• 한국산업융합학회 논문집
• /
• 제25권5호
• /
• pp.811-816
• /
• 2022

Even though continuous management and supervision of reinforcement of policies to safeguard accidents at workplace and work sites were implemented. Accident prevention activities such as inspection and diagnosis are urgently required to induce a preliminary investigation to identify the risk factors for each type of work, before the work task to eliminate risks at the worksites. Since safety inspections at work sites were generally conducted through visual inspections, the results of safety inspections may vary depending on the findings and proficiency of the safety officers. The results of those inspections may have loopholes to prevent potential accidents at work. Therefore, the purpose of this study was to develop a risk identification checklist that can effectively perform safety inspections to prevent accidents at work sites. This study initially analyzed the previously developed accident checklist to identify current complications and issues in safety checklists. Based on the findings of major industrial accidents over the past three years, the relationship between accident, workplace, and work type were analyzed refereeing the safety inspection standards. A risk recognition-checklist was developed to provide basic data on identifying risk factors, and inspection guidance at work sites. To prepare for potential accidents by identifying and taking countermeasures to mitigate the high risk and serious accidents at sites by the guidelines of the checklist. The developed inspection checklist has been practically used by experts at work sites to perform safety inspections, and it has been verified its suitability, and feasibility, to prevent or mitigate workplace accidents, including securing the safety and health of field workers. The role of the developed safety checklist has been considered effective at worksites.

• 정보보호학회논문지
• /
• 제32권2호
• /
• pp.329-339
• /
• 2022

최근 개인정보가 데이터로 활용되면서 다양한 신산업 등이 발굴되고 있지만 체계적인 관리체계 구축 미흡등 으로 개인정보 유출 및 오남용 사례가 연이어 발생되고 있다. 또한 지난 '20년 8월, 데이터 3법 시행 이후 개인정보를 가명·익명 처리하여 활용하는 서비스가 등장하고 있지만 불충분한 가명처리 및 가명정보 처리에 대한 안전성확보조치, 혐오표현 등의 민감정보의 처리 미흡으로 개인정보 이슈가 발생하고 있다. 이에 본 연구는 개인정보를 안전하게 활용하기 위하여 캐나다의 Ann Cavoukian[1]이 제시한 개인정보보호 중심 설계(Privacy by Design, 이하PbD) 원칙을 기반으로 가명정보 생명주기에 적용할 수 있는 새로운 PbD원칙을 제안하였다. 또한, 제안한 방법에 대하여 개인정보보호 관련 전문가 30명을 대상으로 설문조사를 통하여 제안 방법의 유의미함을 확인할 수 있었다.

• 한국IT서비스학회지
• /
• 제21권3호
• /
• pp.27-42
• /
• 2022

Open Data is an essential resource for the data industry. 'Act On Promotion Of The Provision And Use Of Public Data', enacted on July 30, 2013, mandates public institutions to manage the quality of Open Data and provide it to the public. Via such a legislation, the legal basis for the public to Open Data is prepared. Furthermore, public institutions are prohibited from developing and providing open data services that are duplicated or similar to those of the private sector, and private start-ups using open data are supported. However, as the demand for Open Data gradually increases, the cases of refusal to provide or interruption of Open Data held by public institutions are also increasing. Accordingly, the 'Open Data Mediation Committee' is established and operated so that the right to use data can be rescued through a simple dispute mediation procedure rather than complicated administrative litigation. The main issues dealt with in dispute settlement so far are usually the rights of third parties, such as open data including personal information, private information such as trade secrets, and copyrights. Plus, non-open data cannot be provided without the consent of the information subject. Rather than processing non-open data into open data through de-identification processing, positive results can be expected if consent is provided through active rights processing of the personal information subject. Not only can the Public Mydata Service be used by the information subject, but Open Data applicants will also be able to secure higher quality Open Data, which will have a positive impact on fostering the private data industry. This study derives a plan to establish a rights processing platform to enhance the usability of Open Data, including private information such as personal information, trade secrets, and copyright, which have become an issue when providing Open Data since 2014. With that, the proposals in this study are expected to serve as a stepping stone to revitalize private start-ups through the use of wide Open Data and improve public convenience through Public MyData services of information subjects.

