Stereo vision and it's application to vehicle detection

Nguyen, Vinh Dinh Sungkyunkwan University 2012 국내석사

Stereo-vision-based vehicle detection and distance estimation (SDE) is a convenient method to use to build a warning system to support the driver. Many algorithms have been proposed based on SDE but they are very complex or cannot be used in a real-time system. In this thesis, two efficient algorithms are proposed for detection and distance estimation of a preceding vehicle. First algorithm is the fusion of knowledge based and stereovision based methods. This algorithm avoids noise from different objects appearing along the road. Moreover, the algorithm also works well when many vehicles appear on the road. The algorithm has been successfully tested under various road conditions and weather condition using two CCD cameras mounted on a vehicle. For 1570 experimental images, the average processing-time is approximate 25ms/frame, the maximum detection distance is 150m, the detection rate is 87.6% and the false detection rate is 5.86%. First algorithm can detect multiple vehicles at the same time but the accuracy cannot compete to current systems. Therefore, second algorithm is proposed. It is the fusion of stereovision and evolutionary algorithm (SEA). SEA utitilzes our recent work on FPGA implementation of census-based correlations for stereo matching. Then, SEA uses the gray scale left image and disparity information obtained from the FPGA system to detect the preceding vehicle and estimate its distance. This algorithm introduces an effective fitness function that allows our proposed method to have an improved performance and higher accuracy when compared with the existing evolutionary algorithm (EA) based methods. A new crossover types, tournament crossover, is introduced to reduce the convergence time of our proposed. This thesis also introduces a new approach for estimating the fitness function parameters. This estimation differ from the traditional EA because these parameters were generally created via experiments. Moreover, the processing time and accuracy of SEA can be improved by converting the global search to the local search with V disparity map. The robust experiments have proved that SEA successfully detect vehicles in front and sustain noise from different object appearing along the road. The detection range is 10m?140m, the detection rate is 95 % and the average processing time is approximately 31 ms/frame on CPU.

Knowledge-based Question Answering via Semantic Embedding Space

Min-Chul Yang 고려대학교 대학원 2015 국내박사

Semantic transforming a natural language question to its corresponding logical form is central to the knowledge-based question answering system. Most previous methods have tried to achieve this goal by using syntax-based grammar formalisms and rule-based logical inferences. However, such approaches are usually limited in coverage of the lexical trigger which performs a mapping task from words to logical properties of the knowledge base, and easy to ignore implicit and broken relations between the properties, not interpreting the full knowledge base. In this thesis, our goal is to answer questions in any domains via semantic embedding space in which the embeddings are encoded the semantics of words and logical properties. In the latent space, semantic associations between the existing features can be leveraged through their embeddings without using any hand-craft lexicon and rules. The embedding-based inference for question answering provides the ability to map factoid questions posed in natural language and the logical representations of the correct answers guided by the knowledge base. Our method is organized as follows: 1) semantic embedding space construction and 2) embedding-based question answering. The first stage involves learning low-dimensional embeddings of a word and a logical property, both of which are a semantic-associated pair extracted from unstructured textual data with distant supervision, so that vector representations of semantically similar features are close to each other in the semantic embedding space. That is, the meaning representations of words can be collaboratively specified using their relations with the logical properties as regards the knowledge base consisting of conceptual data in the hierarchical and multi-relational structure. The second stage focuses on ranking potential answers based on semantic similarities between embeddings of bag-of-words represented in the given question and those of logical representations of the potential answers. Here, the set of candidate answers is generated by the facts, as the knowledge base constituents, of possible entities appearing in the question statement. We then formulate an answer statement according to the expected answer type. In terms of the overall performance of question answering, experimental results and case examples demonstrate that our proposed method outperforms previous KB-QA baseline methods on a publicly released QA evaluation dataset: WebQuestions. 자연어 질문을 이에 상응하는 정규화 형태로 바꿔주는 의미적인 변환 과정은 지식 기반 질의 응답 시스템에서 중요한 부분이다. 최근 연구들은 이를 위해 구문 기반 문법적 형식화 기법이나 규칙 기반 논리적 추론 기법 등을 이용한 방법들을 제안하고 있다. 하지만 위 연구들은 어휘 단계 정규화 작업에서 다양한 어휘를 지식 베이스 내 논리적 자질로 연결하지 못하는 문제가 있고, 또한 지식 베이스 내 자질 간의 내포되거나 생략된 관계를 발견하지 못하므로, 지식 베이스를 충분히 파악하지 못하는 단점이 있다. 본 연구의 목표는 의미적인 임베딩 공간을 통하여 모든 분야의 질문에 대하여 답변하는 것이다. 여기서 임베딩은 단어들과 논리적 자질들의 의미적인 정보가 수치화된다. 임베딩 공간에서는 자질 간 의미적인 연관성이 사람에 의한 어휘 목록이나 규칙 없이도 자질들의 임베딩에 의해 계량화 된다. 이런 질의 응답을 위한 임베딩 기반 추론은 자연어로 표현된 주어진 질문과 이에 상응하는 답변의 지식 베이스로부터 제공되는 논리적 표현들을 연결시켜 줄 수 있다. 본 방법은 1) 의미적인 임베딩 구축과 2) 임베딩 기반 질의 응답으로 구성된다. 첫번째 단계는 의미적인 임베딩 공간에서 의미적으로 유사한 단어와 논리적 자질의 저차원 임베딩이 서로 비슷한 값을 가진 벡터가 되도록 학습한다. 의미적으로 연결된 단어와 논리적 자질 쌍은 distant supervision 기법을 이용하여 비정형 텍스트 데이터로부터 자동으로 추출된다. 즉, 지식베이스는 계층적이고 다단계로 구조화 된 개념을 표현하는 데이터로 구성되어 있기 때문에, 단어들의 의미적인 표현은 이들의 논리적 자질 간 관계를 통해 협력적으로 구체화된다. 두번째 단계는 주어진 질문을 표현하는 단어 묶음들의 임베딩들과 후보 답변의 논리적 표현들의 임베딩들 간 의미적 유사도를 통해 답변들을 순위화하는데 초점을 둔다. 여기서 후보 답변들의 집합은 질문에서 나타나는 가능한 개체들의 지식 베이스 정보에 의해 생성된다. 그 다음, 예상되는 답변 유형에 따라 답변을 표현화한다. 질의 응답 성능 관점에서, 본 논문에서 보고하는 실험 결과와 구체적인 사례 등은 본 연구가 기존의 지식 기반 질의 응답 연구들보다 효과적임을 증명하고 있다. 성능 평가는 공개된 질의 응답 평가 데이터인 WebQuestions를 이용하였다.

