行展示非接触AI健康检测机器人；北京银行展示的科创企业技术已在智慧农业、农产品溯源等场景落地。","## 数智助农成服贸会金融专题展亮点\n\n2026-09-14 10:12·[中国网三农](https:\u002F\u002Fwww.toutiao.com\u002Fc\u002Fuser\u002Ftoken\u002FMS4wLjABAAAAbAvyMQxxaA9sEwA2aUWsAOJjrH-pK95hBvQUB6GaSEg\u002F?source=tuwen_detail)\n\n近日，北京首钢园迎来2026年中国国际服务贸易交易会（以下简称“服贸会”），金融服务专题展区以“智焕新生 金融聚力”为主题。记者在展区现场看到，各家展台前人头攒动，数智金融如何下沉乡村，成为不少观众关注的焦点。\n\n中国农业银行展台以绿色为主调，“数智助农”“乡村振兴”字样醒目。一位观众朝屏幕招手，AI数字人“农小耘”随即回应。工作人员介绍，农行此次展示了智能化金融服务体系，AI能力可嵌入信贷评估、风险核查等环节，农户在村里线上提交申请后，系统可辅助完成信息核验与初步授信，不必再往返县城网点。服务半径由此延伸到村口，数字技术让金融支农更直达。\n\n![Image 1](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-axegupay5k\u002Fc75fa582de054ded9aaf0da332891c36~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1790121915&x-signature=%2FPNFaboCnXA%2F%2F0wpGhoOm4cWxLI%3D)\n\n中国农业银行展区，观众正在智能点餐机前扫码点单\n\n另一侧，北京农商银行展区排起小队。观众刘女士坐到一台智能检测机前，无需接触硬件，数秒后健康筛查报告呈现在显示屏上。“这是我们研发的非接触AI健康检测机器人，数秒完成多维度生理指标筛查。”耀眼科技副经理王俊才介绍，该设备已在北京亦庄荣华街道智慧康养机器人养老驿站试点，千余名体验者中农民占三分之一。由于部分农村地区居民健康筛查不便，银行网点正好可以成为服务下沉的入口。北京农商银行与耀眼科技合作，计划将AI健康检测机器人嵌入网点，依托养老助残卡、第三代社保卡，构建“金融网点+健康服务”便民模式。\n\n北京银行展台同样热闹。环抱式“金融会客厅”里，中轴线手绘长卷与科技产品相映成趣。环廊陈列的25件前沿科技产品，多来自该行长期支持的科创企业。记者了解到，这些企业的技术已在智慧农业、农产品溯源等场景落地，部分科创企业还获得该行专项信贷支持，将AI、物联网技术带入设施农业园区。从支持科创企业到技术反哺农业，金融正成为连接科技与乡村的桥梁。\n\n![Image 2](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-6w9my0ksvp\u002F21c30bfe934940439449f459252e41d8~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1790121915&x-signature=9CqAJQPWYzu3ZBgbNjWmOZ%2FwyxA%3D)\n\n北京银行展区，观众驻足观看机器人演奏\n\n穿行展区，记者感到，本次服贸会金融服务专题展示的不只是技术，更是服务民生、助力“三农”的路径探索。资金、科技与乡村加速融合，数智技术正让服务更贴近乡土，为乡村全面振兴提供持久支撑。\n\n来源：农民日报","中国网三农 2026-09-14","2026-09-14T02:00:00Z","报道",55,{"impact":65,"substance":66,"depth":65,"authority":67,"freshness":68,"relevant":52,"comment":69},12,13,11,7,"服贸会金融展上的数智助农展示，属企业宣传性综合报道，有AI数字人、非接触健康检测等具体案例，但无政策条款或新数据，适合作为主题页聚合素材而非每日精选头条。",[71],{"name":60,"url":57},[73,74,75,76,22],"数字乡村","智慧农业","农业人工智能","乡村振兴",[78,79],"农业人工智能 乡村振兴 数字乡村 数字金融","农业人工智能 乡村振兴","农业人工智能乡村振兴数字乡村数字金融-2552","2026-09-16T00:03:45.025672Z",{"id":83,"title":84,"url":85,"summary":86,"summary_zh":9,"content":9,"source_name":87,"source_url":9,"published_at":88,"category":62,"cover_url":9,"hotness":13,"is_selected":14,"score":89,"score_detail":90,"sources":95,"tags":97,"search_phrases":101,"slug":104,"view_count":15,"doi":9,"paper":9,"created_at":105},922,"中国农业银行农业AI场景应用首批名单发布 数字金融赋能智慧农业","https:\u002F\u002Fwww.farmer.com.cn\u002F2026\u002F08\u002F23\u002Fwap_991026650.html","中国农业银行近日发布农业AI场景应用首批名单，覆盖种植、畜牧、渔业、农机、农产品加工等全产业链。聚焦大田种植精准管理、设施农业智能控制、畜禽养殖健康监测、水产养殖智能管控、农机作业智能调度、农产品质量追溯等核心场景，推动数字金融与智慧农业深度融合，为新型农业经营主体提供低门槛、低成本、易获取的数字普惠金融服务。","中国农网 2026-08-23","2026-08-23T00:00:00Z",73,{"impact":91,"substance":92,"depth":93,"authority":65,"freshness":68,"relevant":52,"comment":94},22,18,14,"农行发布农业AI场景应用名单，覆盖全产业链，具行业示范意义，但细节有限。",[96],{"name":87,"url":85},[74,98,22,99,100],"农业AI","农业银行","普惠金融",[102,103],"农业银行 数字金融 普惠金融 智慧农业","农业银行 数字金融","农业银行数字金融普惠金融智慧农业-922","2026-08-25T00:03:26.671160Z",{"id":107,"title":108,"url":109,"summary":110,"summary_zh":9,"content":111,"source_name":112,"source_url":9,"published_at":113,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":114,"score_detail":115,"sources":119,"tags":121,"search_phrases":125,"slug":128,"view_count":129,"doi":9,"paper":9,"created_at":130},791,"数字乡村发展与牧户非农就业:来自甘青川藏26县2790户牧户调查的证据(原标题:From Grasslands to Markets: Digital Rural Development and Herders' Non-Farm Employment)","https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503","兰州大学草地农业生态系统国家重点实验室吴忠安等团队基于2023-2025年甘青川藏26县2790户牧户调查数据,考察县级数字乡村发展与牧户非农就业关系。研究发现数字乡村发展与户主主要从事非农就业的可能性正相关,在系列稳健性和内生性检验后结果稳定;在第一产业基础较大的地区以及从事饲草种植的牧户中,正向关联较弱,表明当地生产结构和家庭劳动配置制约了数字机会向就业多元化转化。进一步分析表明数字金融和数字消费是数字乡村发展支持非农就业的两种潜在渠道。研究强调牧区就业收益既取决于数字接入,又取决于家庭劳动配置灵活性及非农机会可得性。","Logical Operator Operator\n\nSearch Text\n\nSearch Type\n\n_add\\_circle\\_outline_\n\n_remove\\_circle\\_outline_\n\n[![Image 1: sustainability-logo](https:\u002F\u002Fpub.mdpi-res.com\u002Fimg\u002Fjournals\u002Fsustainability-logo.png?3798e4e58c765aed)](https:\u002F\u002Fwww.mdpi.com\u002Fjournal\u002Fsustainability)\n\n## Article Menu\n\nFont Type:\n\n_Arial_ _Georgia_ _Verdana_\n\nFont Size:\n\nAa Aa Aa\n\nLine Spacing:\n\n__ __ __\n\nColumn Width:\n\n__ __ __\n\nBackground:\n\nOpen Access Article\n\nby \nZhongan Wu\n\n *[](mailto:wuzha2023@lzu.edu.cn), \nToba Stephen Olasehinde\n\n and \nYubing Fan\n\nState Key Laboratory of Herbage Improvement and Grassland Agro-Ecosystems, Chinese Grass Industry Development Strategy Research Center, College of Pastoral Agriculture Science and Technology, Lanzhou University, Lanzhou 730020, China\n\n*\n\nAuthor to whom correspondence should be addressed.\n\nSubmission received: 3 July 2026 \u002F Revised: 11 August 2026 \u002F Accepted: 12 August 2026 \u002F Published: 19 August 2026\n\n## Abstract\n\nAlthough natural resources and geographical factors limit herders’ employment options, the development of digital rural helps herding households compete on an equal footing by expanding employment opportunities. Using survey data collected from 2790 herder households in 26 counties across Gansu, Qinghai, Sichuan and Tibet from 2023 to 2025, this study examines the relationship between county-level digital rural development and non-farm employment. We find that digital rural development is positively associated with the likelihood that household heads engage primarily in non-farm employment. This finding remains stable across a series of robustness and endogeneity checks. Moreover, the positive association is weaker in areas with a larger primary-industry base and among households engaged in forage cultivation, suggesting that local production structures and household labor commitments constrain the conversion of digital opportunities into employment diversification. Further analysis indicates that digital finance and digital consumption are two potential channels through which digital rural development supports non-farm employment. These findings highlight that the employment benefits of digital development depend not only on digital access but also on the flexibility of household labor allocation and the availability of non-farm opportunities in pastoral areas.