Recently, the Jiangsu Provincial Data Bureau and eight other departments jointly issued the "Implementation Plan for the Construction of High Quality Data Sets in the Development of Data Labeling Industry in Jiangsu Province (2025-2027)", proposing to increase the high-quality and efficient supply of public data, strengthen the data labeling industry, and accelerate the construction of high-quality data sets. The Plan specifies that for the development of the service big model, it will focus on data enrichment areas within the province such as industrial manufacturing, transportation, financial services, healthcare, green and low-carbon, urban governance, as well as innovative areas such as low altitude economy, embodied intelligence, intelligent driving, smart ocean, and biomanufacturing. The main units of the production data usage data chain will be selected, high-quality datasets will be constructed, and supporting application guidelines will be formed. Publish a list of high-quality dataset construction and promote the opening, sharing, and circulation of high-quality datasets in related fields.
Full text attached
Implementation Plan for High Quality Dataset Construction of Jiangsu Province's Development Data Labeling Industry (2025-2027)
In order to implement the Implementation Opinions of the National Development and Reform Commission and other departments on promoting the high-quality development of the data annotation industry, as well as the requirements of relevant documents such as the National Data Administration on the construction of high-quality datasets, fully leverage the advantages of industry data enrichment and diverse application scenarios in our province, adopt measures tailored to local conditions, and systematically promote the development of the data annotation industry and the construction of high-quality datasets, seize the high ground of artificial intelligence development, and comprehensively empower high-quality economic and social development, the following implementation plan is formulated.
1、 Development Goals
By the end of 2027, the level of refinement, specialization, intelligence, and systematization of the data annotation industry in the province will significantly improve, and a nationally leading and globally influential data annotation industry cluster will be established, with a national proportion of over 10% and an average annual compound growth rate of over 20%. We will focus on building three data annotation bases, cultivating around 10 key data annotation enterprises with strong innovation leadership, element aggregation, and industry influence. We will construct 1000 complete, standardized, accurate, and practical high-quality datasets, select 100 replicable and promotable typical application cases, and form a more dynamic, open, and warm industry and innovation ecosystem, laying a solid data foundation for the high-level development of artificial intelligence.
2、 Strengthen the supply of data resources
（1） Increase the high-quality and efficient supply of public data
Deepen the three-year action plan to tackle public data and establish an integrated public data resource system throughout the province. Improve the management mechanism of public data inventory, release high-value public data inventory and electronic license issuance inventory annually, establish a unified data directory for the whole province and dynamically update data resources, and promote the management of public data "one account". Comprehensively promote the joint construction and sharing of provincial basic databases, accelerate the construction of thematic and special databases in key areas such as medical and health, social security, ecological environment protection, credit system, and safety production. Collaboration between government, enterprises, academia, research, and application is being carried out to construct large model datasets in the field of government affairs, promoting the inclusion of services such as data annotation and model training in the scope of government procurement. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（2） Accelerate the high-level development and utilization of enterprise data
Conduct a survey on enterprise data resources, improve the mechanism for the formation, protection, and distribution of enterprise data rights and interests. Encourage enterprises to accelerate the implementation of digital transformation driven by data elements, and strengthen data integration and convergence. Promote the construction of data resource system for state-owned enterprises and implement the "Action to Improve Data Efficiency of State owned Enterprises". Support industry leading enterprises and platform enterprises to take the lead in building industry trusted data spaces, and guide industry leading enterprises to open up data resources to upstream and downstream enterprises. Support various public service platforms and industry service platforms to establish enterprise data service zones, and increase support for data governance and application of small and medium-sized enterprises. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（3） Accelerate high-quality interaction of industry data
Focusing on industries such as agriculture and rural areas, industrial manufacturing, scientific research, education and teaching, healthcare, financial services, cultural tourism, etc., accelerate the formation of industry data products, industry data standards, and supply-demand docking lists. Encourage research institutions, industry associations, and other organizations to establish a common data resource library for the industry. Encourage industry leaders to use key technologies such as 3D spatial perception modeling, multimodal data fusion, and knowledge graph enhanced annotation to empower new scenarios and applications of data annotation. On the premise of implementing the requirements of the data classification and grading protection system and safeguarding the legitimate rights and interests of all parties, we encourage industry leaders, platform enterprises, and other companies to fully interact with data resources. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
3、 Strengthen the data annotation industry
（4） Strengthen the cultivation of labeled enterprises
Persist in promoting and nurturing, actively strive for the data business layout of platform enterprises and industry leading enterprises in Jiangsu, support data annotation technology enterprises and small and micro enterprises to root in Jiangsu and grow stronger. Build a provincial database of data annotation enterprises and select benchmark enterprises for provincial data annotation. By the end of 2027, the province will cultivate 10 key data annotation enterprises with strong innovation leadership, strong factor aggregation, and strong industry influence. (Led by the Provincial Data Bureau, the Provincial Development and Reform Commission, Provincial Department of Science and Technology, Provincial Department of Industry and Information Technology, and Provincial State owned Assets Supervision and Administration Commission are responsible for their respective duties)
