安山 新都市 住居地域 街路邊의 用途와 環境의 變化에 관한 硏究

이석환 서울大學校 1991 국내석사

Thermo-fluid diagnostic methods using laser-induced plasma and ultrasound wave

이석환 서울대학교 대학원 2015 국내박사

We investigated the LIBS (laser-induced breakdown spectroscopy) characteristics in various phase conditions including single-phase (gas) and two-phase (liquid gas, solid gas) for thermo-fluid applications. For single-phase (gas) analysis, two-dimensional mapping of the LIBS signals from chemical species information was performed in propane flames with in situ diagnostics. Key combustion information, such as density, fuel concentration, and fuel/air equivalence ratio were provided by the LIBS measurement. For the two-phase spray flame, the simultaneous laser ignition and spectroscopy enabled rapidly determining the local equivalence ratio and condensed fuel concentration during the reaction. In parallel with the laser ignition, the equivalence ratio and droplet characteristics, such as concentration, size, and distribution of spray combustion, were simultaneously obtained for a feedback control system. For the two-phase aerosol analysis, LIBS detected solid carbon particulates in a flow system that was designed to replicate aircraft exhaust flow conditions. The detected signals from the emissions stream at velocities of up to 70 m/s showed that in-situ characterization of carbon particulates in the high-speed exhaust were proven feasible. The results obtained through this research will allow for more practical diagnostics in various applications such as biology, combustion and the environment. In addition, basic research was conducted for the thermo-fluid diagnostics using LIUS (laser-induced ultrasound) waves. We demonstrated that the reduced-graphene-oxide-coated thin aluminum film (rGO-Al) and rGO-coated polydimethylsiloxane (rGO-PDMS) were effective optoacoustic transmitters for generating high-pressure and high-frequency ultrasound waves. Under pulsed laser excitation, rGO-Al and rGO-PDMS transmitters generated enhanced optoacoustic pressure that was 64 and 76 times stronger, respectively, than Al alone. Promising optoacoustic wave generation was made possible by optimizing the thermoelasticity of metal, the PDMS film and the thermal conductivity of rGO in the proposed transmitter for laser-induced ultrasound applications. LIBS and LIUS have great advantages as analytical techniques, namely, real-time rapid analysis and stand-off detection capability, which are all material phases for diagnosing thermo-fluid phenomena.

Effect of Elastic Collision and Asymmetric Sheaths on Standing Wave Property in Capacitively Coupled Plasma : 탄성 충돌과 비대칭 쉬스가 축전결합형 플라즈마의 정상파 특성에 미치는 영향

