“Broader industry use, including small enterprises and individuals on the ‘AI sharing platform’ means less need for personal information and improved privacy protection. Through AI sharing, businesses will no longer need to share or exchange sensitive customer information or even integrate personal data.”
The ‘AI Seoul Forum 2022’ was held on August 29, 2 pm, at the COEX, Seoul, Korea. Dr. Kyoung Jun Lee, Professor at Kyung Hee University (School of Business and Department of Big Data Analytics), delivered the keynote presentation, emphasizing that “the medical field has been the most active field in AI sharing.” Dr. Lee introduced “the Nature Medicine paper published in October 2021, on EXAM (EMR CXR AI Model) federated learning, ‘Federated learning for predicting clinical outcomes in patients with COVID-19’.”
Dr. Kyoung Jun Lee pointed out, “researchers found that using data from 20 institutions to train EXAM resulted in superior AI-based predictions over locally trained models. The greatest benefits were found among smaller hospitals with relatively smaller datasets, where the FL model markedly outperformed the local models. Notable benefits were also found in the largest hospital with the largest dataset.” He explained that digitalization and artificial intelligence could lead to polarization, but model sharing offered a means to bridge that divide.
During the forum, Dr. Lee unveiled a case of AI sharing in the industry, the “User Centric Sharing Platform, developed with Harex InfoTech that will benefit customers and businesses, heralding a new order in the shared economy based on the AI Sharing Platform.”
Dr. Lee states, “Even if it is not a huge platform, small businesses of franchises and traditional markets participating in the AI sharing platform can grow while enjoying the benefits of attracting new customers and promoting sales in various ways such as repurchase, upsell, and cross-selling through the User Centric AI (UCAI) based recommendation service. The professor expected that “the use of the model sharing platform will allow SMEs to compete with volume-based data network benefits previously exclusive to giant-scale platforms.”
He also introduced B2B services driven by hyper-personalized recommendation systems, explaining with the example, “if the price of Company B’s fried chicken is based on the purchase history of Consumer A, then Company B can target-market its fried chicken to Consumer A. The previous method of targeted marketing was based on group data and demographic information, but this system empowers hyper-personalized targeted marketing.
This system has already recommended a menu item for a major Korean franchise, which delivered that menu within a couple of months. In the context of product development, the system can produce a product (name) that does not yet exist, and that knowledge can inform the product development process.”
Dr. Lee chaired the forum's discussion panel. Panel member Dr. Seong Lyun Kim, Professor at Yonsei University, forecasted that “AI sharing is a very reasonable solution to polarization. It has foreseeable application, particularly in the manufacturing sector, handling data from multiple manufacturing facilities.” He added, “Korean industries already have an advanced understanding of this technology, and it should be possible for Korea to take the initiative and develop an advanced model to take the lead.”
Dr. Jung Hye Lee, Professor at the Ulsan National Institute of Science and Technology, pointed out that “autonomous driving would particularly benefit from AI sharing, at least to earlier effect in the short term, because the healthcare sector faces challenges in data being tied to not only the individual, but the medical service provider such as hospitals, doctors, and physicians, which will need to be navigated for the sharing model to be implemented.”
She affirmed that “the Korean industry competence for this technology does stack up against the U.S., and if we take the initiative on the technology to develop it to a world-class standard, it can secure a leading position in the field of AI.”
Dr. Kyoung Jun Lee raised the importance of policy measures, stating the need to “establish promotion structures and stimulation initiatives to advance AI sharing.” He suggested that “the AI sharing model may fundamentally address the data scarcity issue and provide the means to the world’s most capable AI.”
The Seoul AI Forum 2022 agenda included: △Prof. Seong Lyun Kim of Yonsei University demonstrated techniques for Federation and Split in Wireless AI: 5G Smart Factory Use Case Perspective △Prof. Jung Hye Lee of Ulsan Institute of Science and Technology presented on Smart City and Federated Learning △Prof. Hyeong Joo Kim presented on Kant and Artificial Intelligence, where he deliberated on ethics with relevance to AI.
Seoul Digital Foundation President Yo Sik Kang concluded the “meaningful gathering of top leaders in AI to forecast Seoul’s bright future in AI service and digital trends” and with the benediction for “the Seoul AI Forum to contribute to AI as a core technology in the digital transformation toward Smart Seoul shortly.”
The Korean Academic Society of Business Administration President Sang Man Han offered congratulations, emphasizing that “the vision of Seoul Mayor Se Hoon Oh, ‘Going Together with the Socially Neglected,’ can be visualized through ‘Smart Seoul, AI Sharing,’ the theme of today’s forum.” He continued, “I sincerely hope today’s forum will pave the way towards transforming Seoul into the most competitive city in the world, as well as strengthening the future competitiveness of the Republic of Korea.”
By Staff reporter Joong Kwon Lim(lim9181@etnews.com)
Original Source: Korean version of the ETNEWS on August 29, 2022