（5） Carry out core technology research and development
Encourage technology innovation oriented enterprises, research institutes, etc. to strengthen core technologies such as cross domain and cross modal semantic alignment, 4D annotation, and data synthesis. Accelerate the development and application of intelligent tools such as multimodal annotation, annotation review, quality assessment, expert annotation, and thought chain annotation. Promote the development of intellectual property work in the data annotation industry, accelerate the transformation and application of scientific and technological achievements. (Led by the Provincial Department of Science and Technology, with the Provincial Data Bureau and the Provincial Intellectual Property Bureau responsible for their respective duties)
（6） Establish sound data labeling standards
Encourage the Provincial Data Standardization Technical Committee to focus on the key aspects of data annotation, and to undertake the drafting and technical review of relevant standards based on the specific needs of multimodal data annotation such as text, image, video, and voice, in order to promote the construction of a hierarchical classification data annotation standard system. Support leading enterprises, universities, research institutes, and other institutions in the province to participate in the development of international standards, national standards, industry standards, local standards, group standards, and enterprise standards. Relying on the Provincial Data Standardization Technical Committee and other standardization technical committees to promote and train on data annotation standards, guiding enterprises to achieve standard implementation in annotation tasks, annotation environments, process control, quality assurance, management mechanisms, and other aspects. (Provincial Data Bureau and Provincial Market Supervision Bureau are responsible according to their respective duties)
（7） Promote the development of industrial clusters
Develop the data annotation industry according to local conditions, support regions with conditions to build data annotation bases based on regional industrial advantages, promote the "high-end and specialized" development of the southern Jiangsu region, and enhance the "platformization and automation" of the central and northern Jiangsu regions, achieving gradient connection and industrial synergy within the province. Cooperate with industry leading enterprises, large model enterprises, Internet platform enterprises, scientific research institutes and other entities to draw the provincial data annotation industry map, forming an integrated development ecology composed of data resource units, data annotation enterprises, and data set application institutions. Encourage regions with conditions to formulate supporting industrial policies. Encourage base operating agencies to provide entrepreneurial guidance, personnel training, policy consultation, supply and demand matching and other services for settled enterprises. (The Provincial Development and Reform Commission, Provincial Department of Industry and Information Technology, and Provincial Data Bureau are responsible for their respective roles and responsibilities.)
4、 Accelerate the construction of high-quality datasets
（8） Development of Service Model
Targeting data enrichment fields within the province such as industrial manufacturing, transportation, financial services, healthcare, green and low-carbon, urban governance, as well as innovative fields such as low altitude economy, embodied intelligence, intelligent driving, smart ocean, and biomanufacturing, we will select the main units of the production and usage data chain, establish high-quality datasets, and form supporting application guidelines. Publish a list of high-quality dataset construction, promote the opening, sharing, and circulation of high-quality datasets in related fields, empower large-scale models in industries such as transportation, healthcare, and education, and train and apply large-scale models in industries such as industry, agriculture, and government. Support large model enterprises, data service providers, research institutes and other entities to use high-quality datasets to carry out industrial applications of natural language processing, multimodal interaction, knowledge graph, embodied intelligence training and optimization of large models, and assist in the innovation of artificial intelligence large model technology, scene application and industrial ecological prosperity. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（9） Support multi-party construction
Encourage leading enterprises and large state-owned enterprises in the industry to build high-quality datasets for key industries in batches, guide the development of data annotation bases and artificial intelligence bases in the province with regional advantages, support science and technology innovation enterprises and research institutes to build high-quality datasets for segmented scenarios and research directions, covering general knowledge, industry general knowledge, industry specialized knowledge and other types. Publish a batch of high-quality dataset construction unveiling tasks every year, and encourage regions with conditions to provide support to enterprises, institutions, and organizations that build high-quality datasets. By the end of 2027, the province will have established no less than 1000 complete, standardized, accurate, and practical high-quality datasets. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（10） Explore applications in multiple fields
In response to major national strategic needs, we will explore a "guarantee model" for dataset supply and promote the free and conditional opening of high-quality datasets to artificial intelligence industry application bases. In response to the needs of social welfare and technological innovation, explore a "customized model" for dataset supply, and open it according to demand orientation. In response to the demand of key industries with high data maturity and good market-oriented foundation, explore the "e-commerce model" of dataset supply, and promote the paid and conditional opening of high-quality datasets. In response to the needs of key industries with good data resource endowments but requiring long-term development, we will explore the "paired model" of dataset supply, guide industry enterprises, large model enterprises, and digital solution providers to form a virtuous cycle, and jointly promote the open use and iterative updating of high-quality datasets. Explore the establishment of an authorization and sharing mechanism for industry general knowledge datasets. Taking into account the differences in marketization levels of different types of data, guide the construction and operation entities of the dataset to explore on-demand and on-demand data supply methods for small and medium-sized enterprises, reduce the threshold for small and medium-sized enterprise data acquisition, and strengthen the supply of data resources for small and medium-sized enterprises. Establish a sound dataset price monitoring system, and operating entities shall not engage in monopolistic or unfair competition practices. Widely solicit and regularly publish application cases of key domain datasets from the society. By the end of 2027, 100 replicable and promotable typical application cases will be selected throughout the province. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（11） Improve integrated governance