\n\n## 1. Introduction\n\nArtificial intelligence, Big Data, and mobile internet are examples of digital technology that are spreading to all parts of the economy and society, changing how production is organized and people work [[1](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B1-sustainability-18-08503)]. Given the above, China released the Digital Rural Development Plan in 2018 to promote the construction of a digital rural society and rural revitalization. As 5G networks expand into rural and pastoral areas, the all-weather demonstration project for e-commerce in rural areas continues to advance, and the coverage of digital inclusive finance is gradually expanding. At the same time, digital infrastructure and services are being rolled out in the grassland region at an ever-increasing rate [[2](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B2-sustainability-18-08503)]. However, because of their remote areas, fragile ecosystems and unique cultures, these large pastoral areas have been lacking in development resources for a long time. Traditional livestock farming is limited by the productivity of natural grasslands, ecological red lines and fluctuations in market prices, and a single grazing-based livelihood model cannot provide stable income growth for herder households [[3](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B3-sustainability-18-08503)]. Therefore, expanding non-farm employment opportunities and promoting the transfer of the pastoral workforce to the secondary and tertiary sectors are urgently needed to consolidate the results of poverty alleviation and prevent a relapse into poverty, as well as to achieve all-round rejuvenation of pastoral areas.\n\nNon-farm employment provides herder households with an alternative to relying entirely on traditional livestock production. This resulting income mix can reduce exposure to livestock losses and strengthen resilience to market, climatic and health shocks [[4](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B4-sustainability-18-08503)]. Especially against the backdrop of increasing resource and environmental constraints in pastoral areas and shrinking growth prospects for traditional livestock farming, promoting the rational flow of labor into non-farm sectors is not only a practical necessity for herding households to enhance the resilience of their livelihoods, but also a key direction for industrial restructuring and socioeconomic transformation in pastoral areas. However, the formation of non-farm employment is not simply a matter of individual choice, as herders’ decision-making is influenced by a range of factors, including their ability to access information, job search costs, financial constraints, and the extent of market access [[5](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B5-sustainability-18-08503),[6](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B6-sustainability-18-08503)]. The reason why the development of digital rural may have an impact on herders’ non-farm employment is that it can, to some extent, alleviate these constraints.\n\nExisting research has provided a wealth of discussion on the relationship between the digital economy, information and communication technologies, and rural labor migration, as well as the growth of farm household income [[7](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B7-sustainability-18-08503),[8](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B8-sustainability-18-08503),[9](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B9-sustainability-18-08503),[10](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B10-sustainability-18-08503)]. A large body of the literature indicates that digital development can improve the information environment, expand market access, alleviate financial exclusion, and, to some extent, facilitate the shift in rural labor toward the non-farm sector [[11](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B11-sustainability-18-08503),[12](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B12-sustainability-18-08503),[13](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B13-sustainability-18-08503)]. Some studies have also indicated that the development of internet usage, digital finance, and e-commerce has a positive impact on farmers’ non-farm employment, entrepreneurial activities, and income growth [[14](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B14-sustainability-18-08503),[15](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B15-sustainability-18-08503),[16](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B16-sustainability-18-08503),[17](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B17-sustainability-18-08503)]. More broadly, infrastructure development and technological progress can promote economic diversification and strengthen livelihood resilience in peripheral and resource-dependent regions [[18](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B18-sustainability-18-08503)]. In terms of research subjects, most of the existing studies have focused on farmers and have paid relatively little attention to herders, another group in pastoral areas. Herders’ employment decisions are influenced by the general labor market, but they are also restricted by other factors, such as grassland management systems, livestock production cycles, household livestock assets and grazing practices; therefore, their non-farm employment behavior is relatively context-specific and constrained [[19](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B19-sustainability-18-08503),[20](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B20-sustainability-18-08503)]. Most studies have focused on the general inclusive benefits of digital technology for research, but there is a lack of specific research on how digital rural initiatives operate in the particular spatial conditions of pastoral areas and whether their effects differ among different groups and regions.