이석환 서울대학교 대학원 2013 국내박사

축전결합형 플라즈마원(CCP)에서 여기 주파수의 파장 길이에 비해 전극의 크기가 무시할 수 없을 정도로 운전 주파수가 높아짐에 따라 정상파 형성에 따른 불균일한 플라즈마 분포가 나타나게 된다. CCP에서 형성되는 정상파의 특성 및 플라즈마 분포 특성을 이해하기 위하여 해석적 모델들이 제시되었다. 하지만 기존의 모델은 낮은 압력 조건을 가정하여 탄성 충돌의 영향을 무시하였으며 상하부 전극에 형성되는 쉬스와 전자기장의 분포가 대칭인 비현실적인 조건을 가정하였기 때문에 실제 상황에서 나타나는 현상을 해석하는데 있어 제한이 존재하였다. 최근에 이중 주파수를 이용하거나 복수의 동일 주파수간의 위상을 제어하여 전력을 공급하는 방법을 통해 플라즈마 분포 특성을 제어하는 연구가 활발히 진행되고 있으나 기존의 상하부 쉬스에서 전자기파가 대칭적으로 진행하는 모델에서는 이와 같이 상하부 전극에 다른 전력이 인가되는 경우의 해석이 불가능하였다. 본 연구에서는 탄성 충돌 및 비대칭적인 쉬스를 고려하여 기존 모델을 개선하였으며, 정상파의 특성 중 플라즈마 밀도에 중요한 영향을 미치는 파장 길이 및 축전형 전기장과 유도형 전기장에 의한 전자 가열을 분석하였다. 기존에 단일 주파수의 전력이 인가된 원통형 구조에 국한되어 있던 모델을 확장하여, 사각형 반응기에서도 정상파 형성에 따른 전자기장 분포를 해석하였으며, 이중 주파수 및 위상 제어 CCP에서 인가 전력의 분포를 분석하였다. 다양한 조건에서 실험을 수행하여 그 결과를 기존 모델 및 본 연구에서 제시된 모델과 비교하였다. 60 MHz의 운전 주파수와 1 x 1.2 m의 전극 크기 및 200 mm의 전극 간 간격을 갖는 사각형 반응기에서 파장 길이의 변화 및 축전형 전자 가열형 플라즈마 분포에서 유도형 전자 가열형 플라즈마 밀도 분포가 변화하는 것을 관찰였다. 100 MHz와 2 MHz의 운전 주파수와 0.3 m 의 전극 지름 및 25 mm의 전극 간 간격을 갖는 원통형 반응기에서 단일 주파수, 이중 주파수 및 위상 제어 운전에 따른 플라즈마 밀도 분포 특성 변화를 관찰하였으며, 13.39 MHz와 3.39 MHz의 운전 주파수와 2.1 x 2.4 m의 전극 크기 및 300 mm의 전극 간 간격을 갖는 사각형 반응기에서 이중 주파수의 인가 방식에 따른 플라즈마 밀도 분포 특성을 분석하였다. 기존의 모델은 실험 결과를 잘 설명하지 못하는데 비해 본 논문에서 제시한 모델은 전극 간 간격, 운전 주파수 및 압력, 반응기 및 전력 인가 구조 등 다양한 조건에서 실험 결과와 잘 일치하였다. 실험 결과 운전 주파수가 탄성 충돌 주파수보다 매우 높은 경우, 운전 압력 증가에 따라 축전형 전기장에 의한 전자 가열이 감소하여 전극 중앙의 플라즈마 밀도가 감소하며, 특히 위상 제어형 CCP의 경우 플라즈마 밀도 분포가 두드러지게 변화하는 것을 관찰하였다. 전극 간 간격이 플라즈마의 침투 깊이에 비해 큰 경우, 파장 길이는 플라즈마 밀도에 거의 영향을 받지 않으며, 낮은 전자 밀도 조건에서도 유도형 전기장에 의한 전자 가열이 크게 나타날 수 있음을 확인하였다. 사각형 반응기에서 유도형 전기장에 의한 플라즈마 밀도 분포는 마름모의 꼭지점에 해당하는 위치에서 정점을 가지며, 이중 주파수 CCP의 경우 각각의 전력을 동일한 전극에 인가하는 경우와 별도의 전극에 인가하는 경우 플라즈마의 밀도 분포가 달라짐을 확인하였다. 탄성 충돌은 벌크 플라즈마 내에서 정상파의 저항성 전력을 분석하는데 있어 중요한 인자이며, 이를 무시할 경우 쉬스 에서의 전자 가열에 비해 벌크 플라즈마 내에서의 축전형 전기장에 의한 전자 가열을 과소 평가하게 된다. 비대칭적인 정상파의 전자기장 성분의 분포로 인해 전력이 인가된 전극의 쉬스 부근에 집중된 강한 전기장이 존재하여 벌크 플라즈마 내에 강한 저항성 전자 가열을 형성한다. 비대칭 전자기장 분포의 효과는 침투 깊이에 비해 전극 간 간격이 커질수록 두드러지게 나타나며, 유도형 전기장에 의한 전자 가열을 이해하는데 있어 중요하다. 특히 이중 주파수와 위상 제어 CCP와 같이 복수의 전력원에 의한 정상파 형성 및 그로 인한 플라즈마 분포를 해석함에 있어 비대칭성 쉬스를 고려하는 것은 필수적이다. As the driving frequency increases, the wavelength of the excitation frequency becomes comparable to the dimension of the electrode and plasma is nonuniformly distributed due to the standing wave in capacitively coupled plasma (CCP) sources. Analytic models have been suggested to explain the properties of the standing wave and plasma distribution in CCP. Since the impractical conditions such as the collisionless plasma and symmetric field distributions with the symmetric sheaths at the powered and counter electrodes are assumed, the previous model has limitation in analysis of the practical situations. Recently, dual-frequency and phase-controlled CCP sources attracts interests but the analysis of plasma distribution with different voltages applied at the electrodes is impossible with the previous models due to the assumption of the symmetric propagation of waves in the top and bottom sheaths. In this study, the previous models are modified considering elastic collision and asymmetric sheaths to analyze wavelength and electron heating powers due to the capacitive and inductive fields which are important parameters of the standing wave for the plasma density distribution. The previous models which are limited to the single frequency driven cylindrical reactor are extended to derive the field components in the rectangular reactors and to analyze the distribution of heating powers in dual-frequency driven and phase controlled CCP sources. Experimental results are obtained in various conditions and compared with the results of the previous and newly derived models. Variations of the plasma distribution from the capacitive to inductive heating mode and the wavelength are observed in a 60 MHz driven rectangular reactor with the electrode of 1 1.2 m and the gap distance of 200 mm. Distributions of heating powers in a cylindrical reactor with the electrode of 0.3 m in diameter and the gap distance of 25 mm in single-frequency, dual-frequency and phase controlled operations was analyzed. Plasma distributions are compared depending on the way the input powers are applied in a dual rectangular reactor driven by 13.39 and 3.39 MHz powers with the electrode of 2.1 2.4 m and gap distance of 300 mm. The results from the model in this study agree well with the experimental results while the results from the previous models do not explain the experimental results. With the driving frequency much higher than the elastic collision frequency, the capacitive heating decreases as the operating pressure increases and the plasma density near the center of the electrode decreases., which is significant in the phase controlled CCP source. For the gap distance much longer than the skin depth, the wavelength of the standing wave is insensitive to the plasma density and the inductive heating is substantial even in the low density discharge. Plasma density due to the inductive heating is distributed in the rhombus shaped profile in the rectangular reactors. In the dual-frequency driven CCP sources, distribution of plasma density changes depending on the configuration of the input powers whether they are applied to the separate electrodes or to the same electrode. Elastic collision is an important factor to analyze the ohmic power dissipation of the standing wave in the bulk plasma. Neglecting the elastic collision results in the underestimation of the ohmic heating powers in the bulk plasma compared to the heating at the sheath such as stochastic heating and the ohmic heating in the sheath. Asymmetric field distribution of the standing wave with the asymmetric sheaths results in the strong electric field concentrated in the bulk plasma near the sheath at the powered electrode and the larger ohmic heating power. The effect of the asymmetric field distribution is significant with the gap distance much larger than the skin depth and important in the analysis of the inductive heating power. To understand the plasma distribution due to the standing waves from the multiple power sources as in the dual-frequency driven and phase controlled CCP sources, consideration of the asymmetric sheath is essential.

