Disclosure of progress on AI training data construction project
Available through the NIA website and AI Hub
The 2021 business candidate projects are scheduled to be announced in January next year.


On the 31st, the National Intelligence Service (NIS) disclosed to the public the progress of the mid- to long-term roadmap for each major field and the major fields of the AI learning data construction project being promoted by the Ministry of Science and ICT.

The AI Learning Data Construction Project aims to revitalize the AI ecosystem and create jobs by building and opening large-scale learning data required in the AI technology development process. A total of approximately 1,300 types of data are expected to be built by 2025.

Through the progress of the mid- to long-term roadmap for each major sector being released this time, companies and organizations interested in the project can check the basic direction of the project, the implementation system, and the implementation schedule, which reflect national strategy, the role and importance of the project, and technological trends.

Additionally, you can check the core areas (① voice/natural language, ② vision, ③ healthcare, ④ transportation, ⑤ disaster/safety/environment, ⑥ agriculture, livestock and fisheries) and strategic areas (all areas excluding core areas such as manufacturing, finance, education, and cultural content).

NIA will disclose the information through its website and AI Hub and receive various opinions from the private and public sectors. We plan to incorporate these proposals into the roadmap (airoadmap@nia.or.kr). We plan to announce the 2021 project candidates in January of next year and finalize the final project through online feedback gathering.

Meanwhile, approximately 100 domestic AI experts, including AI professors, researchers, public institutions, companies utilizing the technology in various fields, and doctors, are participating in the project.