A Study on the Importance of Data Balance for Digital Transformation: Based on the result of data balance analysis for AI learning
Yeungjoo Kwon, Ji Hwan Park, Sunil Chon, Sung-Taek Lee · National Association of Korean Local Government Studies · 2022
This study shows the need of data balance by analysis the balance of data collected by the government. As a result of analysis of vehicle recognition data, only 231(3.08%) were significant data, even though 7,500 pieces of diversity were required, and the rest were all listed in the overlapping data. This means that biased artificial intelligence is highly likely to be created. The introduction of artificial intelligence technology is too important and fundamental a change, it cannot be achieved only by the technical efforts of individual companies in quality. Only when the government or public institutions provide necessary standards and provide minimum infrastructure can the competitiveness of companies as well as social stability related to new technologies be secured.