崔漢綺『人政』<鑑枰>의 人事評價에 關한 硏究 : 現代 企業의 人事評價方式과 關聯하여

이석환 성균관대학교 2007 국내석사

ABSTRACT A study of HR management on『Injeong(人政)』<Gampyung(鑑枰)> by Choi han gi Lee Seog-hwan Graduate School of Confucian Studies Eastern philosophy and culture division Sungkunkwan University The most essential thing applied to run most organizations is to find out how to choose human resources, so the good beginning of all things can be started by proper human management. All the time, a lot of effort has been put into hiring excellent human resources and managing organizations. Our society has badly focused on merit system. The merit system and materialism have been important to our society since the 'IMF era'. Also, efficiency and the merit system are key roles to take control of managing businesses. In addition, fierce competition and reward for winners replace the importance of manpower. Thus, our society has stuck to superficial outcome and ignored natural process. The selfishness which is prevalent in our society causes environmental problems and makes people be neglected as well. Most of the businesses also have faced those social problems. However, they can develop new values and give them to our society. Therefore, most businesses can play a leading role in building business ethics and new values which show human being is of importance to our society by focusing on reasonable human management system, the essential factor in taking care of their own businesses. Some problems in our society and companies have been caused by materialism and the values focusing on the merit system. That's why I study human management system and problems of modern companies by comparing them to the values in our tradition. There are several traditional ideas in modern Korean society, one of which is Confucianism that has a strong effect on Korean society. Confucianism sets standards for both Koreans' ways of thinking and their ways of behaving. The study tells that it was based on human being and then come to human being, for studies on human being in Confucianism have been researched and practised for a long time. The studies of Confucianism in the late Joseon Dynasty in 19th century are the basic ideas of my study. That's the first reason why I choose Injeong(人政, human politics) written by Choi han-ki(崔漢綺, 1803-1877) as the basic study. Secondly, he accepted western science, studied it, and improved his own study based on Confucianism. Injeong(人政) consists of 25 volumes telling how to manage human being. The book is divided four chapters; Cheukinmun(測人門, chapter on understanding people), Gyoinmun(敎人門, chapter on teaching people), Seoninmun(選人門, chapter on choosing people), and Younginmun(用人門, chapter on stationing people). They are divided by each role and related to the order of hiring people. Cheukin(測人) is the start of human management and Youngin(用人) is the purpose of it. Gyoin(敎人) and Seonin(選人) are the process of human management. There is Gampyung(鑑枰, the table of evaluation for people) in the last part of Cheukinmun(測人門). Cheukin(測人) is a way of measuring personality while Gampyung(鑑枰) is a list of how to classify people by basing on practical demands. Gampyung(鑑枰) is made up of Ogu(五具, five factors used for evaluating some qualities of a person) and Obal(五發, five assisting factors used for helping Ogu(五具)). Every factors is used as the standard for Sojang(消長, giving high scores and low scores by evaluating each quality of a person's) and according to the good and bad of a person's character, high and low scores are given to a person. Ogu(五具) have Gipum(氣稟, gift), Simdeok(心德, attitude to all), Cheyung(體容, appearance), Mungyun(聞見, education), and Cheoji(處地, circumstance). These essential qualities are what human being has. Obal(五發) consist of Jaeguk(才局, improved gift), Eongbyun(應變, ability to cope with things), Pungdo(風度, appearance), Gyungrun(經綸, social experience), and Gosi(措施, ability to deal with work). Those assisting factors are manifested by Ogu(五具). Gipum(氣稟) has four types of a person's character; Gang(强, determined and robust person), Yak(弱, helpless person), Cheong(淸, gentle and flexible person), and Tak(濁, negative and hot-tempered person), which tell what he is. Gang(强) and Chung(淸) are persons who have good qualities and they can get high scores while Yak(弱) and Tak(濁) bad qualities and get low scores. Gipum(氣稟) is the mental and physical quality which a person has when he/she was born. Simdeok(心德) also has four types of a person's character; Sung(誠, true person), Ui(僞, hypocritical person), Soon(純, innocent), and Pak(駁, rude and uneducated person). Sung(誠) and Soon(純) get high scores while Ui(僞) and Pak(駁) low scores. Cheyoung(體容) has four types of a person's character; Hu(厚, generous person), Pak(薄, selfish person), Mi(美, beautiful person), and Chu(醜, ugly and dirty person). Hu(厚) and Mi(美) are good to get high scores and Pak(薄) and Chu(醜) to low scores. However, Cheyoung(體容) also means that a person takes care of himself with right attitude including beautiful appearance. Mungyun(聞見) has four types; Ju(周, sophisticated and experienced person), Bi(比, narrow-minded person), Ah(雅, graceful person), and Sok(俗, materialistic person). Ju(周) and Ah(雅) are persons who have good qualities for getting high scores while Bi(比) and Sok(俗) for low scores. Mungyun(聞見) is the ability to do study and work. Cheoji(處地) has four types; Gui(貴, noble person), Cheon(賤, common person), Bu(富, rich person), and Bin(貧, poor person). Gui(貴) and Bu(富) can get high scores while Cheon(賤) and Bin(貧) low scores. Cheoji(處地) is a person's situation when he/she was born and can be changed by his/her will. The value of Gipum(氣稟) is 4 scores, Simdeok(心德) is 3 scores, Cheyung(體容) is 2 scores, Mungyun(聞見) is 1 score, and Cheoji(處地) is a half score. The high and low scores of each quality are similar to the process of human being's growth. Those qualities come from individual to society and from inside to outside and make relationships among them. Ogu(五具) has the combination of each five types, which shows 1,024 persons' types. Hyegang(惠岡, Choi han-ki's pen name) created useful study by using mathematical measurement. That's because he established practical study, not abstract one. He wrote Gampyung(鑑枰) by basing on his Ki(氣) philosophy. His energy theory is his own philosophy added traditional energy theory which is showing all in universe to experiential philosophy in western science. Everything in universe is formed by energy, which moves in certain ways and then formed certain things. If energy formed a human being, it is called Ilsinwunhwa(一身運化, an organic structure created by Wunhwa), and if energy formed a society, it is called Tongminwunhwa(統民運化, Wunhwa of building relationships between a human being, society, nation). If it formed the universe, it is called Cheonjiwunhwa(天地運化, Wunhwa of controling the Universe and its principle). The three kinds of Wunhwa(運化, transformative action) are made by energy, and they are the basic of interacting among all things. When energies from each area meet one anther and move in certain ways and formed things, we call it Gyojeobwunhwa(交接運化, Wunhwa of interaction) and it builds new energy's moving in certain ways to form new things. Gyojeob(交接, interaction among all things) is short from Gyoinjeobmul(交人接物), meaning all things under the sun interact one anther. All is made by interaction. A person can understand others by reflecting himself, and from outside, the inside of all things can be understood, which is based on Wunhwa(運化) and Gyojeob(交接). Recently, some businesses have judged their employees only by their ability to work. Thus, some of them have been in trouble because of some employees who try to achieve their goal immorally. Most of the businesses have come to realize the importance of ethics management and CSR(corporate social responsibility), so they make organizations which are to take care of corruptions, have the code of business ethics, and try to donate certain amount of their profit. However, the main judgement which is used to hire employees working for those organizations is decided by how well they work, so it is hard to believe such employees who do not have business ethics. It is the time that we need new evaluation factors to choose good employees. Fortunately, we have the evaluation factors which came from our tradition. I expect Korean companies to use the evaluation factors. For example, Gipum(氣稟) can be used to evaluate a person's basic quality, Simdeok(心德) a person's attitude to all, Cheyoung(體容) a person's improvement, and Mungyun(聞見) a person's knowledge for his work and his way of dealing with his work. It is not right to absorb in western values by ignoring our traditional values, because some western values are not always perfect in solving our problems. Gampyung(鑑枰) can be the key not only for solving our problems but also for building new standards to judge a person's quality. There have been some wrong examples which proved how important human management was in our society. Whether he/she runs big organizations or not, he/she has to have certain right standards and principles. Though Injeong(人政) was written 150 years ago, it shows that proper human management can play a leading role in managing companies. Besides, we should find out our ancestors' excellent ways of thinking and then study their values for improving our society. Also, we should be proud of our traditional values and continue succeeding to and improving it.