서양음악사 web based 학습 system을 위한 prototyping 평가에 관한 연구

강정경 東國大學校 經營大學院 2000 국내석사

컴퓨터의 비약적인 발전은 인간과 각종 기술뿐만 아니라 교육분야에도 일대 변화 양상을 맞이하는 계기가 되고 탈 산업사회, 정보화 사회, 다원화, 세계화 등의 사회적 변화의 요구에 발맞추어 교육도 진보된 교육으로 변화가 시도 되고 있다. 본 논문에서는 Web Based Instruction의 프로그램 가운데 서양음악사 Web Based Instruction를 구현해 실제 현장 수업에 적용해 봄으로써 학습효과를 분석하는데 밝히는데 있다. 본 연구는 서양음악사 웹학습 시스템 프로토타이핑에 관한 평가로 본인이 설계한 프로그램을 가지고 O 여자 정보고등학교 3학급을 대상으로 실제로 수업한후 설문조사하였다. 설문의 결과는 다음과 같다. 첫째, 본 연구에서 구현한 서양음악사 Web Based학습 System의 교육 내용의 하위변인 중 교육내용의 타당성은 학습 효과에 영향을 미치는 것으로 분석되었다. 둘째, 본 연구에서 구현한 서양음악사 Web Based학습 System의 교육 내용의 하위변인 중 교육내용 구성의 합리성과 교육내용 구성의 명확성은 학습 효과에 영향을 미치는 않는 것으로 분석되었다. 셋째, 본 연구에서 구현한 서양음악사 Web Based학습 System의 학습의 이성의 하위변인 중 피드백 제공 정도와 동기유발은 모두 학습 효과에 영향을 미치는 것으로 분석되었다. Making great strides in computer technology develop an opportunity for receiving a big change in the educational field as well as for mankind and technologies of various kinds. The informationized soiety, plual society, globalization, and to get out of industrial society, the educational field is also trying to pursue a progressive education keeping pace with a demand of social change. In this research, among the Web Based Instruction program, the Western Music History program is embodied and applied through on-line teaching, through which the learning effect is more clearly analyzed. This research to assess the performance evaluation of the Western Music History Web Learning System Prototype. The survey was design to analyzes. The results of this research were as follows. First, from this research, the lowest variation of the educational contents of the Western Music History Web Based learning system analyzed the validity of the educational contents giving influence to the learning effect. Second, from this research, the lowest variation of the educational contents of the Western Music History Web Based learning system analyzed the clarity and rationality of the construction educational contents not giving influence to the learning effect. Third, from this research , the lowest variation of the faciliation of application to the Western Music History Web Based Learning System analyzed to what extent the Feedback and Motivation gave to influencing the learning effect.