\n\nThis paper takes as its focus the non-farm employment of herders and studies the effect of digital rural development on the transformation of herders’ employment, as well as the mechanisms through which this change occurs. Based on this, this paper seeks to answer the following questions: can digital rural development promote non-farm employment among herders?; through what mechanisms does it primarily achieve this?; and does this impact vary depending on individual characteristics, spatial conditions, and the level of development in each county? This study makes three contributions. First, it conceptualizes digital rural development in pastoral areas as a mechanism for translating digital access into non-farm employment opportunities, emphasizing the role of household labor constraints in livestock-based livelihoods. In doing so, it extends the literature on digital rural development and labor transition by showing that improved digital access does not automatically lead to labor reallocation, because pastoral production commitments may constrain herders’ ability to respond to new employment opportunities. Second, it integrates survey data from 2790 herding households across Gansu, Qinghai, Sichuan, and Tibet, with a county-level Digital Rural Development Index, providing evidence from underrepresented pastoral production systems. Third, it examines how structural and household-level conditions shape the relationship between digital rural development and non-farm employment, highlighting the importance of contextual heterogeneity in pastoral regions. By linking digital development with livelihood diversification, the study also contributes to the rural sustainability literature by showing that the employment benefits of digitalization depend on household labor flexibility and local economic conditions.\n\n## 2. Theoretical Framework\n\n### 2.1. Direct Effects of Digital Rural Development on Herders’ Non-Farm Employment\n\nCompared to typical rural areas, pastoral regions face long-standing constraints such as geographical dispersion, poor transportation, limited access to information and insufficient market access. As a result, employment options for herding households are often confined to traditional livestock farming. Access to non-farm employment depends largely on personal network or occasional face-to-face contact [[21](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B21-sustainability-18-08503)]. In light of the above, the construction of an information infrastructure, the extension of platform coverage and the broad application of digital tools in the development of digital rural have created new employment opportunities for herders outside of traditional sectors.\n\nAccording to the theory of information asymmetry, if market participants cannot obtain the required information in time and to their full satisfaction, they will bear a higher search cost, be less efficient in matching, and thus lose opportunities in their decision-making [[22](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B22-sustainability-18-08503),[23](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B23-sustainability-18-08503)]. The development of digital rural is changing how herders get jobs and market information. New technologies are now available to help herders stay informed about new employment prospects outside farming, wage levels, working locations and required skills [[24](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B24-sustainability-18-08503)]. In this way, information barriers between herders and the outside labor market have been reduced. Thus, employment information is no longer scarce and fragmented, but rather open and accessible. As a result, herders’ awareness of non-farm employment and their opportunities for employment have significantly increased [[25](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B25-sustainability-18-08503)].\n\nAs is well-known, non-farm employment not only involves the movement of labor across sectors, but also entails costs associated with information search, transportation, payment settlements and job matching. Transaction cost theory suggests that improvements in market efficiency depend largely on reductions in transaction costs [[26](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B26-sustainability-18-08503),[27](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B27-sustainability-18-08503)]. The development of digital rural has reduced the time costs, communication costs, and institutional friction faced by herders when seeking non-farm employment through information dissemination, digital payments and platform integration. Looking further, improvements in the information environment and reductions in transaction costs do not stop at the level of access to employment opportunities. Rather, they continue to influence labor allocation within herding households. In this context, non-farm employment reflects the reallocation of household labor from livestock production to non-farm sector [[28](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B28-sustainability-18-08503),[29](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B29-sustainability-18-08503)]. Consequently, household labor may gradually shift away from traditional livestock production toward a wider range of market-based non-farm activities. In light of the above, this paper proposes the following hypothesis:\n\n**H1.**\n\nDigital rural development is positively associated with household non-farm employment.\n\n### 2.2. Boundary Conditions of the Employment Effects of Digital Rural Development\n\nDigital rural development can improve information access, reduce job-search costs, and strengthen links to external markets. Nevertheless, these improvements do not automatically generate non-farm employment. Whether herders convert digital opportunities into employment also depends on local industrial structure and household production systems. We therefore examine the moderating roles of primary-industry development and forage cultivation.\n\nStructural transformation involves the reallocation of labor from primary production toward secondary and tertiary activities [[30](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B30-sustainability-18-08503)]. Primary-industry value added at the county level measures the absolute scale of agricultural and livestock output, rather than the sector’s share of the local economy or its relative return. A higher value may nevertheless indicate a larger agricultural and livestock production base. In such counties, digital tools may be used more extensively for production management, input procurement and product marketing within the primary sector [[31](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B31-sustainability-18-08503)]. Digitalization may lower search, matching, and market-access costs while also improving production within the primary sector. Where digital services are used mainly for production management and product marketing, these benefits may be absorbed within agricultural and pastoral activities rather than translated into labor reallocation [[32](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B32-sustainability-18-08503)]. The positive association between digital rural development and herders’ participation in non-farm employment may therefore be weaker in counties with a larger primary industry base. We propose:\n\n**H2a.