친자 및 혈연관계 분석시스템의 성능을 개선하기 위한 지식기반 전문가 시스템 개발

구교찬(Kyo-Chan Koo),김선욱(Sun-Uk Kim) 한국산업경영시스템학회 2015 한국산업경영시스템학회 학술대회 Vol.2015 No.추계

This paper considers a paternity and kinship analysis system(PKAS) being currently used in real sites. A knowledge-based expert system is proposed to improve the performance of PKAS in terms of accuracy, speed, training time, and satisfaction, which are common measures for evaluation. The knowledge base, one of the most important components in the knowledge-based expert system(KBES), consists of a rule made from random matching algorithm, decision rules of allele types and guide rules of options. The last two rules are learning incrementally from sample data. The results show that PKAS armed with the expert system ensures the better performance with regard to these criteria than the existing system. Especially as far as speed is concerned, as the sample size increases, it outperforms the existing one. As the number of samples increases, while processing time increases nearly exponentially in the existing PKAS, it does linearly in our proposed system.

유전자 분석 자료에 의한 친자 및 혈연관계 분석시스템 개발 및 활용

구교찬(Koo, Kyo Chan),김선욱(Kim, Sun-Uk) 한국산학기술학회 2015 한국산학기술학회논문지 Vol.16 No.10

최근 실종자, 변사자, 미아 등의 유전자 분석 자료는 지속적으로 증가하고 있으나, 현재 친자확인을 위한 통계학적 계산은 대부분 수기에 의하거나 엑셀을 통해서 이루어지고 있다. 따라서 유전자 분석 자료 중 상염색체 Short Tandem Repeat (STR)을 체계적으로 관리하고 효과적으로 분석할 수 있는 소프트웨어의 개발이 필요하다. 친자관계 및 혈연관계를 다양한 옵션 하에서 용이하게 분석하는 웹 기반 유전자자료 분석시스템이 광범위한 테스트 없이 약 20개월의 연구를 통해서 개발되 었다. 친자관계 분석을 위해서 부계지수 계산 알고리즘을 사용하였고, 혈연관계 분석을 위해서 Identity by descent (IBD) 공식 을 사용하였다. 이 시스템은 실제 데이터를 기반으로 혈연관계지수와 친자확률이 검증됨으로써 신뢰성이 확보됨은 물론, 대 량 재난·재해 시 발생될 유전자 분석 자료의 관리 및 분석에 효과적으로 이용될 수 있을 것이다. 이 외에도 본 시스템은 데이 터베이스와 알고리즘의 통합 환경, 사용자 중심 인터페이스, 프로세스 자동화 등 고급기능을 포함한다. Recently, DNA data of missing person, killed person, and missing child continue to increase but most of statistical calculation for paternity confirmation is being done through manual methods or Excel. Therefore, we need development of a software which is able to facilitate both systematic management and effective analysis of Short Tandem Repeat (STR) derived from DNA data. Without extensive testing, through a twenty-month study was developed a web-based system which performs paternity analysis and kinship analysis easily based on the various options. The former uses an existing algorithm for paternity index and the latter does Identity by descent (IBD) formula. Due to our system validated over real datasets in terms of likelihood ratio and probability of paternity, it ensures increased reliability as well as effective management and analysis of DNA data in mass disaster. In addition, it includes advanced features such as an integrated environment, user-centered interface, process automation and so on.

들기 작업의 허리부하 평가 및 자세 추천을 위한 작업자 중심의 웹기반 전문가시스템

한승조,구교찬 한국융합과학회 2023 한국융합과학회지 Vol.12 No.2

Purpose Low back pain caused by lifting tasks accounts for the highest incidence of occupational diseases. This study aimed to analyze the effect of lifting tasks on low back pain and construct a web-based assessment system that can aid to prevent low back pain. Methods Lifting task-related factors (trunk angle, knee angle, stoop/squat lifting, trunk bending, foot position, lifting velocity, horizontal distance, and asymmetrical lifting) were investigated for their effect on the compressive force through theoretical, experimental, and biomechanical approaches. The compressive force was adjusted to reflect the influence of the load center of gravity. The information obtained from these analyses was used as the knowledge base for training the assessment system, termed as the “expert system.” Result The expert system provides the final compressive force and recommends lifting posture based on the values input or selected by the worker. The worker can obtain information about the low back loading level directly on the screen and can adjust the low back load by changing the posture according to the recommended lifting postures. Conclusion We designed a web-based expert system that can be self-operated by workers to measure low back loading and recommend lifting postures. This system can help in overcoming the limitations of the existing low back loading measurement models and can be useful in small businesses that cannot hire experts, thereby helping in reducing direct and indirect costs related to musculoskeletal disorders, including low back pain.

