• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.417-419
• /
• 2020

데이터 3법 시대에 접어들면서 기업들에는 가명화된 개인정보를 활용할 수 있는 길이 열렸다. 하지만 현 데이터 3법은 데이터를 생성하고 유통하며 활용할 기업들의 책임과 혜택에 내용이 맞춰져 있어 아쉬운 감이 있다. 개인의 기본권을 보장하면서도 마이데이터 유통 및 활용을 도울 방법은 없을까? 본 논문에서는 데이터의 주체인 개인이 데이터 주권을 행사하고 실질적인 혜택을 받는 마이데이터 서비스의 활성화를 위한 ID 관리 기술로 블록체인 기반 분산 ID(Decentralized Identification, DID)를 제안하고, DID 수용의도에 영향을 마치는 요인을 연구함으로써 마이데이터 서비스 개발 활성화를 위한 정책적, 실무적 시사점을 도출하고자 한다.

• Journal of the Korean Society of Radiology
• /
• v.83 no.4
• /
• pp.759-770
• /
• 2022

The importance of ethics in research and the use of artificial intelligence (AI) is increasingly recognized not only in the field of healthcare but throughout society. This article intends to provide domestic readers with practical points regarding the ethical issues of using radiological images for AI research, focusing on data security and privacy protection and the right to data. Therefore, this article refers to related domestic laws and government policies. Data security and privacy protection is a key ethical principle for AI, in which proper de-identification of data is crucial. Sharing healthcare data to develop AI in a way that minimizes business interests is another ethical point to be highlighted. The need for data sharing makes the data security and privacy protection even more important as data sharing increases the risk of data breach.

• Informatization Policy
• /
• v.28 no.3
• /
• pp.49-72
• /
• 2021

This study analyzes the major content, significances, and future outlook of Three Data Acts amendment enacted in August 2020 in South Korea, with the focus on their impact on the financial and data industries. It seems that the revision of the Credit Information Act will enable the specification of a business which had previously only been regulated as the business of credit inquiry, and also enable the domestic data industry to activate the MyData industry, data trading and platforms, and specify data pseudonymization and trading procedures. For the rational and efficient implementation of the amendments to the Three Data Acts, the Personal Information Protection Committee must be as transparent and lawful in its activities as possible, and fairness must be guaranteed. Even in the utilization of personal information, the development or complementation of the related data processing technologies is essential, and clear data processing methods and areas must be regulated. Furthermore, the amendments must be supported with guarantees and the systematization of a fair competitive system in the data market, stricter regulations on penalties for illegal acts related to data, establishment and strengthening of the related security systems, and reinforcement of the system of cooperation for data transfer.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.47-49
• /
• 2021

According to the revision of the Data 3 Act in 2020, personal information of medical data can be processed anonymously for statistical purposes, research, and public interest record keeping. However, unidentified data can be re-identified using genetic information, credit information, etc., and personal health information can be abused as sensitive information. In this paper, we derive the need for homomorphic encryption to protect the privacy of personal information separated by sensitive information.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.04a
• /
• pp.633-635
• /
• 2016

Convolutional Networks (ConvNets) are powerful models that learn hierarchies of visual features, which could also be used to obtain image representations for transfer learning. The basic pipeline for transfer learning is to first train a ConvNet on a large dataset (source task) and then use feed-forward units activation of the trained ConvNet as image representation for smaller datasets (target task). Our key contribution is to demonstrate superior performance of multiple ConvNet layer features over single ConvNet layer features. Combining multiple ConvNet layer features will result in more complex feature space with some features being repetitive. This requires some form of feature selection. We use AdaBoost with single stumps to implicitly select only distinct features that are useful towards classification from concatenated ConvNet features. Experimental results show that using multiple ConvNet layer activation features instead of single ConvNet layer features consistently will produce superior performance. Improvements becomes significant as we increase the distance between source task and the target task.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.6 no.7
• /
• pp.9-16
• /
• 2016

As life in a rapidly changing Internet age at home and abroad, large amounts of information are being used medical, financial, services, etc. Accordingly, especially hospitals, is an invasion of privacy caused by leakage and intrusion of personal information in the system in medical institutions, including clinics institutions. To protect the privacy & information protection of personal health medical information in medical institutions at home and abroad presented by national policies and de-identification processing technology standards in accordance with the legislation. By comparative analysis in existing domestic and foreign institutional privacy and de-identification technique, derive a advanced one of pseudonymization and anonymization techniques for destination data items that fell short in comparison to the domestic laws and regulations, etc. De-identification processing technology for personal health information is compared to a foreign country pharmaceutical situations. We propose a new de-identification techniques by reducing the risk of re-identification processing to enable the secondary use of domestic medical privacy.