**\n\nHigher primary industry value added in a county weakens the positive association between digital rural development and household non-farm employment.\n\nForage cultivation is closely integrated with livestock production and requires labor for planting, management, harvesting, transport and storage [[33](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B33-sustainability-18-08503)]. These tasks increase household labor demands, particularly during busy seasons. With limited household labor, forage cultivation may reduce the time available for non-farm employment [[34](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B34-sustainability-18-08503)]. It also binds cropland, forage, livestock and family labor into a more integrated production system. Households may therefore have less flexibility to act on digital employment information and market opportunities.\n\n**H2b.**\n\nForage cultivation weakens the positive association between digital rural development and household non-farm employment.\n\n### 2.3. Transmission Mechanisms of Digital Rural Development\n\nIn most cases, funding is a key factor limiting herders’ ability to engage in any activities. Continuing livestock production, starting a business, and seeking employment elsewhere may all require financial resources, while reliable payment services can facilitate related transactions [[35](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B35-sustainability-18-08503),[36](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B36-sustainability-18-08503)]. However, due to factors such as remote locations, a lack of financial service outlets and exclusion from the traditional financial system, herders have long faced challenges in accessing financial services [[37](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B37-sustainability-18-08503)]. The development of digital rural has driven the shift in financial services toward online, convenient and grassroots-level delivery, providing herders with more accessible channels for payments, credit and financial services. The expansion of financial services to grassroots levels represents not only an increase in the scale of financial operations but also an extension of service coverage and an improvement in the efficiency of resource allocation. Digital rural development can facilitate the expansion of digital finance. Greater access to digital financial services may ease the credit and liquidity constraints faced by herding households. This, in turn, can lower the financial barriers to participation in non-farm employment and entrepreneurial activities [[38](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B38-sustainability-18-08503)]. At the same time, development of digital finance has reduced the transaction costs for labor transfer and provided a relatively easy way for herders to access external labor markets [[39](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B39-sustainability-18-08503)]. Therefore, the development of digital rural areas should not only increase the supply of financial products but also improve the financial situation for herders’ non-market activities at a deeper level and enhance their ability to engage in non-farm employment. Based on the above analysis, this study puts forward the following hypothesis:\n\n**H3.**\n\nDigital rural development promotes herders’ non-farm employment by expanding digital finance.\n\nWith the spread of online shopping, digital payments, and platform services, herding households have become more closely connected to digital markets. These developments have changed how households purchase goods, make payments, and access consumer services [[40](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B40-sustainability-18-08503)].\n\nThe shift in consumption patterns is not merely a change in lifestyle. It also influences household employment choices through adjustments in income needs and labor allocation. Traditional herding households often rely on self-produced goods and local supplies. Digital rural development has expanded their access to external markets. Their consumption may therefore become more dependent on purchased goods and services. While households now have a wider range of choices regarding goods and services, they also have a greater need for a stable cash flow to support these choices [[41](https:\u002F\u002Fwww.mdpi.com\u002F2071-1050\u002F18\u002F16\u002F8503#B41-sustainability-18-08503)]. The increase in digital consumption suggests that more money will be needed by households. As a result, there has been a change in demand for labor, and many people have begun to leave traditional pastoralism for non-agricultural work in search of a more stable in","MDPI Sustainability 18(16):8503 2026-08-19","2026-08-19T00:00:00Z",81,{"impact":92,"substance":91,"depth":116,"authority":66,"freshness":117,"relevant":52,"comment":118},20,8,"基于2790户牧户调查的实证研究，揭示数字乡村对非农就业的促进作用及条件，数据详实，结论可靠，对牧区数字乡村政策有参考价值。",[120],{"name":112,"url":109},[73,122,22,123,124],"Sustainability","非农就业","牧户",[126,127],"数字乡村 数字金融 非农就业 牧户","数字乡村 数字金融","数字乡村数字金融非农就业牧户-791",2,"2026-08-21T00:05:57.571383Z",{"id":132,"title":133,"url":134,"summary":135,"summary_zh":9,"content":136,"source_name":137,"source_url":9,"published_at":138,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":139,"score_detail":140,"sources":145,"tags":147,"search_phrases":151,"slug":154,"view_count":143,"doi":9,"paper":9,"created_at":155},418,"金融科技倍增效应:将政府关注转化为中国农业经济的抗风险能力","https:\u002F\u002Fwww.ebiotrade.com\u002Fnewsf\u002F2026-8\u002F20260808082614715.htm","发表于《Finance Research Letters》\u002FarXiv等平台。基于中国地级市面板数据,研究地方政府对金融科技重视程度如何影响农业经济抗风险能力,从而弥补既有文献多从市场层面金融供给(如数字金融普惠、农业保险)出发、对政府推动金融科技发展关注不足的空白。研究框架从政府重视程度入手,分析其对农业经济抗风险能力产生的因果影响,使用地级市而非省级综合指数,体现不同地级市之间政府行为差异。