음악개념 기반 음악교육에 대한 장애영유아 담당교사의 인식과 실제

윤혜진 창원대학교 2018 국내박사

Music can be a good tool for young children with disabilities to express themselves. Like general infants, young children with disabilities like to sing with their children of their age and enjoy body expression along with music. Young children with disabilities can learn the factors of music concept, such as rhythm, dynamic, beat, melody, tone, pitch, and tempo, through the music activities in nursing facilities, and the teacher of young children with disabilities who teach them should have actual experience and training of music concept. This research established following research problems, aiming to analyze difference and relation between the cognition of teachers who take care of young children with disabilities, who are in a decisive period of musical development, on music concept based education, and the actual level. First, what is the cognition of teacher of young children with disabilities on music concept based music education and the actuality? Second, is there any difference in the actual level according to the young children with disabilities teachers’ level of cognition on music concept based music education? Third, what is the relation between the cognition on music concept based music education and the actual level, and what is the influence of cognition on music concept based music education on the actual level? To settle the research problems, this research conducted a survey targeting 250 teachers who take care of young children with disabilities at the kindergartens and Day care centers in Gyeongnam. This research performed frequency analysis, descriptive statistics, reliability analysis, independent sample t-test and F-test, correlation, and simple regression analysis using SPSS program. Based on the analysis result, this research derived following conclusion. First, cognition level of teacher of young children with disabilities on music concept based music education was high, however the level of actual application was low, they focused on melody, beat, and tempo among the components of music concept, particularly, they preferred repeated melody class, regular beat lesson, and comparison of quick/slow tempo. Therefore, it is necessary to suggest balanced music education based on various components of music concept such as rhythm, dynamic, melody, rhythm, tone, harmony, pitch, tempo, and form, to develop their musical balance by improving music education based on the music concept of young children with disabilities education, which are currently focused with melody, beat, and tempo. Second, a difference appeared between the level of cognition on music concept based music education and the actual level. This implies a possibility that the actual level could be high when the cognitive level in the components of music concept based music education is high. Accordingly, it is necessary to reinforce practical field training program to share the knowledge and experience of music concept based music education targeting the infant education teachers. Third, a significant relation appeared between the cognition of music concept based music education and the actual level, and the cognition of music concept based music education displayed a significant and positive influence. Since the infants express their emotion through music, it is necessary to develop music concept activity teaching plan to let them express themselves utilizing various methods by intensively extracting the music concepts such as rhythm, dynamic, melody, beat, tone, harmony, pitch, tempo, and form in music. This research is significant in respect that it promoted infant education teachers’ understanding of music concept based music education by investigating their cognition on music concept based music education of teacher of young children with disabilities and the actual level, and prepared a basis to develop various teaching methods applied with music concept, also, it suggested the material for the understanding of the actuality of teaching activity of music concept based music education through analysis of difference of the actuality of music concept based music education according to teacher of young children with disabilities’ music experience and characteristic and for the efficient teacher training.

Study of highly sensitive and reproducible paper-based SERS sensors for detection of trace pesticides