• Informatization Policy
• /
• v.30 no.2
• /
• pp.46-67
• /
• 2023

There are criticisms that, despite the proactive government policy on open government data (hereinafter "open data"), certain highly demanded data remains restricted due to legal constraints. In this study, we aim to analyze the factors that limit the opening and utilization of open data, focusing on cases wherein requests for open data provision have been denied. We will explore possible approaches that are in harmony with the Open Data Law while examining the constitutional value of open data, considering the foundational Open Data Charter that underpins the government's data policy. We will also examine cases wherein requests for data provision have been denied for institutional reasons, with nearly half of these cases involving open data that includes personal information. It is necessary to explore the potential for improvement in these cases. Furthermore, considering the recent amendment to the Personal Information Protection Act, which allows for the processing of pseudonymous information without the consent of the data subject for limited purposes, it is an opportune time to consider the need for amending the Open Data Law to facilitate broader access and utilization of open data for the nation. Lastly, we will propose institutional improvement directions aligned with the opening and utilization of open data by examining the constraints of and need for improvement in the selected target laws.

• Informatization Policy
• /
• v.26 no.1
• /
• pp.25-45
• /
• 2019

With the advent of the data economy, interest in using big data has increased, but conflicts with protecting personal information have been also steadily raised. In this regard, major countries are accelerating use of big data by exempting de-identified, pseudonymous personal information from protection. However, these policies have been made without the understanding that the economic value of personal information has been actually changing slowly. This paper presents the concept of 'collected information' and defines it as having public interest and therefore, not the exclusive property of the collector of such information. The paper shows the collected information has public interest in terms of personal information protection, connectivity, and universal service and public goods. It also specifies that the 'data governance' cannot be applied to the current data utilization framework that depends upon the holder's consent; rather, it raises the need to improve the practices of information provision consent or provide the beneficiary right of information use to the information holder in order to ensure the proper 'data governance' that will turn market failure into success.

• The Journal of the Korea Contents Association
• /
• v.21 no.10
• /
• pp.1-10
• /
• 2021

Log data accumulated in the Learning Management System (LMS) provide high-quality information for the learning process of students. Until now, various studies have been conducted to predict students' academic achievement using LMS log data. However, previous studies were based on relatively small sample sizes of students and courses, limiting the possibility of generalization. This study developed and validated a deep neural network model for the early prediction of academic achievement of college students using massive LMS log data. To this end, we used 78,466,385 cases of LMS log data and 165,846 cases of grade data. The proposed model predicted the excellent-grade students with a high level of accuracy from the beginning of the semester. Meanwhile, the prediction accuracy for the moderate and underachieving groups was relatively low, but the accuracy improved as the time points of the prediction were delayed. This study is meaningful in that we developed an early prediction model based on a deep neural network with sufficient accuracy for practical utilization by only using LMS log data.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.5
• /
• pp.915-932
• /
• 2022

According to the Personal Information Protection Act and Pseudonymization Guidelines, the mapping is processed to the hash value of the combination key generation items including Salt value when different combination applicants wish to combine. Example of combination key generation items may include personal information like name, phone number, date of birth, address, and so on. Also, due to the properties of the hash functions, when different applicants store their items in exactly the same form, the combination can proceed without any problems. However, this method is vulnerable to combination in scenarios such as address changing and renaming, which occur due to different database update times of combination applicants. Therefore, we propose a privacy preserving combination key generation scheme robust to updates of items used to generate combination key even in scenarios such as address changing and renaming, based on the thresholds through probabilistic record linkage, and it can contribute to the development of domestic Big Data and Artificial Intelligence business.