结论指出,地方政府对金融科技的重视程度直接影响各种制度安排,包括数字金融基础设施和银企对接平台,数字乡村发展和保险投资共同提升农业经济抗风险能力。","《Finance Research Letters》：The fintech multiplier: turning government attention into agricultural economic resilience in China\n\n**编辑推荐：**\n\n摘要农业是传统金融服务最不足的领域：抵押品限制和信息不对称使得农业生产者容易受到冲击。本研究探讨了上游层面的政府干预是否能够缓解这些问题。我们通过中国地级市的政府工作报告来衡量地方政府对金融科技的重视程度，并将其与农业经济抗风险能力指标联系起来。那些将金融科技纳入治理议程的城市，\n\n## 摘要\n\n农业是传统金融服务最不足的领域：抵押品限制和信息不对称使得农业生产者容易受到冲击。本研究探讨了上游层面的政府干预是否能够缓解这些问题。我们通过中国地级市的政府工作报告来衡量地方政府对金融科技的重视程度，并将其与农业经济抗风险能力指标联系起来。那些将金融科技纳入治理议程的城市，其农业经济抗风险能力更强：重视程度每增加一个标准差，抗风险能力就会提升约十分之一个标准差；这一效应在工具变量估计和多种稳健性检验中依然成立。证据表明，风险分担是主要作用机制——政府对金融科技的重视推动了农业保险市场的发展，而产业结构升级的作用则相对较弱。在数字金融服务覆盖更广的地区，这一效应更为显著，这说明政府重视与市场化金融之间存在互补关系。因此，政府重视可作为一种政策手段，帮助金融资源惠及那些市场服务不足的领域。\n\n## 引言\n\n农业领域面临较高的风险（Feng等人，2025年）。根据中国应急管理部的数据，2024年的自然灾害影响了9413万人，直接经济损失达4011.1亿元人民币。再加上国际大宗商品市场的波动以及生产成本的上升，农业经济长期处于多重冲击之下。在这种背景下，农业经济的抗风险能力已成为衡量农业可持续发展和国家粮食安全的重要指标。它指的是农业系统在外部冲击面前能够承受干扰、维持核心功能，并实现恢复与结构转型的能力（Holling，1973年；Folke等人，2010年；Martin和Sunley，2015年）。因此，通过更好的风险分担和资源分配来增强农业抗风险能力，已成为一项重要的政策目标（Kramer等人，2022年）。\n\n然而，农业却是传统金融体系服务最不足的领域。农业生产者规模较小，缺乏合格的抵押品，信用记录也不透明（Karlan等人，2014年）。因此，传统的信贷和保险产品长期以来都无法满足农业需求，金融摩擦也限制了农业系统抵御冲击的能力。金融科技相关研究指出，数字金融和替代数据有助于筛选没有信用记录的借款人（Berg等人，2020年），而无现金支付产生的可验证交易记录则降低了小企业的融资门槛和违约风险（Ghosh等人，2026年）。金融科技能够服务于传统银行无法覆盖的客户（Di Maggio和Yao，2021年），还能通过与银行信贷的互补作用提升企业的融资渠道（Beaumont等人，2026年）。这些机制非常适合信息严重不对称的农业领域。除了信贷渠道外，金融科技和数字化转型还与新兴市场中的银行和企业的可持续发展及治理效果有关（Alim和Mansour，2026年；Mansour等人，2026年）。然而，在那些金融摩擦最为严重的农业领域，这类研究却相对较少。最近的一些研究开始从市场角度弥补这一空白：数字乡村发展和保险投资共同提升了农业经济的抗风险能力（Liu等人，2025年），而数字金融则推动了农业的技术升级（Zhu和Gao，2026年）。另一个空白在于，金融科技的本地采用和普及并非纯粹由市场驱动，而是依赖于地方政府的政策和资源支持。地方政府对金融科技的重视程度直接影响着各种制度安排，包括数字金融基础设施和银企对接平台。有证据表明，政府对此的重视程度会影响数字经济的效率（Li和Yue，2025年）。因此，要了解金融科技对农业抗风险能力的影响，就必须从政府重视程度入手。\n\n目前，相关研究尚未就此问题提供系统的证据，存在两个方面的局限。首先，关于农业经济抗风险能力的金融驱动因素的研究多集中在市场层面的金融供给，如数字金融普惠和农业保险（Biagini，2025年；Wang和Zhang，2025年）。很少有研究从更上游的政策制定层面，即政府重视程度出发，来分析政府推动的本地金融科技发展对农业抗风险能力的因果影响。其次，现有研究通常使用省级综合指数来衡量金融发展水平，但这无法体现不同地级市之间政府行为的差异。而地方政府的年度治理议程——政府工作报告，则为识别政府的重视程度提供了可靠依据：报告中与金融科技相关的表述强度是可以量化的，不同城市之间具有可比性，而且还在资源投入之前出现（Li和Yue，2025年）。\n\n鉴于这些不足，本研究利用2011年至2023年中国各地级市的面板数据展开分析。我们通过文本分析从地级市政府工作报告中提取与金融科技相关的词汇频率，进而构建出地方政府对金融科技重视程度的指数。同时，我们还建立了城市层面的农业经济抗风险能力指数，以研究政府对金融科技的重视程度对其抗风险能力的影响及其作用机制。研究得出三个结论：首先，地方政府对金融科技的重视程度显著提升了农业经济的抗风险能力，且这一结果在多种稳健性检验中依然成立。其次，风险分担是主要作用机制——政府对金融科技的重视程度提升了农业保险的普及率，而产业结构升级的作用则相对较弱。第三，当数字金融服务的覆盖范围更广时，这一效应会更加明显，尤其体现在数字金融的覆盖度和使用率方面。此外，异质性分析表明，这一效应在不同类型的城市中都有体现，并非仅限于少数城市。\n\n本研究在三个方面丰富了相关文献。首先，它拓展了关于农业经济抗风险能力决定因素的研究视角。现有研究多关注数字金融普惠和农业保险等市场层面的金融工具对农业抗风险能力的影响（Biagini，2025年；Wang和Zhang，2025年），而本研究则从地方政府重视程度这一角度揭示了金融科技提升农业抗风险能力的机制，为理解转型经济体中政府、金融与农业之间的传导链条提供了新证据。其次，在测量方法上，我们通过分析地级市政府工作报告的文本内容来构建金融科技重视程度指标，这一方法避免了以往测量方法的两个缺陷：省级综合指数无法反映省内的差异，而事后的金融发展指标也无法区分政策意图与市场结果。这样一来，基于文本的分析方法也被引入了农业金融研究领域（Li和Yue，2025年）。第三，我们发现了通过农业保险普及实现风险分担的机制。金融科技相关研究强调技术在缓解信息不对称中的作用（Berg等人，2020年；Ghosh等人，2026年），而农业抗风险能力研究则侧重于保险的收入平滑功能（Biagini，2025年）。我们通过研究表明，政府对金融科技的重视是通过完善农业保险市场来提升农业经济抗风险能力的。\n\n## 章节节选\n\n## 理论分析与假设\n\n这一分析基于三个核心要素。基于重视程度的观点解释了上游机制：一个组织关注什么，就会决定如何配置资源（Ocasio，1997年）。信贷配给理论则解释了其中的摩擦：信息不对称导致那些缺乏抵押品和可验证信用记录的借款人难以获得贷款（Stiglitz和Weiss，1981年）。风险管理理论则说明了下游机制：市场保险能够在不同自然状态下平滑收入，从而补充自我保护措施（Ehrlich）\n\n## 样本与数据\n\n我们根据北京大学数字金融普惠指数的可得性，构建了2011年至2023年间278个中国地级市的面板数据。地方政府对金融科技的重视程度是通过从政府门户网站和官方公报中收集的地级市政府工作报告来确定的；农业产出、行业附加值以及控制变量则来自《中国城市统计年鉴》和各省年鉴；农业保险保费数据则来自《中国保险年鉴》\n\n## 基准结果\n\n表2展示了逐步加入控制变量后的基准估计结果。在所有四列中，金融科技重视程度的系数均为正，且在1%的水平上显著，而且随着控制变量的加入，其数值保持稳定，这说明该估计结果并非由其他相关遗漏变量所导致。在第四列中，重视程度每增加一个标准差（7.33个单位），农业经济抗风险能力就会提升0.14，约为一个标准差的十分之一，对于单一政策渠道而言，这一效应相当显著。\n\n## 结论\n\n提升农业抗风险能力关乎一个社会如何组织风险分担和资本配置。本研究指出了金融与农业研究领域中被忽视的一个关键主体——那些重视金融科技发展的地方政府。当一个地级市将金融科技纳入治理议程后，其农业经济就能更好地抵御各类冲击。这一效应相当显著：即使在排除省内差异的条件下，这一效应依然存在，而且并非\n\n## 作者声明\n\n我们声明，本手稿为原创内容，未曾发表过，也未被其他机构考虑发表。\n\n我们确认，所有署名作者均已阅读并同意本手稿的内容，也没有其他符合作者资格但未列入名单的人士。同时，我们也确认手稿中作者的排序已得到所有人的认可。\n\n## 资金支持\n\n本研究得到了“农村建设与农村治理指标体系研究项目”以及“‘十五’规划时期农村发展目标”项目（项目编号GXZC2024-C3-005884-JZZB）、“广西自然科学基金联合专项”：‘数据要素与人工智能技术协同推动广西制造业培育新质生产力的机制研究’（项目编号2025JJH180035）的支持。\n\n## CRediT作者贡献说明\n\n**陆琪：**概念设计、数据整理、形式分析、方法论、软件应用、验证、初稿撰写。**文军：**资金筹集、项目管理、资源协调、修改润色。**朱天奇：**数据收集、初稿撰写、修改润色。**曹金华：**资金筹集、监督指导、修改润色。\n\n陆琪|文军|朱天奇|曹金华\n\n广西大学农村与区域发展研究院，中国南宁市530004","Finance Research Letters 2026-08-08","2026-08-08T00:00:00Z",82,{"impact":91,"substance":141,"depth":142,"authority":93,"freshness":143,"relevant":52,"comment":144},24,19,3,"核心期刊论文，研究地方政府金融科技重视度对农业抗风险能力的影响，方法新颖，数据详实，但时效性稍差。",[146],{"name":137,"url":134},[73,148,149,22,150],"金融科技","农业抗风险","政府关注",[152,153],"农业抗风险 政府关注 数字乡村 数字金融","农业抗风险 政府关注","农业抗风险政府关注数字乡村数字金融-418","2026-08-11T23:57:03.260961Z",{"id":157,"title":158,"url":159,"summary":160,"summary_zh":9,"content":161,"source_name":162,"source_url":9,"published_at":163,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":164,"score_detail":165,"sources":167,"tags":169,"search_phrases":174,"slug":177,"view_count":117,"doi":178,"paper":9,"created_at":179},175,"Leveraging Large Language Models (LLMs) for GeoAI-enabled Digital Agro-advisory","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract","International Rice Research Institute (India) 的 Shalini Gakhar 和 Raj Kumar Singh 在综述中系统分析 LLMs 在将复杂 GeoAI 数据转化为面向农户的可操作建议方面的潜力。区分了 LLM 已经验证的能力与农业遥感中的潜在应用，强调 RAG（检索增强生成）与 SFT（监督微调）在缓解误读风险中的重要性；探讨多模态方法在作物健康评估和生物物理异常识别的应用，并主张'人在回路'决策方式。综述最后展望了从田间结果中迭代学习的前瞻性研究范围和闭环系统。","## Abstract\n\nDigital agriculture has undergone a profound transformation driven by rapid advances in Earth observation, unmanned aerial vehicles and advanced data processing techniques. Despite this progress, climate-smart agriculture management faces a critical challenge: a farmer-centric agro-advisory system. Farmers and extension workers have limited access to GeoAI-based agro-advisory services, primarily due to limited data access and insufficient technical support. This study synthesises the potential of Large Language Models to address this challenge by translating complex GeoAI data into actionable and human-centric advice. The review examines gaps, recent advancements, and the model architectures required to adapt general-purpose LLMs for remote-sensing-based crop monitoring and management. We distinguish between experimentally validated capabilities of LLMs and the broader prospective applications proposed for agricultural remote sensing, noting that many current advances are primarily driven by multimodal foundation models, computer vision, and GeoAI systems rather than standalone LLMs. The study highlights the importance of advanced Retrieval-Augmented