• 한국컴퓨터정보학회논문지
• /
• 제28권9호
• /
• pp.81-92
• /
• 2023

프라이버시 모델이란 프라이버시 공격을 통한 개인정보의 유출 가능성과 위험 정도를 정량적으로 제한하는 기법이다. 대표적인 모델로 k-익명성, l-다양성, t-근접성, 차분 프라이버시 등이 있다. 지금까지 많은 프라이버시 모델들이 연구되어 왔지만, 주어진 데이터에 대해 가장 적합한 모델을 선택하는 문제에 대한 연구는 미흡하다. 본 연구에서는 개인정보 유출 문제를 막기 위한 최적의 프라이버시 모델 추천 시스템을 개발한다. 본 논문에서는 프라이버시 모델 선택 시 고려해야 할 데이터 특성(예: 데이터 타입, 분포, 빈도, 범위 등)을 분석하고 데이터 특성과 모델 간의 연관관계정보를 포함하는 프라이버시 모델 배경지식에 기반한 최적 모델을 추천한다. 마지막으로 타당성과 유용성을 검증하기 위해 추천 프로토타입 시스템을 구현하였다.

• Animal Bioscience
• /
• 제36권5호
• /
• pp.704-709
• /
• 2023

Objective: In tropical, subtropical and arid zones, heat stress is the main cause of productivity reduction in cattle. When climate stressors occur, animals become thermal adapted through differential expression of some genes, including heat shock proteins (HSP) family. The aim of this study was to determine levels of expression of HSP60, HSP70, and HSP90 genes in Simmental cattle raised in tropical environments of Mexico. Methods: In this study, expression of HSP60, HSP70, and HSP90 genes was analyzed in 116 Simmental cattle from three farms with tropical climate located in western Mexico. Animals were sampled twice a day, in the morning and noon. Gene expression was evaluated by quantitative polymerase chain reaction using probes marked with fluorescence. The MIXED procedure of SAS with repeated measures was used for all statistical analysis. Results: HSP60 gene expression differences were found for sex (p = 0.0349). HSP70 gene differences were detected for sampling hour (p = 0.0042), farm (p<0.0001), sex (p = 0.0476), and the interaction sampling hour×farm (p = 0.0002). Gene expression differences for HSP90 were observed for farm (p<0.0001) and year (p = 0.0521). HSP70 gene showed to be a better marker of heat stress than HSP60 and HSP90 genes. Conclusion: Expression of HSP70 gene in Simmental herds of the tropical region of western México was different during early morning and noon, but the expression of the HSP60 and HSP90 genes was similar. Identification of resilient animals to heat stress will be useful in the genetic improvement of the Simmental breed.

• 예술인문사회 융합 멀티미디어 논문지
• /
• 제7권2호
• /
• pp.311-319
• /
• 2017

최근 대용량 의료정보의 이차적인 활용에 대한 관심과 함께 의료정보 내의 개인정보에 대한 프라이버시 침해 문제에 대한 관심 또한 대두되고 있다. 대용량 의료정보의 경우 질병 연구, 예방 등 다양한 분야에서 활용될 수 있는 매우 유용한 정보이다. 이러한 대용량 의료정보의 경우 환자, 의료인 등에 대한 개인정보를 포함하고 있기 때문에 개인정보보호법과 같은 프라이버시 관련 법률에 저촉되어 활용에 많은 제한이 존재한다. 현재까지 k-익명성, l-다양성, 디퍼렌셜 프라이버시 등 의료정보 내의 개인정보를 보호하면서 대용량 의료정보의 이차적인 활용을 가능하게 하는 다양한 방법들이 개발되어 활용되어오고 있다. 본 논문에서는 지금까지 개발된 다양한 방법들 중 디퍼렌셜 프라이버시의 처리 절차에 대해 알아보고 라플라스 노이즈를 사용하는 디퍼렌셜 프라이버시가 가지고 있는 문제점들에 대해 알아본다. 또한 AES와 같은 대칭키 암호화 알고리즘과 Shamir의 비밀 분산 기법을 이용하여 이에 대한 해결책을 새롭게 제안한다.

• 융합보안논문지
• /
• 제24권2호
• /
• pp.133-140
• /
• 2024

이용자가 직접 유전체 검사를 의뢰하는 DTC(Direct-to-Consumer) 유전자검사가 확산되고 있다. 수요 확대에 따라 인증제도를 통한 비 의료기관에 검사자격을 부여하고, 검사항목을 확대하였다. 그러나 제약이 적은 국외 사례와 달리 국내 제도에서는 여전히 질병 검사항목은 제외한다. 기존의 비식별 방식은 유전체 정보의 고유성과 가족 공유성에도 영향을 미쳐 충분한 활용 가능성을 보장하지 못한다. 따라서 본 연구는 서비스 활성화 및 검사 항목 확대를 위한 방안으로 분석과정에 완전동형암호를 적용하여 유전체 정보의 유용성을 보장하되, 유출 우려를 최소화한다. 또한 정보주체의 자기결정권 보장을 위해 Opt-out을 기반한 프라이버시 보존 모델을 제안한다. 이는 유전체 정보보호와 활용 가능성 유지를 목표로 하며, 이용자의 의사를 반영한 정보의 활용 가능성을 보장한다.