우리나라 연안의 동․하계 검조위와 유속 관측자료의 상관성 연구

이석환 국립부경대학교 대학원 2024 국내석사

This paper aims to examine the correlation between tidal levels and field observation current velocity during the winter and summer seasons off the coast of Korea. The data used in the study are current velocity data collected and obtained by the researcher from 2019 to 2023 (a total of 5 years) in the summer and winter seasons at a total of 28 points in the East Sea, South Sea, and West Seas, as well as 20 tidal level data located near the field observation peak among the tidal stations of the Korea Hydrographic and Oceanographic Agency(KHOA). Through this, the correlation between the tidal level and the water current of the entire coast of Korea, which is proposed in this study, makes it possible to estimate the approximate current velocity of the target sea area using the tidal level data provided by Korea Hydrographic and Oceanographic Agency(KHOA), even if the tidal current field observation is not carried out, which requires a lot of time and cost. This can be provided as basic data for researchers and marine engineering designers when researching the marine environment. The detailed findings obtained from this study are summarized as follows. (1) As a result of analyzing the observed current velocity off the coast of Korea, it was found that the maximum current velocity ranges from 21.95 cm/s in the bottom layer of Yeongdeok (2023.01) to 195.41 cm/s in the surface layer of Boryeong (2022.08) and from 5.55 cm/s in the bottom layer of Yeongdeok (2023.01) to 54.68 cm/s in the surface layer of Boryeong (2022.08) in the case of mean current velocity. In general, the observed current was weak on the East Sea, where the tidal level is small, and strong on the West Sea, where the tidal level is large. (2) According to the tidal level data and analysis results of the tidal station, both the Spring range and the Mean sea level showed a tendency to increase in the order of East Sea≤ South Sea < West Sea. Analysis of the tide form factor showed that the tidal pattern was mixed tide with a predominance of diurnal tide in Hupo in East Sea, and semidiurnal tide predominantly in the entire sea area of East Sea except Hupo, Wando in the South Sea, Mokpo and Heuksando in the West Sea. In other areas, tidal patterns in the form of semidiurnal tides were exhibited. (3) As a result of analyzing the correlation between tidal level and observed current velocity, both spring tide and mean sea level showed a larger coefficient of determination when analyzed with mean velocity than maximum current velocity. In addition, compared to the surface, which is relatively affected by external factors such as wind, the tidal form factor was found to be larger at the bottom. Seasonally, the value was greater in winter than in summer. In addition, in the analysis of the correlation between the tidal level and the tidal current harmonic constant, the coefficient of determination was not significantly different from that of the tidal level and the observed current velocity. (4) As a result of determining the correlation between tidal level and observed current velocity by sea area and season, the correlation equation was y=0.0448x+0.1472 and the coefficient of determination was 0.3843 for mean sea level and mean current velocity on the sea surface during the entire observation period. In addition, in the case of bottom, the correlation equation was y=0.0472x+0.0998, and the coefficient of determination was 0.4775, which showed a relatively large value. The largest coefficient of determination was the relationship between the spring tide and the mean velocity, with a correlation of y=0.0287x+0.0896 and a coefficient of determination of 0.5503.