VR 자전거의 사용자 경험 설계 모델 예측에 관한 연구

조재형,구교찬,한승조,김선욱 한국디지털정책학회 2018 디지털융복합연구 Vol.16 No.11

MBTI 성격유형에 따른 투자성향 연구

한승조,고현민,구교찬 대한산업공학회 2015 대한산업공학회 춘계학술대회논문집 Vol.2015 No.4

This study investigated the relationship between MBTI personality types of potential investors and their cognitive - behavioral investment disposition. 3 questionnaires relating to MBTI, cognitive investment disposition, and behavioral one from 400 subjects were collected and analyzed. Based on both cognitive and behavioral investment disposition scores, potential investors having E, S, and T types had a tendency to endure the risk resulting from their investment better than ones who having I, N, and F types. However, the difference between J and F types was not significant. Also investor group with ENF combination had the most aggressive investment disposition among other groups. On the contrary the group with ISF tended to avoid the investment risk. This research is expected to be used as basic data with which investment companies and banks recommend adequate financial instruments to consumers.

VR 영상이 신체 안정성에 미치는 영향 : 손 안정성을 중심으로

한승조,김선욱,구교찬,이근주,조민수 한국디지털정책학회 2017 디지털융복합연구 Vol.15 No.8

다양한 유기용매와 물 경계면에서의 양친매성 금나노입자의 흡착 거동

임연수,이영민,구교찬,이희영 한국접착및계면학회 2024 접착 및 계면 Vol.25 No.1

양친매성 금나노입자는 표면에 친수성 및 소수성 리간드가 동시에 결합된 입자로 에너지, 바이오, 광학, 전자 기술 분야 등 그 응용분야가 다양하다. 특히 이러한 양친매성 금나노입자는 친수성과 소수성 특성을모두 지니고 있어 서로 섞이지 않는 두 유체의 경계면에서 계면을 활성화하고 정렬된 구조체를 형성할 수 있다. 여기서, 금나노입자의 표면 특성은 두 유체의 경계면에서 금나노입자의 거동에 중요한 영향을 미친다. 따라서본 연구에서는 양친매성 금나노입자의 표면 특성 및 유기용매의 종류에 따른 금나노입자의 유기용매-물 경계면에서의 흡착 거동을 확인하였다. 소수성 리간드의 탄화수소가 길수록, 유기용매의 탄화수소가 짧을수록 계면에흡착되는 금나노입자의 양이 증가함을 확인하였다. 또한 Langmuir 흡착등온식을 토대로 양친매성 금나노입자가 단일층을 이루고 있음을 확인함과 동시에 유의미한 열역학적 값(KL, Cmax)들을 얻었다. Amphiphilic gold nanoparticles, synthesized by the simultaneous binding of hydrophilic and hydrophobic ligands on their surfaces, find diverse applications in energy, bio, optical, electronic technologies, and various other fields. Particularly, these amphiphilic gold nanoparticles possess both hydrophilic and hydrophobic characteristics, enabling them to activate interface at the interface of immiscible fluids and form org anized structures. T he surface p roperties of g old nanoparticles play a c rucial r ole in influencing the behaviors of amphiphilic gold nanoparticles at the interface of two fluids. Therefore, this study investigated the adsorption b ehaviors o f g old nanoparticles at t he o rganic s olvent-water i nterface b ased o n the surface characteristics of amphiphilic gold nanoparticles and the type of organic solvents. It was observed that the amount of adsorbed gold nanoparticles at the interface increased with the length of hydrocarbon chains in hydrophobic ligands and increased with shorter hydrocarbon chains in the organic solvent. Furthermore, using the Langmuir isotherm model, the study confirmed the formation of a monolayer by amphiphilic gold nanoparticles and obtained significant thermodynamic parameters simultaneously.