이민우 서울대학교 대학원 2018 국내박사

미량의 농약 검출을 위한 효율적인 방법 중에 하나로 각광받는 필터 페이퍼 기반의 표면 증강 라만 산란 (SERS) 센서는 지난 십수년간 연구되어 왔다. 그러나 필터 페이퍼의 친수성 성질에 의해 용액들은 종이 표면에 고정되지 않고 빠르게 종이 내부로 흡수되게 될 뿐만 아니라 종이를 구성하는 셀룰로오스 섬유의 구조적, 크기적 다양성에 의해 많은 수의 공극과 높은 종이 표면 단차를 보여주고 있다. 이러한 종이의 특성으로 인해 낮은 종이 기반의 SERS 센서의 감도와 재현성이 나타났다. 또한 분자와 금속 사이의 결합력 차이에 의해 일부의 분자만 검출이 가능하며 다양한 분자를 검출할 수 있는 범용적 센서로는 사용이 어려웠다. 본 연구에서는 이러한 기존의 종이의 특성에 의해 발생하는 단점을 극복하고 고감도, 고재현성의 SERS 신호를 내는 필터 페이퍼 기반 SERS 센서를 개발하였다. 우선 필터 페이퍼 기반의 SERS 센서의 감도를 높이기 위하여 알킬케텐다이머를 (AKD) 종이에 처리하여 필터 페이퍼의 친수성 성질을 소수성으로 개질하였다. 또한 소수성 처리된 필터 페이퍼 표면에 셀룰로오스 나노섬유를 (CNF) 도입하여 종이 표면의 단차 및 공극을 줄여 필터 페이퍼 기반 SERS 센서의 표면의 균일성을 증가시켰다. 이를 기반으로 필터 페이퍼 기반의 SERS 센서의 재현성을 높이고자 하였다. 마지막으로 기존에 검출이 어려웠던 극성 유기 분자의 검출이 가능한 필터 페이퍼 기반의 SERS 센서를 개발하기 위해 서로 다른 전하를 가지는 나노입자를 도입하여 전하 선택적 검출이 가능한 필터 페이퍼 기반의 SERS 센서를 제작하였고 이를 통해 SERS 기반의 분자 검출의 보편성을 확장하고자 하였다. 그렇기에 본 연구는 크게 소수성 처리를 통한 고감도 필터 페이퍼 기반 SERS 센서 제작, 셀룰로오스 나노섬유의 도입을 이용한 고재현성을 나타내는 필터 페이퍼 기반 SERS 센서 제작, 마지막으로 극성 유기 분자 검출을 위한 전하 선택적 검출을 도입한 필터 페이퍼 기반 SERS 센서를 제작 및 적용 이렇게 세 파트로 구성되어 있다. 챕터 1에서는 소수성으로 개질된 필터 페이퍼의 개발에 대해서 다루고 있다. 기존의 종이 기반의 SERS 센서는 친수성의 성질로 인해 용액이 종이 내부로 빠르게 흡수되어 종이 표면에 나노입자가 거의 존재하지 않았으며 또한 분석 물질이 나노입자와 결합하기에 시간이 충분치 않아 낮은 SERS 신호 민감도와 재현성을 나타냈다. 이를 개선하기 위해 종이에 AKD 처리를 통해 소수성 성질을 나타내도록 하였다. 소수성 개질을 통해 종이 표면에서 수용액의 접촉각과 유지 시간이 증가하였으며, 증가된 접촉각으로 인해 용액의 접촉 면적이 감소하였다. 이로 인해 종이 표면에 은 나노입자 (AgNP) 의 밀도가 증가하게 되어 수많은 SERS hot-spot 을 형성하였고, 이로 인해 SERS 신호의 급격하게 증가하였다. 또한 AgNP 용액의 농도를 조절하여 필터 페이퍼 위의 AgNP 분포를 조절하였으며, 이를 통해 SERS 신호의 민감도와 재현성을 최적화 하였다. 이를 통해 개발한 종이 기반의 SERS 센서는 6.19%의 SERS 신호의 표준 편차를 나타냈으며 thiram 과 ferbam 이 각각 0.461 nM 과 0.491 nM 의 검출 한계를 나타냄을 확인하였다. 이 결과를 통해 종이의 소수성 처리를 통해 저렴하고 쉬운 방법으로 미량의 농약을 높은 재현성으로 검출 할 수 있는 필터 페이퍼 기반의 SERS 센서를 개발 할 수 있음을 확인하였다. 챕터 2에서는 CNF 코팅 도입을 통해 종이 표면의 단차 및 공극을 줄여 종이 기반의 SERS 센서의 신호 재현성을 증가시키는 것에 대해 다루고 있다. 많은 수의 공극과 높은 표면 단차에 의해 종이 표면은 불균일한 특성을 나타내며 이로 인해 기존의 종이 기반의 SERS 센서는 낮은 SERS 신호 재현성을 나타내고 있다. 이를 극복하기 위해 CNF 코팅을 도입하여 종이 표면의 공극을 채우고 표면 단차를 줄이고자 하였다. 두 번의 CNF 코팅의 도입에 따라 필터 페이퍼의 표면에서의 AgNP 커버리지가 87% 에서 95% 까지 증가됨을 확인하였으며, CNF 코팅 도입에 따른 종이 표면의 표면 단차 감소 및 공극의 수 감소로 인해 AgNP 가 균일하게 도입됨을 FE-SEM 이미지 분석을 통해 확인할 수 있었다. 이를 통해 두 번의 CNF 코팅을 필터 페이퍼에 도입함에 따라 SERS 신호의 상대 표준 편차는 28%에서 9%로 감소하였으며 4-ATP 의 검출 한계는 3.782 nM 에서 0.426 nM 로 감소하는 것을 확인하였다. CNF 코팅 도입에 따른 표면 균일성 증가는 SERS 신호 재현성 뿐만 아니라 신호 민감도 까지 증가시킴을 확인하였고 이는 종이 기반의 SERS 센서의 신호 재현성의 증가에 기본 연구가 될 수 있음을 확인하였다. 마지막으로 챕터 3 에서는 극성 유기 분자를 검출 할 수 있는 전하 선택적 종이 기반의 SERS 센서 개발에 대해서 다루고 있다. 분자에 있는 작용기에 따라 분자와 금속은 서로 다른 친화도를 나타내며, 이로 인해 기존의 종이 기반의 SERS 센서는 특정한 분자만 검출 가능하였다. 이를 극복하고 SERS 기반의 분자 검출의 보편성을 확대시키기 위해 본 연구에서는 전하 선택적 검출이 가능한 종이 기반의 SERS 센서를 제작하였다. 이를 위해 서로 다른 전하를 가지는 AgNP를 필터 페이퍼에 도입하여 전하 선택적 종이 기반의 SERS 센서를 개발하였다. 기존의 AgNP은 시트르산 이온에 의해 음전하를 나타내며, 이러한 AgNP 표면에 양전하를 띄는 고분자인 poly(allyldimethylammonium chloride) (PDDA)을 도입한 AgNP@PDDA를 개발하여 양전하를 나타내는 나노 입자를 개발하였다. 서로 다른 전하를 나타내는 AgNP와 AgNP@PDDA를 종이 표면에 도입하여 전하 선택적 검출이 가능한 종이 기반의 SERS 센서를 개발하였다. AgNP에 PDDA 고분자가 도입됨을 UV/Vis 스펙트럼, 제타 전위 변화 및 고해상도 투과 전자 현미경을 이미지 변화를 통해 확인하였다. 제작한 센서의 전하 선택적 검출의 적용 가능성을 확인하기 위해 전하를 띄는 라만 염료를 각각의 센서에 적용하였으며 AgNP 가 처리된 센서에서는 양전하를 띈 라만 염료가, AgNP@PDDA 가 처리된 센서에서는 음전하를 띈 라만 염료가 검출됨을 확인하였고, 이는 나노 입자와 전하를 띄는 분자 사이에 작용하는 정전기적 인력에 의해서 나타남을 확인하였다. 이를 통해 개발된 전하 선택적 종이 기반의 SERS 센서의 전하 선택적 분자 검출 적용 가능성을 확인하였고, 극성 유기 분자의 검출 확인을 위해 아닐린과 벤조산 유도체를 도입하였다. 각각의 극성 유기 분자는 AgNP 및 AgNP@PDDA 센서에서 분리 검출됨을 확인되었으며, 아닐린과 벤조산이 수십 마이크로 몰 농도의 수준까지 검출됨을 확인하였다. 또한 10% 이하의 상대 표준 편차를 나타냄을 확인하였으며 이를 통해 개발한 전하 선택적 종이 기반의 SERS 센서를 이용하여 기존의 검출이 어려웠던 분자를 고감도, 고재현성 검출이 가능함을 확인하였다. 본 연구에서 제시한 종이 기반의 SERS 센서 개발은 종이 기반의 SERS 센서의 기초 연구로서 다른 연구의 길잡이가 될 수 있을 것이라 생각한다. As a cost-effective approach for detecting trace amounts of pesticides, paper-based surface-enhanced Raman scattering (SERS) sensors have been the subject of intensive research. However, one of the hurdles to overcome is the difficulty of retaining nanoparticles on the paper surface due to the hydrophilic nature of the cellulose fibers in paper. Furthermore, another hurdle is high surface roughness and non-uniformity of the paper surface due to the size variation and conformation of cellulose fibrils in paper. These hurdles reduce the sensitivity and reproducibility of paper-based SERS sensors due to the low density of nanoparticles, short retention time of analytes and non-uniform surface by many pores and high surface roughness on the paper surface. Furthermore, conventional SERS sensors have difficulty to use as universal SERS sensors due to difference of binding affinities between molecules and metal nanoparticle depending on functional groups of molecules. In this study, a highly sensitive and reproducible filter paper-based SERS sensor was developed. To increase the sensitivity and reproducibility of the sensor, the hydrophilic nature of the filter paper was modified to become hydrophobic one by using alkyl ketene dimer (AKD) treatment. In addition, cellulose nanofibril (CNF) coating was applied to the hydrophobic filter paper to increase the uniformity of the paper surface by filling pores and reducing the surface roughness of the filter paper. Finally, using CNF coated hydrophobic filter paper, a charge-selective paper-based SERS sensor was developed to detect polar organic pollutants for expanding the generality of SERS based molecular detection. Thus, this thesis contains three chapters that focus on: fabricating a highly sensitive and reproducible filter paper-based SERS sensor by hydrophobic modification, developing a uniform surface for the filter paper-based SERS sensor for highly reproducible SERS detection by introducing CNF coating, and applying the developed paper-based SERS sensor for fabricating of a charge-selective SERS detection to expand the