트리거(trigger)를 이용한 튜브 충돌의 최적화

이석환 서울대학교 대학원 2007 국내석사

협력적 필터링 추천 시스템의 성능 향상

이석환 인하대학교 대학원 2010 국내박사

정보기술과 인터넷 기술의 발전은 정보의 양을 크게 증가시키고 있다. 그러나 정보량의 증가는 사용자가 원하는 정보를 찾기 위한 노력과 시간을 증가시킨다. 데이터마이닝은 이와 같은 대량의 정보 중에서 사용자가 원하는 정보를 적은 노력으로 단시간에 찾을 수 있도록 도와준다. 데이터마이닝은 마케팅, 생산관리, 의료 등의 다양한 분야에서 응용되고 있다. 마케팅 분야에서 추천 시스템은 데이터마이닝을 이용하여 사용자나 상품 등과 같은 정보를 분류하고 분석하여 목표고객의 기호에 맞는 상품을 추천해준다. 추천에는 내용기반 필터링과 협력적 필터링의 두 가지 기법이 있다. 내용 기반 필터링은 상품의 특징을 나타내는 속성과 목표고객의 기호를 가지고 있는 프로파일을 비교하여 이 두 가지 요소가 일치하는 상품을 찾아서 추천한다. 반면에 협력적 필터링은 목표고객에 대해서 유사한 기호를 가진 이웃사용자를 찾고 그 이웃사용자가 평가한 상품의 선호도를 이용하여 목표고객에게 상품을 추천한다. 협력적 필터링은 내용 기반 필터링으로 찾을 수 없는 잠재적인 추천 상품을 발견할 수 있기 때문에 추천 시스템에서 많이 사용되는 방법이다. 협력적 필터링에서 추천의 정확도를 향상시키기 위해서는 목표고객에 대한 이웃사용자를 정확하게 선정해야하고 상품에 대한 목표고객의 선호도를 정확하게 예측하며 예측된 선호도를 이용하여 목표고객의 기호에 맞는 추천목록을 정확하게 생성해야한다. 본 연구에서는 협력적 필터링을 기법을 적용한 추천 시스템에서 추천의 정확도를 향상시키기 위해 다음과 같은 세 가지 방법을 제시하였다. 첫째, 사용자를 유사도에 따라 정확하게 군집할 수 있는 중심이동확장(EMC ; Expanded Moving Center) 휴리스틱 군집 알고리즘을 고안하였다. EMC 휴리스틱 군집 알고리즘은 초기에 군집의 중심배치 방법을 개선함으로써 초기해보다 개선된 해를 찾을 수 있는 확률을 증가시켰고 군집의 중심간 간격을 변경해나가면서 최적해에 근사한 해를 체계적으로 탐색할 수 있다. 한편으로 군집에 소요되는 계산량을 줄여서 군집의 효율을 향상시켰다. EMC 휴리스틱 군집 알고리즘의 성능은 실제로 고객의 선호도를 조사한 MovieLens 데이터를 사용하여 실험하였다. 군집의 수를 변화시키면서 EMC 휴리스틱 군집 알고리즘의 성능을 실험한 결과, 모든 군집 수에서 EMC 휴리스틱 군집 알고리즘에 대한 제곱오차(군집의 성능을 평가하는 척도)는 타 알고리즘의 제곱오차에 비해 항상 좋은 결과를 나타내었다. 그리고 군집의 수가 많아질수록 타 알고리즘에 비해서 제곱오차의 개선 폭이 점차적으로 증가하였다. 이 군집 알고리즘이 추천 시스템의 추천 정확도를 향상시킬 수 있는지 확인하기 위해 이 군집 알고리즘을 추천 시스템에 적용한 결과, 추천의 정확도는 기존의 알고리즘을 적용했을 경우에 비해서 향상되었다. 둘째, 기존의 선호도 예측방법을 개선한 이웃사용자 중심의 선호도 예측방법을 제시하였다. 이웃사용자 중심의 선호도 예측방법은 기존의 선호도 예측방법에 비해서 선호도 예측의 정확도를 높일 수 있다. 기존의 선호도 예측방법은 목표고객이 입력한 선호도 정보를 선호도 예측에 사용한다. 그러나 입력한 선호도 정보가 적을 경우 선호도 예측의 정확도가 낮아지는 문제가 발생한다. 본 연구에서 제시한 이웃사용자 중심의 선호도 예측방법은 목표고객의 이웃사용자가 입력한 선호도를 바탕으로 선호도를 예측한다. 이웃사용자의 선호도 정보량은 목표고객에 비해서 항상 많기 때문에 정보량이 증가하여 선호도 예측의 정확도를 높일 수 있다. 본 연구의 이웃사용자 중심 선호도 예측방법을 추천 시스템에 적용한 결과, 추천의 정확도는 기존의 선호도 예측방법을 사용했을 경우에 비해서 향상되었다. 셋째, 연령속성을 반영한 추천순위 결정방법을 고안하였다. 기존 추천 시스템에서는 상품의 추천순위를 결정할 때 예측 선호도가 높은 순서로 추천순위를 결정하였다. 그러나 본 연구에서는 예측 선호도 외에도 목표고객과 이웃사용자의 연령속성을 반영한 추천순위 결정방법을 사용하여 추천목록을 생성하기 때문에 보다 정확한 추천목록을 생성할 수 있다. 본 연구에서 제시한 추천순위 결정방법을 추천 시스템에 적용한 결과, 추천의 정확도는 기존의 예측 선호도만을 반영한 추천순위 결정방법에 비해서 향상되었다. 본 연구에서 제시한 세 가지 방법을 적용한 추천 시스템의 실험결과는 본 연구와 동일한 MovieLens 데이터를 사용한 기존 연구의 추천 실험결과와 비교하였다. 그 결과 본 연구의 추천 정확도는 기존의 연구에 비해서 향상되었음을 확인하였다. The development of information and internet technology has added great amount of information. However, this mass information causes users to spend much effort and time to find information they need. Data mining is a technology which helps the user to find information they need in a short time with low effort out of mass information. Data mining is applied in various fields such as marketing, product management and medical treatment. In marketing, a recommender system recommends some items to a target customer using data mining by classifying and analysing information concerning the customers or the items. There are two methods of recommendation. One is content-based filtering, and the other is collaborative filtering. In content-based filtering, the system compares characteristics of an item and the profile of the target customer's preference and recommends some items which fit to the target customer's preference. In collaborative filtering, the system searchs for the user groups sharing similar preferences, predicts the preference of the target customer based on the preferences of the neighbour customers, and recommends the items. Collaborative filtering is applied in many recommender systems because it can find potential items which can not be