山东省工业和信息化厅:关于组织开展2026年度重点行业语料库揭榜挂帅项目申报工作的通知

Municipal Industry and Information Technology Bureaus, Finance Bureaus:
为深入贯彻全省“人工智能+”创新应用工作会议精神,认真落实《山东省人民政府办公厅印发〈关于支持人工智能全产业链创新发展的若干政策措施〉的通知》要求,加快打造一批高质量、标准化、可流通的行业语料库,赋能一批行业大模型、工业智能体等应用落地,根据《山东省省级“语料券”奖补资金管理办法及实施细则》,现组织开展2026年度重点行业语料库揭榜挂帅项目申报工作。有关事项通知如下:
1、 Task content
Focusing on the basic theoretical research, product development and design, production management and operation, process quality inspection and other key links and specific scenarios of key industries in industrial manufacturing, a knowledge corpus is gathered. Based on structured data (such as table data in databases), unstructured data (such as text, images, speech, etc.) and semi-structured data (such as log files, XML files, etc.), through cleaning, denoising and unified format, it is used to support tasks such as natural language processing, computer vision, machine learning, deep learning, etc., to meet the high-quality corpus needs of industry or scene large model development, training and fine-tuning.
2、 Recommended conditions
(1) Declaration unit conditions
1. Enterprises with independent capacity to bear civil liability, normal operating status, no serious dishonesty record in the past three years, and meeting the requirements of the "green threshold" system for financial enterprise funds;
2. Possess basic capabilities such as core data, key technologies, and independent intellectual property related to the construction of industry corpora;
3. Possess the necessary software and hardware facilities as well as a professional team for corpus collection, organization, and annotation. Encourage project stakeholders to collaborate with universities or research institutions with advantageous disciplines to develop construction plans;
When purchasing or managing data corpus, it is not allowed to violate relevant policies and regulations such as national security, trade secrets, personal privacy, and technological ethics, and necessary security and compliance processing must be completed.
(2) Declaration project conditions
1. The project is an ongoing or planned project with the necessity, importance, and urgency to accelerate implementation. The corpus product can significantly improve the efficiency, accuracy, safety, and reliability of large model training through continuous optimization and high-quality expansion. The construction period generally does not exceed one year;
2. Significant innovation in technology or product performance, high specialization and applicability in specific fields or industries, clear market demand and strong market competitiveness;
3. The project should have clear application promotion plans and ecological construction expectations, including but not limited to new sales revenue, serving large model enterprise fields and quantities, as well as organizing supply and demand docking, communication and promotion activities;
4. The project should have a clear implementation plan and financial support to ensure that the construction can be completed according to the plan;
At the time of project acceptance, the industry related corpus data volume should not be less than 100000, with high data quality, domain coverage, potential value, and application effectiveness. The project acceptance should pass third-party evaluation;
6. Encourage corpus projects in various industries to accelerate the optimization and integration of corpus resources, and actively open up public corpora.
(三)重点行业
为加快实现对制造业行业门类全覆盖,本次申报重点围绕高端装备、烟草制品业、农副食品加工业、家具制造业、木材加工、皮革毛皮羽毛及其制品和制鞋业、仪器仪表制造业、废弃资源综合利用业等行业,重点推进行业关键数据技术攻关、行业数据语料标准研制、高质量行业语料库打造、语料应用场景落地等。
3、 Declaration Procedure
(一)企业申报。各市工业和信息化局会同市财政局按照企业自愿原则组织申报,省属及中央驻鲁单位按属地申报。申报企业按要求编制《山东省重点行业语料库揭榜挂帅项目申报书》(见附件1),申报材料须符合国家保密相关规定。
(二)初审推荐。各市工业和信息化局牵头对申报项目进行评审论证,于2月27日前将推荐文件、汇总表、申报材料纸质版(一式两份),以两部门名义正式行文报至省工业和信息化厅、省财政厅,电子版申报材料(盖章PDF版及Word版)同步发至指定邮箱。
(3) Review announcement. The Provincial Department of Industry and Information Technology organizes experts or third-party professional institutions to evaluate the declared projects, and comprehensively adopts various methods such as material review and on-site defense to determine the list of eligible projects for unveiling. After obtaining approval from the Provincial Department of Finance, the evaluation results are publicly announced.
(四)验收评估。省工业和信息化厅将进行动态监督,根据项目实施实际情况,于2026年下半年进行集中验收评估,并对项目实施期内的实际投入资金、建设应用成效等进行核查,公布验收评估结果。
4、 Regarding requirements
(1) The project application unit shall be responsible for the authenticity of the indicators, data, relevant certificates, and attachment materials involved in the application materials, and shall bear the consequences caused by providing false materials.
(2) The Provincial Department of Industry and Information Technology does not entrust any unit or individual to engage in project application agency services. The application form and data of the applying unit are only used for project evaluation and process management.
5、 Contact Information
省工业和信息化厅:杨洪于 0531-51782640,18888302292
省财政厅:马子腾 0531-51769593
邮 箱:cytjc@shandong.cn
地 址:济南市历下区省府前街1号省工信厅数据产业推进处(邮编250011)
Attachment: 1
山东省2026年度重点行业语料库揭榜挂帅项目申报书.docx
2.
Manufacturing Industry Catalogue. pdf
山东省工业和信息化厅 山东省财政厅
2026年2月3日
