Establishing Data Quality Concepts, Scope, and Detailed Requirements
On the 5th, the Ministry of Science and ICT announced that it will pursue international standardization of AI data quality standards, a core resource of the Digital New Deal 'Data Dam'.

In this regard, it will develop a standard that establishes the concept and scope of quality and detailed requirements, and on the 6th will officially propose it to the group standardization organization (TTA PG 1005, AI-based Technology) within the Korea Information and Communication Technology Association, with plans to make it a TTA group standard.

Subsequently, it is expected to undergo a process of collecting opinions from related experts and achieve final adoption and confirmation in June of next year. Additionally, there are plans to proceed with related content toward international standardization achievements.

This standard was developed by supplementing and concretizing the previously developed 'Common Criteria for Building and Quality Management of AI Training Data ('19)' through the Korea Information and Communication Technology Association and the National Information Society Agency, based on analysis of overseas cases and review of AI and quality management.

Therefore, the quality of AI data focuses on user-oriented useful value, taking into account the diversity, accuracy, and validity of AI technology (models and algorithms) data.

However, even renowned open datasets built and updated over 4-6 years by global information and communication technology companies such as Microsoft and Google show data accuracy levels of only 43-83%, demonstrating that a systematic methodology for AI data quality has not yet been established globally.

Accordingly, the standard is being developed in the form of a universal standard applicable across diverse fields including natural language processing, autonomous vehicles, healthcare, agriculture and fisheries, and manufacturing.

The Ministry of Science and ICT announced that it is already applying the standard developed this time to projects in ten priority sectors of supplementary budget, and will rapidly advance standardization through collecting opinions from industry, academia, and research experts in the group standardization organization (TTA PG 1005, AI-based Technology) within the Korea Information and Communication Technology Association.

Furthermore, there are plans to propose related matters to the 'AI International Standardization Conference (ISO/IEC JTC1/SC42)' to achieve outcomes in international standardization.