selected by content-based filtering. Selection of the correct target customer is necessary to raise the accuracy of the recommendation in collaborative filtering. The accurate estimation of the target customer's preference for the item and creating a proper recommend list of the items which fit to the target customer's preference should be proceeded. In this paper, the author proposed following three methods to improve the accuracy of the recommender system. First, the author devised a clustering algorithm, that is, EMC(Expanded Moving Center) heuristic algorithm, which can classify the users more accurately according to preferences. The developed EMC heuristic algorithm increases probability to find better solution than the initial solution by the improving method to assign the cluster center at the beginning of clustering process, and it can search for the optimal solution systematically by changing interval from the center of clusters. Additionally, the efficiency of clustering is increased by reducing the computing complexity in clustering. Performance of EMC heuristic algorithm was tested using actual data surveyed from customers, MovieLens. In the experiments of changing the number of clusters, the Squared-error(evaluation value of clustering) of the EMC heuristic algorithm always showed good results than other clustering algorithms in all cluster numbers, and the Squared-errors are improved gradually than other algorithms as the numbers of clusters increased. Also, in the recommender system, the recommendation accuracy is improved when the EMC heuristic algorithm was applied to cluster users than other clustering algorithms. Second, the author proposed the Neighborhood-oriented preference prediction method that improved the conventional preference prediction methods. Neighborhood-oriented preference prediction can improve the accuracy of the preference prediction than the conventional preference prediction methods. The current preference prediction method has used the target customer's preference for preference prediction. It can cause the low accuracy in preference prediction if the data of the target customer's preference is not enough. But, the preference prediction method suggested in this paper can improve the accuracy of the preference prediction because the quantity of the preference information increased by using neighborhood's preference instead of the target customer's. By applying the Neighborhood-oriented preference prediction method to the recommender system, the accuracy of recommendation improved more than the current preference prediction method. Third, the author devised a new recommendation ordering method attributed by the user's age. In the conventional recommendation system, the recommendation order is decided according to the high preference prediction. However, in this paper, the recommendation accuracy is improved by the decision recommendation order method that reflects the age attributes of the target customer and neighborhood in preference prediction. By applying the decision recommendation order method to the recommender system, the recommendation accuracy is improved more than that of the conventional ordering method of recommendation. The test result of the recommendation system which adapted the above three methods suggested in this paper was compared with the result from the conventional recommendation when sharing the same MovieLens data. The recommendation method that was proposed in this paper confirmed the improvement of the recommendation accuracy compared with other studies.