generality of SERS based molecular detection. Chapter I describes the fabrication of hydrophobically modified filter paper to increase sensitivity and reproducibility of paper-based SERS sensor. Conventional paper-based SERS sensors quickly absorb nanoparticle and analyte solutions because of their hydrophilic nature. In addition, the short retention time of the analyte solution on the paper surface provides insufficient time for the analyte to bond on the surface of the nanoparticle. Thus, conventional paper-based SERS sensors show low sensitivities and reproducibilities. To overcome disadvantages of conventional paper-based SERS sensor, the surface of filter paper was modified hydrophobically by introducing AKD on filter paper. By introducing AKD on filter paper, the retention time of the silver nanoparticle (AgNP) and analyte solutions on the paper surface was increased because the AKD treatment changed hydrophilic nature of filter paper to hydrophobic one. The AKD treatment increased the contact angle of the aqueous AgNP solution, which consequently increased the density of AgNP on the paper-based SERS sensor within reduced contact area. In addition, the retention time of the aqueous solution was increased by preventing its rapid absorption into the filter paper, and the AgNP solution was dried on the paper surface without absoprtion to the filter paper. As a result, because the increased density of AgNP on a small contact area on hydrophobic filter paper, the number of increased SERS hot-spots, and strongly enhanced the SERS signal. The sensitivity and reproducibility of the SERS signal were optimized by controlling the distribution of AgNP on the surface of the filter paper, which was achieved by adjusting the concentration of the AgNP solution. The spot-to-spot variation of the SERS intensities of 4-aminothiophenol (4-ATP) at 25 AgNP spots on hydrophobic filter paper-based SERS sensor was approximately 6.2% of relative standard deviation (RSD), and the limits of detections (LODs) of thiram and ferbam were 0.461 and 0.491 nM, respectively. These proof-of-concept results indicate that this low-cost and easily fabricated paper-based SERS sensor can provide highly sensitive pesticide detection. Chapter II describes reducing the surface roughness and pores of the filter paper-based SERS sensor using CNF coating for highly reproducible SERS detection. Because of non-uniform filter paper surface originated from the large number of pores and high surface roughness, conventional paper-based SERS were difficult to measure uniform SERS signals, resulting in low reproducibility. To overcome the low reproducibility of paper-based SERS sensors, CNF coating was introduced on the surface of filter paper to fill the pores and to flatten the surface of filter paper. Double CNF coatings on the surface of the hydrophobic filter paper increased the coverage of AgNP on the paper surface from 87 to 95%. Furthermore, the AgNP were uniformly introduced onto the surface of the filter paper by reduced surface roughness and the number of pores on the paper surface were redcued, as confirmed using field-emission scanning electron microscopy (FE-SEM). Applying double CNF coatings on hydrophobic filter paper-based SERS sensor reduced the RSD of the SERS intensity from 28 to 9% and the LOD of 4-ATP from 3.782 to 0.426 nM. These CNF surface modifications on paper-based SERS sensor provided a base to fabricate a highly reproducible paper-based SERS sensor. Chapter III describes the development of a paper-based SERS sensor for detecting polar organic pollutants to expand the generality of SERS based molecular detection. Conventional paper-based SERS sensors are difficult to use as universal SERS sensors because the functional group in molecule have different affinity toward the metal. To expand the generality of paper-based SERS sensor, a charge-selective paper-based SERS sensor was developed by modifying the surface charge of AgNP. The citrate ion capped AgNP were exhibited negative surface charge due to negative charged citrate ion and attracted positively charged molecule by electrostatic attraction. However, it was hard to detect negatively charged molecule by electrostatic repulsion. To detect negatively charged molecules, the positively surface charged AgNP was fabricated by modifying the surface charge of AgNP by applying poly(diallyldimethylammonium chloride) (PDDA) on AgNP surface. The PDDA encapsulated AgNP (AgNP@PDDA) exhibited positive surface charge, and attracted negatively charged molecule. The introduction of PDDA on AgNP was confirmed by zeta-potential change and high resolution transmission electron microscopy (HR-TEM) images. Charged Raman dyes were applied to the charge-selective paper-based SERS sensor to verify its feasibility. By using charge-selective paper-based SERS sensor, positively charged Raman dyes were successfully detected using the AgNP spots, and negatively charged Raman dyes were successfully detected using the PDDA encapsulated AgNP (AgNP@PDDA) spots by electrostatic attraction.. To expand its applicability for field test, polar organic pollutants (e.g., aniline and benzoic acid derivatives) were detected using the charge-selective paper-based SERS sensor. Aniline and benzoic acid derivatives were detected by AgNP and AgNP@PDDA spots, respectively. Furthermore, aniline and benzoic acid were detected at low concentration of tens of micromolar concentration with a less than 10% RSDs. Thus, the developed charge-selective paper-based SERS sensor could expand the generality of SERS based molecular detection with a high sensitivity and reproducibility by modifying the surface charge of the AgNP.

