Cryptography, Evaluation Modelling, Blockchain with its applications, Mathematical Aspects of Artificial Intelligence
Professor Chul Kim is a member of our mathematics faculty. Prior to joining us, he had full-time teaching positions at Shaw University, University of South Dakota, Kwangwoon University, Gwangju Institute of Science and Technology, Daegu Gyoungbuk Institute of Science and Technology, and Ryerson University. He was also a visiting scholar at the University of British Columbia, Pyongyang University of Science and Technology, and Tianjin University. While at Ryerson University, he was the recipient of the Teaching Excellence Award in both 2005 and 2015.
Prof. Kim’s research interests span both applied mathematics and information technology security. His book, “Introduction of Cryptography”, was published in Korea in 1996. In recent years, he focused on evaluation modelling on security mechanisms. He was also interested in methodologies of evaluation for university admissions. His book, “Introduction to Holistic Evaluation for University Admissions,” published in 2011, was also the first such specialized book in Korea.
Life-time summary
Ph.D. in Applied Mathematics, North Carolina State University (1989)
M.S. in Applied Mathematics, North Carolina State University (1987)
B.S. in Mathematics, Yonsei University (1984)
We propose a mathematical modelling for dynamic signature verification which is one of the key behavioural authentications. In the signature verification area, many papers have been published focusing on feature-based verification and its efficiency with validation only. Those papers have demonstrated various feature selecting methodology and decision process. That is, the results have shown an embedding invariance, feature integrity for verification and convergence of different feature parameters. In this paper, we propose a formal description of feature selection with 3 depths; template itself, segments level and sub-segment level. The second level could be obtained based on a discontinuity of input signature, if there is one; otherwise the segment would be separated by position parameters based on temporal or spatial separations. The third level depth, the sub-segment level could be selected by sub-position parameters. By calculating a rectangular area based on its position and sub-position parameters, the model will recognize the specific two-dimensional segments. And by calculating rectangular volume based on its sub-position parameters with average pressure value, the model will be decided the specific three-dimensional sub-segments. We find that most of the published works fit in the proposed model. For example, when the pressure values set as 1, the model indicates a simple feature captioning verification system. By defining the overall integrity of the model, based on the computations of correlation between those model’s variables, we could get validation of the model.