야간 물류 육체 근로자의 에센셜 오일 선호에 따른 육체적 피로의 Aromachology 효과에 관한 연구

이석환 한밭대학교 산업대학원 2019 국내석사

기존 기술 및 신기술과의 융합으로 복잡해지고, 이런 산업구조는 직업의 전문화, 분업화 등 단순 업무를 증가시켰다. 기업간의 생존 경쟁은 고객의 요구를 충족시키고, 보다 빠른 배송으로 고객 만족을 실현 시키고 있다. 이에 물류는 기존의 생산, 판매, 유통에서 유통 전문 기업이 생겨나고, 지속적으로 성장하고 있다. 이런 기업들은 물류리드타임을 최소화하기 위해 야간 상품 이송 시스템을 개발하였다. 따라서 물류 회사는 야간에 상차, 하차를 하는 야간 육체 근로자를 지속적으로 고용하고 있다. 야간 육체 노동에 의한 육체적 피로 및 스트레스는 크게 증가하고 있으며, 이는 만성 피로, 수면 장애의 원인이 되고 있다. 이런 증상은 초기에는 근육통, 식욕부진 등을 일으키고, 장기간 노출되면, 만성피로, 우울증, 소화기 질환, 심혈관계 질환 등 개인의 생명을 위협하는 질병으로 전환되기도 한다. 본 연구는 물류 야간 육체 노동자의 피로 및 스트레스에 대하여 뇌파를 측정하고, 에센셜 오일 제시에 의해 완화 효과를 검증하였다. 연구 대상은 신체 건강한 30대, 40대, 50대 24명을 대상을 실행하였으며, 대조군은 향에 대한 선호도가 없는 피험자로, 실험군은 향에 대한 선호도가 있는 피험자로 선정하였다. 대조군과 실험군 독립표본 t-검정 결과 노동 후 향 제시 델타파좌 P=0.014, t=2.665, 알파파좌/우 각각 P=0.022, P=0.008, t=-2.462, t=2.895, SMR파좌/우 각각 P=0.042, P=0.044, t=-2.157, t=-2.132로 신뢰수준 95%에서 유의차가 인정되었다. 따라서 실험군의 대응표본 t-검정 분석 결과와 대조군과 실험군의 노동 후 향 제시 독립표본 t-검정 결과 Lemon oil 향을 선호하는 집단은 향에 의해 육체 피로에 효과가 있으며, 심리적 안정에 영향을 미친다. 본 연구를 통해 야간 물류 육체 노동자들의 피로 경감에 의한 1차적, 2차적 안전사고 및 질환을 사전 방지할 수 있다. 더 나아가 생산직 노동자에게 적용하면 생산성 향상과 더불어 개인적, 사회적 적응에 높은 효과를 기대할 수 있으리라 사료된다. 본 실험은 향에 대한 주관적 선호를 기반으로 대조군과 실험군을 선정하였다. 뇌파를 통한 향에 대한 선호도를 정립하고, 객관적 선호 기반으로 실험을 진행 할 필요가 있다.

『寶性論』의 佛性認識에 관한 硏究 : 「一切衆生有如來藏品」을 中心으로

이석환 東國大學校 2008 국내석사

The thought of the Buddha-nature(佛性) is the teachings of the Buddha that give emphasis to the point that even though all sentient beings are suffered from afflictive emotions in reality, the inherent purity of their mind are equal to the Buddha's mind by nature. Among the Buddhist texts which deal with the thought of the Buddha-nature, the text of "Ratnagotra-vibh&#257;ga Mah&#257;y&#257;n&ocirc;ttaratan-tra-&#347;&#257;stra"(『究竟一乘寶性論』, abbreviated to Uttaratantra, 『寶性論』) was written, especially, for the purpose of the systematization of the thought of the Tathagatagrabha(如來藏), and the contents of this commentary shows that it well details the original thought and the fundamental meaning of the Buddha-nature formed before the connection with the Mind-Only school. With Uttaratantra(『寶性論』), the early commentary on the thought of the Tathagatagrabha, as a main text, this study is focused on what differences lie in the recognition of the entity of the Buddha-nature as compared with other texts. And, this study proceeds in the manner of re-criticizing the view of the critical Buddhist thinkers, who have wrong concept of the Buddha-nature, by presenting the thought of the Uttaratantra. Especially, the 5th chapter of the Chinese version of the Uttaratantra, the chapter of ‘all sentient beings posses the Tathagatagrabha’(&#65378;一切衆生有如來藏品&#65379;), shows the way of recognizing the Buddha-nature by explaining the 3 significant meanings and the 10 points of view of the Tathagatagrabha. Accordingly, this study deals with the problem that there might be the miss understanding, based on the chapter of all sentient beings have the Tathagatagrabha, that the entity of the Buddha-nature realistically exists. Thus, as concerns about how to recognize the entity of the Buddha-nature, this study presents the guidelines to find the correct method of the recognition on the Buddha-nature with the help of the commentary, the Uttaratantra. And this study is aimed at proving the fact that the origin of the thought of the Buddha-nature is not in violation of Buddha's right teachings. In order to study on this subject, first, the original form of the Buddha-nature in other Buddhist text was examined. Secondly, it proceeded the study on the Uttaratantra's way of recognizing the Buddha-nature based on the 5th chapter of the Uttaratantra, the chapter of all sentient beings have the Tathagatagrabha. And, lastly, after studying on the critical Buddhist thinkers' wrong view of the Buddha-nature, this work expresses re-criticizing opinion of their view. Those scholars who assert the critical Buddhism tried to understand the thought of the Brahman and the thought of the Buddha as the simple antagonistic system of the monism and pluralism. They say that the thought of the Brahman which believes the one and only absolute being beyond the phenomenal world is the monism, and as the monistic thinking in the Mahayana philosophy is opposed to the Buddha's teachings, it is under the thought of the Brahman. Furthermore, they regard that the theories of the Tathagatagrabha and the innate enlightenment(本覺) are, far from the view of the Dependent Arising in Fundamental Buddhism, transformed into the monistical essentialism. However, the Buddha's thought of dependent arising transcends the antagonistic system of the monism and pluralism. The Buddha never accepted or denied neither monism nor pluralism, but insisted the middle path. The main issue in the view of the critical Buddhist thinkers are that they regard the Buddha-nature as the Dhatu-vada(基體說), and such view is refuted by the point of the ontology and the derivative manifestation. Besides, this work attempted to arrange the method of analysis and the different view of the definition of the Buddha-nature through the original language, and concluded the standpoint of the recognition of the Buddha-nature through the relationship of the time and space. As above, this study is focused on the thought of the Buddha-nature that deliberated in the Uttaratantra, and deals with the assertions of the critical Buddhist thinkers. The text of the Uttaratantra explains that one can realize the non-difference between the sentient beings and the Buddha through the right recognition of the Buddha-nature, and this shows the emphasis on the practice and faith by way of revealing the potentiality of being a Buddha.