Chemical modification of biomass materials for functional bio-based plastics

권혁준 Graduate School, Yonsei University 2023 국내박사

Chemical modification is a technology capable of controlling physical and new chemical properties through an intentional change of materials. For chemical modification, it is necessary to establish a modification method according to the material, and it is essential to study the characteristic behavior according to the intention. Recently, various studies have been attempted with the goal of carbon zero, and the use of carbon-neutral biomass is increasing. Cellulose, representative biomass, is classified as an infinite resource whose annual production is similar to the total oil reserves and is not likely to be depleted if adequately used. Therefore, the interest in cellulose-based materials is significantly increasing, considering the harvestable cycle and non-edible resources. Since cellulose-based materials have different chemical reaction properties depending on their constituents, it is necessary to understand the materials used for modification and the reaction mechanism. Accordingly, this paper introduces the intended functional group through chemical modification using various cellulose-based materials, and the characteristic behavior and mechanism are studied. Hydrocarbon chains are introduced through chemical modification into crystalline cellulose and chitosan induced phonon scattering in the polymer matrix. As a result, the modified chitosan (Chito-D) had lower thermal diffusivity than the control polypropylene (PP) when 1 wt% complexed. Due to the scattering of phonons, it was possible to control the thermal conductivity of the bio-based plastic in which the cellulose-based material, including the crystal structure, was complex. In addition, the cellulose-based adsorbent modified with an amine functional group exhibited carbon dioxide gas adsorption properties in a solid state. And the morphology of the material was changed according to the method of chemical modification, and the amount of carbon dioxide gas adsorption could be controlled. Although it has a similar content to amine groups, further increasing the adsorption amount was induced by controlling the morphology according to the reforming reaction. 화학적 개질은 재료의 의도적인 변경을 통한 물리적 특성 뿐만 아니라, 새로운 화학적 특성 제어가 가능한 기술이다. 이러한 화학적 개질은 제어하고자 하는 재료에 따라 개질 방법을 확립할 필요가 있으며, 의도에 따른 특성 거동에 대한 연구가 필수적이다. 최근에는 탄소 제로를 목표로 다양한 연구가 시도되고 있고, 탄소 중립의 바이오매스에 대한 활용이 증가하고 있다. 바이오매스의 대표적인 셀룰로오스는 지구상에서 1년간 생산량이 석유의 전체 매장향과 비슷한 수준으로, 적정하게 이용하면 고갈될 염려가 없는 무한자원으로 분류되고 있다. 그렇기 때문에, 수확 가능한 주기 및 비식용계 자원 등을 고려한 셀룰로오스 베이스 소재의 활용에 대한 관심이 크게 증가하고 있다. 이러한, 셀룰로오스 베이스 소재는 구성 성분에 따라 화학적 반응 특성이 달라지기 때문에, 개질에 사용되는 재료와 반응 메커니즘에 대한 이해가 필요하다. 이에 따라, 본 논문에서는 다양한 셀룰로오스 베이스 소재를 활용하여 의도된 기능기를 화학적 개질로 도입하고, 특성 거동과 메커니즘에 대해 연구하고자 한다. 결정질을 갖는 셀룰로오스와 키토산에 화학적 개질을 통해 도입된 탄화수소 사슬은 고분자 매트릭스 내에서 포논 산란을 유발하였다. 결과적으로는 개질된 키토산(Chito-D)가 1 wt% 복합화되었을 때는 대조구 폴리프로필렌(PP)과 비교해서 더 낮은 열 확산성을 가졌다. 이러한 포논의 산란으로 인해, 결정구조를 포함한 셀룰로오스 베이스 소재가 복합화된 바이오 베이스 플라스틱의 열전도도를 제어할 수 있었다. 또한, 아민 작용기를 개질한 셀룰로오스 베이스 흡착제는 고상의 조건에서 이산화탄소 가스 흡착 특성을 발현했다. 그리고, 화학적 개질의 방법에 따라 소재의 morphology 변형을 유발하였고, 이산화탄소 가스 흡착량을 제어할 수 있었다. 유사한 아민기의 함량을 갖지만, 개질 반응에 따라 morphology 제어를 통해 흡착량을 더 증가시키는 결과를 유도하였다. 이렇게 기능기 유도한 셀룰로오스 베이스 소재를 고분자와 복합화하여 향후 생분해 플라스틱 또는 바이오 베이스 플라스틱으로의 어플리케이션 가능성을 확인하였다. 화학적 개질된 바이오 베이스 소재는 고분자 복합화 후에도 기능성을 발현하였으며, 식품 포장재로서의 활용이 가능함을 보여줬다.