Generation and Supervised Fine-Tuning in agronomic science and mitigating the risk of misinterpretation. Further, we examine the emerging capabilities of multimodal approaches, which can seamlessly integrate visualisation and textual reasoning to support stakeholders in assessing crop health conditions and biophysical anomalies. The representative case studies, spanning multiple geographies and including voice-based advisory systems, demonstrate the shift from static advisory tools to dynamic, interactive recommendation systems. We highlight current challenges, such as the need for region-specific fine-tuning, data governance, and operation in low-connectivity environments, and advocate for a “human-in-the-loop” approach to decision-making. Through this, the LLMs will function as co-pilots assisting multiple stakeholders rather than as autonomous decision-makers vulnerable to biased output or hallucination problems. The review concludes by outlining a forward-looking research scope and closed-loop systems that iteratively learn from field outcomes.\n\n## 1 Introduction\n\nOver the past decade, digital agriculture has undergone a profound transformation driven by rapid advances in Earth observation, unmanned aerial vehicles (UAVs), Internet of Things (IoT), and proximal sensing technologies. These developments have enabled continuous, multiscale, and unprecedented volumes of spatiotemporal data to be generated for monitoring crops, soils, hydro-ecology, and agroecosystem processes (). Operational satellite missions such as Landsat, Sentinel, MODIS, LISS, Cartosat, WorldView, PlanetScope, combined with very high-resolution UAV imagery, now form the backbone of precision agriculture, supporting applications including crop monitoring, yield estimation, cropping practices and the implementation of climate-smart agricultural practices (; ). Despite the growing maturity of remote sensing technologies, the increasing availability of data has exposed a fundamental limitation in digital agriculture: the interpretation and convergence bottleneck. Remote sensing data are typically expressed as spectral reflectance or backscatter values, vegetation indices, and derived biophysical parameters, which require substantial agronomic and contextual expertise to interpret and apply to improve crop management. Translating these quantitative indicators into actionable recommendations requires an understanding of crop type, phenological stage, soil characteristics, climatic conditions, and management history, which is complex (). This complexity constrains the timely delivery of decision-support information, particularly in smallholder-dominated agricultural systems where access to trained extension services remains limited and costly (). Recent advances in Large Language Models (LLMs) offer a promising framework for addressing these challenges. LLMs such as the Generative Pre-trained Transformer (GPT), Large Language Model Meta AI (LLaMA), and Pathways Language Model (PaLM) are based on transformer architectures that model long-range dependencies, support contextual reasoning, and enable large-scale knowledge synthesis (; ). Unlike conventional machine learning models, which are optimised for narrowly defined prediction tasks, LLMs are designed to interpret, explain, and adapt to complex information across domains. These capabilities position LLMs as an effective interface layer for translating remote-sensing-derived indicators into human-interpretable, decision-relevant guidance for agronomists, extension agents, and farmers.\n\nThe potential of LLM-enabled agricultural advisory systems is further enhanced by recent advances in multimodal foundation models and vision-language frameworks that integrate visual, textual, and geospatial information. It is important, however, to distinguish between standalone LLMs and multimodal computer vision systems. Standalone LLMs are primarily designed for language understanding, knowledge synthesis, reasoning, and natural-language generation, whereas remote sensing image interpretation is largely performed by computer vision, vision-language, and multimodal foundation models. Recent vision-language models (VLMs) and geospatial foundation models have demonstrated increasing capability in analysing remote sensing imagery, field photographs, and UAV observations for applications such as crop stress detection, disease identification, canopy trait assessment, phenological monitoring, and interpretation of spectral anomalies (; ). These models generate structured visual insights that can subsequently be translated into actionable recommendations through LLM-based reasoning and conversational interfaces. This distinction is particularly important given the exponential growth of Earth observation imagery, driven by the expansion of satellite constellations and the widespread adoption of UAV platforms for agricultural management. While multimodal foundation models and GeoAI systems increasingly perform perception and analytical tasks associated with image understanding, LLMs primarily contextualise these outputs, integrate agronomic knowledge, retrieve relevant information from external knowledge bases, and communicate recommendations in a human-interpretable form. Therefore, the emerging opportunity lies not in replacing remote sensing analytics with LLMs, but in integrating multimodal perception systems with language-based reasoning frameworks to support end-to-end agricultural decision-making. Despite this rapid progress, a systematic understanding of how LLMs can be effectively integrated into remote sensing-based agricultural workflows remains lacking. The existing literature predominantly focuses on remote sensing for crop monitoring or the application of LLMs in broader agricultural contexts, leaving a crucial gap in current knowledge. To date and to our knowledge, no comprehensive review has synthesised how LLMs specifically enhance the interpretation, communication, and decision-support functionalities of remote sensing in digital agriculture. Moreover, critical deployment challenges such as the risk of model “hallucination”, the necessity for region-specific fine-tuning, and strategies for effective implementation remain largely unexplored. Existing studies tend to focus either on remote sensing-based crop monitoring and modelling (; ) or on the application of LLMs in broader agricultural advisory and knowledge systems () with relatively little synthesis across these domains. As agricultural monitoring systems increasingly transition from descriptive assessment to operational advisory services, the need for intelligent interpretation frameworks becomes more acute. LLMs are efficient in data mining through historical observations and practices, textual reports, and agronomic knowledge bases, thereby reducing human effort. The Earth observation foundation models highlight the growing integration of geospatial AI, multimodal learning, and LLM-assisted agricultural analytics. For instance, the IBM-NASA _Prithvi_ model uses multi-temporal satellite imagery and transformer-based learning to support applications such as crop monitoring, land-use mapping, and environmental assessment (). Similarly, the European Space Agency (ESA) _WorldCereal_ initiative applies AI-driven analysis of Sentinel satellite data to generate global crop maps for major cereals, including wheat and maize (). Although these systems are not standalone LLMs, they demonstrate how foundation models and multimodal geospatial AI are advancing scalable agricultural monitoring and advisory systems. Future integration of these Earth observation models with conversational LLM frameworks could further improve farmer-centric decision support and agro-advisory services (). Therefore, a coherent understanding of how LLMs can be systematically integrated into remote sensing-driven agricultural workflows will improve farm management and decision support.\n\nThis review aims to address the following vital gaps: (a) To synthesise the potential for integrating advanced remote sensing tools\u002Ftechniques and LLMs for precision\u002Fdigital agriculture and advisory generation. (b) To present representative case studies illustrating the integration of LLMs with satellite- and UAV-based crop monitoring workflows, focusing on rice-based cropping systems. (c) To critically examine technical, operational, and governance challenges, including reliability, validation, and ethical considerations associated with LLM-based advisory. (d) To outline a forward-looking roadmap\u002Fframework that prioritise multimodal model development, regionally adaptive fine-tuning, and edge or low-bandwidth deployment strategies.\n\nBy bridging remote sensing analytics with LLM-based knowledge representation, this review argues that foundation models have the potential to significantly enhance the scalability, contextual relevance, and inclusivity of digital agriculture. When appropriately grounded in agronomic knowledge and validated against observational data, LLMs can transform geospatial information from an expert-centric analytical resource into a human-centred and actionable advisory ecosystem that supports informed decision-making (). [Figure 1](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#F1) illustrates the knowledge translation pipeline from remote sensing data acquisition through biophysical indicator derivation to LLM-driven farm-level advisory, highlighting the interpretation gap that LLMs are positioned to bridge.\n\nFIGURE 1\n\n### 1.1 Scope and methodology of this review\n\nThis review follows a structured narrative and conceptual review approach to map the intersection of LLMs, remote sensing, and agricultural advisory systems. The literature search was conducted across Web of Science, Scopus, Google Scholar, and the arXiv preprint repository using Boolean combinations of key terms including “Large Language Model,” “LLM,” “Generative AI,” “remote sensing,” “digital agriculture,” “precision farming,” “crop advisory,” “farm advisory chatbot,” “RAG agriculture,” and “multimodal geospatial AI.” The search covered publications from 2010 to 2025, with emphasis on post-2020 literature to capture the rapidly evolving LLM landscape. Inclusion criteria required that studies: (i) explicitly addressed the application of LLMs or conversational AI in agricultural contexts, (ii) incorporated remote sensing data or geospatial analytics, or (iii) presented technical architectures (SFT, RAG, RLHF, multimodal fusion) with relevance to agro-advisory generation. Studies that focused exclusively on conventional machine learning for crop classification, without an advisory or interpretive component, were excluded. Grey literature, including technical reports from CGIAR centres, FAO, and national agricultural extension bodies, was also consulted to capture operationally deployed systems not yet represented in peer-reviewed literature. A total of over 330 sources were screened, of which 97 were retained for detailed synthesis ([Figure 2](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#F2)). The review is organised thematically rather than chronologically, progressing from the remote sensing foundation ([Section 2](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#s2)) through LLM adaptation pathways ([Section 3](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#s3)), representative field deployments ([Section 4](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#s4)), deployment challenges ([Section 5](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#s5)), and future research frontiers ([Section 6](https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fremote-sensing\u002Farticles\u002F10.3389\u002Ffrsen.2026.1839369\u002Fabstract#s6)).\n\nFIGURE 2\n\n## 2 Overview of remote sensing in digital agriculture and associated challenges\n\nRemote sensing has become a cornerstone of modern digital agriculture, providing the spatial and temporal multi-dimensional information needed to monitor, assess, and manage complex agroecosystems. The expansion of various satellite platforms, UAV-based imaging systems, and sensor-integrated field networks has revolutionised our ability to observe crop conditions, map spatiotemporal variability, and optimise resource use (; ). These advances fuel precision agriculture, enabling data-driven decisions from crop suitability, fertilisation and pest management, crop health monitoring, water resource utilization, and yield forecasting. Although a large proportion of the latest remote sensing-based crop management uses machine learning models, most of these trained models operate in silos and require human intervention for consistent interpretation and translation into actionable advice for farmers, extension workers, and policymakers.