MBTI 성격유형에 따른 투자성향 연구

한승조,고현민,구교찬 한국경영과학회 2015 한국경영과학회 학술대회논문집 Vol.2015 No.4

This study investigated the relationship between MBTI personality types of potential investors and their cognitive - behavioral investment disposition. 3 questionnaires relating to MBTI, cognitive investment disposition, and behavioral one from 400 subjects were collected and analyzed. Based on both cognitive and behavioral investment disposition scores, potential investors having E, S, and T types had a tendency to endure the risk resulting from their investment better than ones who having I, N, and F types. However, the difference between J and F types was not significant. Also investor group with ENF combination had the most aggressive investment disposition among other groups. On the contrary the group with ISF tended to avoid the investment risk. This research is expected to be used as basic data with which investment companies and banks recommend adequate financial instruments to consumers.

레시틴/담즙염/데케인 유기젤에 대한 물의 영향

추은애,김나현,강민석,구교찬,이희영 한국접착및계면학회 2023 접착 및 계면 Vol.24 No.4

Lecithin self-assembles into reverse spherical micelles in organic solvents as an amphiphilic molecule. When additives such as bile salts and water are introduced into lecithin solutions, it induces structural changes in the molecular form of lecithin, leading to the transformation into reverse cylindrical micelles. In this study, we observe the rheological changes of lecithin/bile salt mixtures in a decane system after the addition of water. The resulting mixtures exhibit high viscosity and characteristics of viscoelasticity, suggesting potential applications in various fields such as drug delivery and edible oil gels.

데이터마이닝을 이용한 메뉴 설계 방법에 관한 연구

손혜민(Hyemin Son),구교찬(Kyo-Chan Koo) 한국산학기술학회 2023 한국산학기술학회논문지 Vol.24 No.2

본 연구는 메뉴 기반 사용자 인터페이스를 최적화하기 위한 데이터마이닝 기반 인터페이스 개발 방법(DaMIM: Data Mining-based Interface development Method)에 관한 연구로서 주요 내용은 DaMIM 절차의 제안과 검증으로 구성된다. DaMIM의 절차는 6단계로 대상 업무 정의, 사용자 선정, 자료수집, 데이터마이닝 기법 적용, 분석 및 평가, 대안 확정으로 제안하였다. 이 제안에서 데이터마이닝 기법으로는 군집분석을 사용하였다. 이후 이를 검증하기 위해 메뉴 선택 반응시간을 측정할 수 있는 프로그램을 개발하여 실험하였다. 실험은 메뉴 수에 따라 저밀도(9개), 고밀도(14개)로 구분되며 메뉴 구성 방법에 따라 알파벳순 방법(AM), 객체 지향적 방법(OM), 프로세스 지향적 방법(PM), 데이터마이닝 기반 방법(DM)으로 구분된다. 실험을 통해 측정한 시간을 일원배치 분산분석한 결과 저밀도에서는 메뉴 구성 방법별 메뉴 선택 반응시간이 통계적으로 유의한 차이를 보이지 않았다. 그러나 고밀도에서는 DM과 AM의 반응시간 이 가장 짧은 그룹으로 나타났으며, DM의 평균 반응시간이 가장 짧은 것을 확인함으로써 효과성을 검증하였다. 이에 본 연구에서 제안한 방법론은 사용자의 의견을 효과적으로 반영함으로써 인터페이스 설계의 객관화와 생산성 향상에 기여할 수 있을 것으로 기대된다. This study proposes and validates a data mining-based interface development method (DaMIM) to optimize menu-driven user interfaces. DaMIM comprises six steps—target task definition, user selection, data collection, application of data mining techniques, analysis and evaluation, and determination of alternatives. Cluster analysis is used as the data mining technique. The proposed method is validated using a program developed to measure the reaction times of menu selection. The experiments are classified as low density (9) or high density (14) depending on the number of menus, and alphabetical method (AM), object-oriented method (OM), process-oriented method (PM), or data mining-based method (DM) are used for menu organization. The results reveal that statistically significant differences are not observed between the menu selection response times of the different menu organization methods at low density experiments, However, the response time of DM is the shortest in the high density case. Therefore, the effectiveness of DM is verified. The conclusions of this study are expected to contribute toward the objectification and productivity improvement of user-based interface design.