한국어 학습자의 어휘추측전략 분석 연구 : 한자권-비한자권 중·고급 학습자를 대상으로

오은희 연세대학교 교육대학원 2009 국내석사

본 연구는 읽기를 통한 어휘 학습에서 한자권 한국어 학습자와 비한자권 한국어 학습자가 모르는 어휘가 있을 때 사용하는 어휘추측전략 유형과 활용 양상을 밝히는 것을 목적으로 한다.1장에서는 본 연구의 필요성과 목적을 기술하고, 한국어교육에서 어휘학습전략에 대한 앞선 연구들을 검토하였다.2장에서는 어휘지식과 어휘학습, 어휘학습전략과 어휘추론전략을 정의함으로써 연구의 이론적 배경을 구성하고 어휘교육의 방향에 대해 재고하였다.3장에서는 연구의 대상과 기간, 절차, 자료의 수집과 분석 방법에 관해 기술하였다.4장에서는 3장의 연구 방법을 통해 얻어진 자료를 분석하여 한자권-비한자권 한국어 학습자의 어휘추측전략을 분석하였다. 첫째, 한자권-비한자권 학습자들의 어휘추측전략 유형을 밝힌다. 어휘추측전략 유형은 어휘의 의미소 분석하기, 유사 발음 활용하기, 주변 문법 및 어휘 활용하기, 문맥 활용하기, 어휘 구조 활용하기, 경험과 배경지식 활용하기, 직관 활용하기 전략으로 총7가지로 정리하고 실제 학습자들의 추측전략 사용의 예를 제시하였다. 둘째, 한자권-비한자권 학습자들의 어휘추측전략 활용 양상을 비교, 분석한다. 전략 활용 양상의 비교 분석은 전략 활용 범위, 추측전략 유형별, 어휘 유형별, 추측 성공률을 통해 확인한다. 학습자들은 두 집단 모두 문맥 활용하기 전략의 빈도가 가장 높았다. 그 다음으로 사용된 전략으로 한자권 학습자는 유사 발음 활용하기 전략, 비한자권 학습자는 어휘의 의미소 분석하기 전략이다. 그리고 한자권 집단은 75.13%, 비한자권 집단은 56.95%로 두 집단 모두 높은 성공률을 보여 전략 사용의 효용성을 확인하였다. 셋째, 어휘추측전략에 대한 학습자들의 의견을 밝힌다. 38명의 학습자들이 어휘추측전략을 사용하는 것이 어휘학습 방법으로서 긍정적이라고 답하였다.5장에서는 결과를 바탕으로 한국어 어휘추측전략 수업 방안을 제시하였다.6장에서는 연구의 결과를 요약하고, 본 연구의 제한점과 후속 연구에서 고려해야 할 점을 제시하였다.본 연구는 한자권 학습자와 비한자권 학습자의 전략 사용 양상과 특징을 비교하여 봄으로써, 어휘추측전략 지도 시 각 집단의 특성을 인식하고, 효율적으로 지도할 수 있도록 한 점에서 의의를 찾을 수 있다. This study is aim to show the type and use of inferencing vocabulary strategy when there exist vocabularies Chinese letter based- Non Chinese letter based Korean learners don't recognize.Chapter 1 describes the necessity and purposes of this study and provides a review of prior studies on Korean vocabulary acquisition strategies.Chapter 2 defines the concepts of vocabulary learning, vocabulary learning strategies and inferencing strategies, and proposes an alternative direction for teaching vocabulary by investigating the concept of word knowledge.Chapter 3 describes the objectives, period and procedures, data collect and analysis of this study.In chapter 4, I analyzed the inferencing vocabulary strategies of Korean learners through the data based on chapter 3. First, I argue the type of strategy on inferencing vocabulary of Chinese letter based- Non Chinese letter based Korean learners. For this process, I present 7-type of strategies on inferencing vocabulary with examples of the actual use of learners. The 7-type of strategies are analysis of the morpheme of vocabulary, the use of similarities between pronunciations, the use of grammar and vocabulary adjacent , the use of context, the use of vocabulary structure, the use of experience and background knowledge, the use of sense. Second, I analyzed the aspect of the use of inferencing vocabulary strategy of Chinese letter based- Non Chinese letter based Korean learners and compare them each other. The comparison and analysis of the aspect of the use of inferencing vocabulary strategy include range of the use of inferencing strategy, type of inferencing strategy, type of vocabulary and inferencing success rate. All groups make the most of the use of context strategy, then Chinese letter based learners use 'the use of similarities between pronunciations' and Non Chinese letter based Korean learners use 'analysis of the morpheme of vocabulary'. In inferencing success rate, Chinese letter based learners got 75.13%, Non Chinese letter based Korean learners got 56.95%. It means that when dealing with vocabularies, the use of various strategies is useful. Third, I show the Korean learners's opinions. 38 Participants recognized the usefulness of a vocabulary inferencing strategy for vocabulary learning and responded that the lessons were helpful.Chapter 5 provides a teaching plan from this study's outcome for Korean teachers.Chapter 6 summarizes the findings from the study, discusses its limitations and provides suggestions for future studies.