\n\n### 2.1 Landscaping the diverse remote sensing applications in agriculture\n\nSatellite remote sensing remains the backbone of national, regional, and global agricultural monitoring systems, providing consistent, spatially explicit observations of cropland dynamics over large areas. A diverse constellation of publicly accessible and commercial Earth observation platforms, including multispectral optical sensors such as Sentinel-2, Landsat-8\u002F9, and PlanetScope, MODIS, WorldView, LISS, as well as synthetic aperture radar (SAR) systems such as Sentinel-1, EOS-04, and hyperspectral data such as HySI and PRISMA, and satellite data-derived products such as soil moisture (by SMAP), now deliver high-frequency data streams essential for operational agricultural applications (; ; ). Despite these advances, a key limitation of satellite remote sensing lies in its spatial, spectral, and temporal resolutions. Most available data have revisit times of 5–6 days–20 days and moderate spatial resolution, which limits their application during the crucial crop-monitoring phase. While freely available sensors typically operate at spatial resolutions of 10–30 m, which are often insufficient to capture within-field variability in small farmlands with field sizes below 1 ha (; ). As a result, satellite products are highly effective for strategic monitoring and policy-level assessments in homogeneous fields. However, they are less suited for heterogeneous small field-level decision-making without costly high-resolution imagery.\n\nTo overcome the spatial resolution limitations of satellite-based observations, UAVs have emerged as a critical component of field-scale agricultural intelligence. UAV platforms provide centimetre-level spatial detail with flexible deployment schedules, allowing crop condition monitoring at user-defined temporal frequencies in key phenological stages (; ). Advanced UAV systems can be equipped with a diverse suite of lightweight sensors, each contributing complementary information on crop status. High-resolution RGB cameras generate detailed orthomosaics that support plant stand assessment, lodging detection, weed mapping, and canopy structure analysis (). Recent advances in deep learning have further expanded the capabilities of UAV and mobile vision systems beyond traditional vegetation index mapping. Convolutional neural networks and real-time object detection frameworks, particularly the YOLO family of models, have enabled automated crop detection, fruit localisation, segmentation, counting, and growth-stage recognition at field and greenhouse scales. For example, developed an integrated YOLO-based system for capsicum detection, segmentation, growth-stage classification, fruit counting, and real-time mobile identification. Such perception-oriented computer vision models provide detailed crop-level intelligence that can complement satellite-derived biophysical indicators and serve as important upstream information sources for multimodal advisory systems. In future GeoAI-enabled agricultural platforms, outputs from UAV-based detection and crop-stage recognition models could be integrated with LLM reasoning frameworks to generate more context-aware recommendations on harvesting, crop management, and resource optimisation. In parallel, thermal infrared sensors enable the mapping of canopy temperature, which serves as a proxy for plant water status and transpiration efficiency, supporting irrigation scheduling and drought stress detection at fine spatial scales (). These high-fidelity datasets are already reshaping agricultural research and innovation pipelines. For farmers and extension agents, a detailed crop health or thermal anomaly map may indicate variability. However, it does not inherently convey why stress is occurring or what management action should follow. Bridging this interpretive gap remains a challenge in translating UAV-based insights into timely, actionable agricultural decisions.\n\nRecent advances in UAV and mobile vision systems have further extended field-scale monitoring capabilities through deep learning-based crop detection, segmentation, counting, and growth-stage classification. demonstrated the application of YOLO-based frameworks for real-time capsicum detection, segmentation, growth-stage identification, and mobile deployment, highlighting the growing convergence between computer vision and precision horticultural management.\n\nRemote sensing observations from satellite and UAV platforms are rarely interpreted in their raw spectral form. Instead, they are systematically transformed into standardized biophysical and agroecological indicators that provide an interpretable description of crop condition, canopy functioning, and landscape-scale processes. Optical satellite sensors capture surface reflectance across visible, near-infrared, and shortwave infrared wavelengths, enabling the development of widely adopted vegetation indices. Metrics such as the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Soil-Adjusted Vegetation Index (SAVI), Normalized Difference Red Edge Index (NDRE), Photochemical Reflectance Index (PRI), and Normalized Difference Moisture Index (NDMI), are routinely used to characterize crop vigour, biomass accumulation, canopy development, and leaf water content throughout the growing season (). Beyond spectral greenness, structural and radiative parameters offer a more process-oriented perspective on crop growth. Leaf Area Index (LAI) characterises canopy structure by quantifying leaf surface area available for light interception, while the fraction of absorbed Photosynthetically Active Radiation (fAPAR) represents the efficiency with which vegetation captures incoming solar energy. These v","Frontiers in Remote Sensing 2026, Data Fusion and Assimilation section","2026-07-07T00:00:00Z",74,{"impact":92,"substance":91,"depth":92,"authority":93,"freshness":129,"relevant":52,"comment":166},"综述论文探讨LLM在GeoAI农业咨询中的应用，具有前瞻性，但时效性较低。",[168],{"name":162,"url":159},[170,171,172,21,173],"大语言模型","GeoAI","数字农情咨询","人在回路",[175,176],"数字农情咨询 大语言模型 人在回路 GeoAI","数字农情咨询 大语言模型","数字农情咨询大语言模型人在回路GeoAI-175","10.3389\u002Ffrsen.2026.1839369\u002Fabstract","2026-08-01T23:53:44.541424Z"]