• 대한영상의학회지
• /
• 제83권4호
• /
• pp.759-770
• /
• 2022

인공지능의 연구 개발 및 활용에서 윤리의 중요성이 의료분야뿐 아니라 전 사회적으로 점차 널리 인식되고 있다. 이 종설은 영상의학 영상데이터를 인공지능 연구에 활용할 때 개인정보의 보호 및 데이터에 대한 권리 측면에서 윤리적으로 고려할 사항들에 대해서 국내 독자들에게 실용적인 정보를 제공하고자 한다. 따라서 이 글에 담긴 내용은 많은 부분이 관련된 국내 법과 정부 제도에 바탕을 두고 있다. 인공지능의 연구 개발 및 활용에서 개인정보 보호는 매우 중요한 윤리적 원칙이며 연구 데이터의 적절한 가명처리는 개인정보 보호를 위한 핵심 방법이다. 아울러 인공지능 연구 개발에 의료 데이터를 상업적 이해관계를 최소화하며 윤리적으로 공유할 필요성도 부각되고 있다. 연구 데이터 공유는 개인정보 유출의 위험을 증가시키므로 개인정보 보호에 더욱 주의가 필요하다.

• Korean Journal of Radiology
• /
• 제24권12호
• /
• pp.1179-1189
• /
• 2023

Objective: We aimed to evaluate the reporting quality of research articles that applied deep learning to medical imaging. Using the Checklist for Artificial Intelligence in Medical Imaging (CLAIM) guidelines and a journal with prominence in Asia as a sample, we intended to provide an insight into reporting quality in the Asian region and establish a journal-specific audit. Materials and Methods: A total of 38 articles published in the Korean Journal of Radiology between June 2018 and January 2023 were analyzed. The analysis included calculating the percentage of studies that adhered to each CLAIM item and identifying items that were met by ≤ 50% of the studies. The article review was initially conducted independently by two reviewers, and the consensus results were used for the final analysis. We also compared adherence rates to CLAIM before and after December 2020. Results: Of the 42 items in the CLAIM guidelines, 12 items (29%) were satisfied by ≤ 50% of the included articles. None of the studies reported handling missing data (item #13). Only one study respectively presented the use of de-identification methods (#12), intended sample size (#19), robustness or sensitivity analysis (#30), and full study protocol (#41). Of the studies, 35% reported the selection of data subsets (#10), 40% reported registration information (#40), and 50% measured inter and intrarater variability (#18). No significant changes were observed in the rates of adherence to these 12 items before and after December 2020. Conclusion: The reporting quality of artificial intelligence studies according to CLAIM guidelines, in our study sample, showed room for improvement. We recommend that the authors and reviewers have a solid understanding of the relevant reporting guidelines and ensure that the essential elements are adequately reported when writing and reviewing the manuscripts for publication.

• 한국간담췌외과학회지
• /
• 제26권3호
• /
• pp.277-280
• /
• 2022

Backgrounds/Aims: Biliary colic is a common cause of emergency department (ED) visits; however, the natural history of the disease and thus the indications for urgent or scheduled surgery remain unclear. Limitations of previous attempts to elucidate this natural history at a population level are based on the reliance on the identification of biliary colic via administrative codes in isolation. The purpose of our study was to validate the use of International Statistical Classification of Diseases and Related Health Problems codes, 10th Revision, Canadian modification (ICD-10-CA) from ED visits in adequately differentiating patients with biliary colic from those with other biliary diagnoses such as cholecystitis or common bile duct stones. Methods: We performed a retrospective validation study using administrative data from two large academic hospitals in Toronto. We assessed all the patients presenting to the ED between January 1, 2012 and December 31, 2018, assigned ICD-10-CA codes in keeping with uncomplicated biliary colic. The codes were compared to the individually abstracted charts to assess diagnostic agreement. Results: Among the 991 patient charts abstracted, 26.5% were misclassified, corresponding to a positive predictive value of 73% (95% confidence interval 73%-74%). The most frequent reasons for inaccurate diagnoses were a lack of gallstones (49.8%) and acute cholecystitis (27.8%). Conclusions: Our findings suggest that the use of ICD-10 codes as the sole means of identifying biliary colic to the exclusion of other biliary pathologies is prone to moderate inaccuracy. Previous investigations of biliary colic utilizing administrative codes for diagnosis may therefore be prone to unforeseen bias.