B2B핀테크 서비스 사용자 수용성에 대한 연구

이석환 동국대학교 일반대학원 2022 국내석사

핀테크는 글로벌 ICT기업이 보유한 풍부한 고객층을 기반으로 결제 및 송금, 대출,자산관리 등의 다양한 금융서비스를 제공하는 혁신적인 기술이다. 2017년을 기준으로 국내 핀테크 거래는 B2C가 603조, B2B가 3,727조(기업간 거래 기준)로 B2B의 규모가 휠씬 더 크지만, 대부분의 핀테크 기업들은 B2C분야에 집중하는 경향을 나타내고 있다. 이에 본 연구에서는 기술수용이론을 바탕으로 B2B핀테크 서비스 사용자의 수용성에 대하여 연구를 진행하였다. 본 연구를 위해 핀테크 서비스, B2B/B2C 전자상거래 및 기술수용이론에 대한 이론적 배경을 바탕으로 연구모형을 설계하여 실증분석을 수행하였다. 먼저 독립변수로는 기술수용이론의 두 가지 신념인 인지된 유용성과 인지된 이용용이성을 설정하였다. 인지된 유용성은 신속성, 안전성, 경제성, 대량의 정보획득을 하위 범주로 구성하였고, 인지된 이용 용이성은 학습의 용이성, 업무수행의 용이성, 시공간의 제약 감소, 정보검색의 용이성을 세부 항목으로 구성하였다. 종속변수로는 수용의도를 설정하였고, 적극적인 이용, 다양한 활용 및 타인에게 추천의향 여부를 하위 항목으로 구성하였다. 본 연구를 통해 B2B핀테크 서비스에 대한 기대감이 매우 높다는 사실을 확인 할 수 있었으며 이에 대한 수용 의지 또한 높다는 사실을 확인할 수 있었다. 다만, 연구의 진행과정에서 참고할 만한 선행 연구가 부족하고 본 연구의 분석 결과를 검증할 방법이 없었다. 이러한 한계점에도 불구하고 본 연구는 B2B핀테크 서비스 사용자의 수용성에 대하여 처음으로 제시하였다는 점에서 의의가 있다. Fintech is an innovative technology that provides various financial services of payment and remittance, loans, and asset management based on the customer base of global ICT companies. As of 2017, B2B of KRW 3,727 trillion (based on inter-corporation transactions) is much larger in domestic fintech transactions compared to KRW 603 trillion for B2C, but most fintech companies tend to focus on the B2C field. Therefore, in this research, based on the theory of acceptance and use of technology, a study was conducted on acceptance of a B2B fintech service user. For this study, an empirical analysis was conducted by designing a research model based on the theoretical background of fintech services, B2B/B2C e-commerce, and theory of acceptance and use of technology. First, as independent variables, two beliefs of theory of acceptance and use of technology, perceived usefulness and perceived ease of use, were set. Perceived usefulness consisted of immediacy, safety, economy, and mass information acquisition into subcategories, and perceived ease of use consisted of ease of learning, ease of performing work, reduction of time and space, and ease of information retrieval. Acceptance intention was set as the dependent variable, and active use, various uses, and whether or not to recommend to others were composed of sub-items. Through this research, it was confirmed that there are high expectation for B2B fintech and volition to use this is also high. However, there was a lack of prior studies that could be referenced in the process, and there was no method to verify the analysis results of this study. Despite these limitations, this research is meaningful in that it was the first to propose acceptance of a B2B fintech service use.