Based on the regional functional nodes of data infrastructure, establish a provincial-level data annotation and high-quality dataset public service zone, and provide services such as policy release, standard specifications, open source resources, technical guidance, and cooperation and exchange. Building a smart annotation platform and an open-source community for data annotation industry has been included in the construction task of Jiangsu Digital Economy Innovation and Development Pilot Zone. Build a high-quality dataset project reserve library, and establish a three-level project reserve system consisting of basic library, characteristic library, and strategic library. Establish a dynamic evaluation and rating system for high-quality datasets, develop a high-quality dataset quality evaluation method and tool set for segmented industries, conduct regular reviews and dynamic management of project datasets based on standardized evaluation tools, and achieve quality control and efficiency improvement throughout the entire lifecycle of high-quality datasets. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（12） Guide the circulation of a platform
Guide and incentivize both the supply and demand sides of the dataset to register and trade on the province wide integrated data trading platform, encourage the trading platform to waive registration fees and provide subsidies based on transaction volume for enterprises that register for the first time and obtain platform certification, and provide filing and other guarantee services for datasets traded outside the platform. Encourage cross provincial transactions of datasets and promote cross regional cooperation. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
5、 Organize and implement
（13） Strengthen top-level coordination efforts
Give full play to the coordinating mechanism of the provincial digital economy development department, and focus on key industry areas according to their respective responsibilities. Collaborate to promote the cultivation of cross regional and cross departmental data annotation industries and the demonstration construction of high-quality datasets. Guide the establishment of a provincial-level data annotation industry alliance. Strengthen the linkage between provinces and cities, and encourage eligible prefecture level cities to introduce supporting implementation rules. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
（14） Strengthen the foundational support capability
Accelerate the construction of data infrastructure throughout the province, consolidate the basic operation foundation of data annotation public services, high-quality dataset circulation and trading, etc. Accelerate the construction and deployment of an integrated computing power dispatch platform throughout the province, optimize the layout of computing power infrastructure, improve the efficiency of computing power resource allocation, and reduce the cost of computing power usage for data annotation enterprises. Strengthen data security protection, implement data classification and grading protection system, and improve data security management rules throughout the entire process. Promote the application of data security technologies such as blockchain and privacy computing. (The Cyberspace Administration of the Provincial Party Committee, the Provincial Development and Reform Commission, the Provincial Department of Industry and Information Technology, the Provincial Department of Finance, and the Provincial Data Bureau are responsible for their respective duties.)
（15） Increase fiscal and financial support
Coordinate relevant financial funds to support data annotation industry projects that comply with regulations. Give full play to the guiding role of the provincial strategic emerging industry mother fund, strengthen the functions of scientific and technological innovation investment, industrial investment, and merger and acquisition investment in the data field, and amplify the leverage effect of state-owned capital on social capital investment in the data annotation industry. Improve the due diligence exemption, fault tolerance and error correction mechanism for fund investment through hierarchical classification, establish and improve a differentiated and refined comprehensive performance evaluation system, and scientifically set the industry fund loss tolerance rate. Encourage financial institutions to increase credit support for data annotation industry related enterprises in accordance with market-oriented principles, and encourage eligible data annotation industry related enterprises to enter the capital market for financing in accordance with the law. (The Provincial Development and Reform Commission, Provincial Department of Finance, Provincial State owned Assets Supervision and Administration Commission, and Provincial Data Bureau are responsible for their respective roles and responsibilities.)
（16） Building a talent cultivation system
Increase support for artificial intelligence talents in provincial-level talent projects such as the "Double Innovation Plan" and the high-level talent training plan. Establish a talent pool for the data annotation industry and conduct professional and technical title evaluations for digital economy professionals in the fields of artificial intelligence and big data. Encourage universities, vocational colleges, and technical schools to offer majors and courses related to data annotation, strengthen cooperation between schools and enterprises, and support the construction of industry education integration training bases. (Led by the Provincial Data Bureau, the Organization Department of the Provincial Party Committee, the Provincial Department of Education, and the Provincial Department of Human Resources and Social Security are responsible for their respective duties.)
（17） Deepen regional exchanges and cooperation
Deepen technical cooperation with key regions such as the European Union and ASEAN, support data annotation enterprises to undertake international business, carry out pilot construction of cross-border data services, explore cross-border data applications such as product digital passports and overseas digital bills of lading, improve cross-border data flow security supervision mechanisms, and strengthen data export classification management. Relying on organizations such as the Data Labeling Industry Alliance, we will carry out hierarchical connections between the data labeling industry in central, southern, and northern Jiangsu regions, prioritize promoting business circulation within the region, and strengthen the promotion of data labeling innovation achievements and cross regional exchange and cooperation on this basis. (Led by the Provincial Data Bureau, relevant departments are responsible according to their respective duties)