A Study on the Development of Middleware Architecture for Context-Aware System : A Study on the Development of Middleware Architecture for Context-Aware System

로비율 광운대학교 2016 국내박사

A context-aware system adapts itself according to the changing contexts of ambient environment and provides service to the user based on available contexts. Context acquiring, processing, reasoning, and disseminating to the services are complex tasks for a context-aware system. An appropriate middleware architecture could handle such complexities. In this paper, we proposed a middleware architecture for a context-aware system. Here, the context is modeled based on ontology using Web Ontology Language (OWL). The middleware architecture consists of a three layer structure. The sensing and managing layer reads sensing data from environment as well as controls necessary appliances to carry out the services. The processing layer generates low-level contexts. It also computes high-level contexts by using an effective reasoning mechanism. The service composition layer makes service commands to provide services based on available contexts. Since it is not possible to collect all the required contexts through the surrounding sensors, an effective reasoning mechanism is needed for the context-aware system. Generally, context reasoning is adopted to deduce highlevel context based on the available other contexts. The rule-based reasoning is one of the most well-known methods for reasoning context to build a contextaware system. However, it is difficult for the rule-based algorithm to reason personalized context because it requires a large number of personalized rules to apply the user’s preferences. To overcome this weakness, we proposed the Profile-Applied Reasoning Engine (PARE) for the middleware architecture. PARE is an enhanced rule-based reasoning method which uses user profile while reasoning contexts. By using profiles, PARE can become aware of the context that is preferred by a specific individual. In this system, all available current contexts are stored in the working memory. Whenever a change is detected in working memory, PARE begins to find proper match of each rule with current contexts in the working memory. If a user has several choices in a context then the PARE picks up the preferences from that user’s profile and puts into the matching rule for better reasoning outcome. We demonstrated the middleware architecture as well as reasoning algorithm in a smart home domain to provide context-aware services. To validate the effectiveness of the proposed reasoning algorithm, we performed a comparison between proposed and traditional rule-based reasoning using a real data-set. The experiment result shows that proposed algorithm has a better outcome for reasoning personalized context than the traditional rule-based reasoning. In addition, by using profiles, a significant number of rules have been omitted and consequently the runtime is also decreased. Moreover, PARE occupies less memory which is restricted with number of variables of a rule. Therefore, PARE optimizes both runtime and memory space, which is valuable when making embedded context-aware system. In this middleware architecture, context-aware service is also selected using the rule-based algorithm, where the service can be extended easily by adding new service rules in the service rule base.

기독교 대안학교에서 기독교적 공동체 역량 함양을 위한 프로그램 개발: 프로젝트 학습을 중심으로

이 다니엘 아신대학교 2022 국내석사

ABSTRACT A Study on the Development of a Program for Cultivating Christian Community Competence in Christian Alternative School: Focusing on Project-based Learning Lee, Daniel Master of Education The Graduate School of Education Asia United Theological University The purpose of this study is to investigate and develop a Program for Cultivating Christian Community Competence. For this purpose, this study examined the concept of community competencies among the core competencies presented in the Korean national curriculum(2015 Revised Curriculum), future core competencies announced by the OECD DeSeCo and CDC, ATCS21s compared Christian community competencies. And this study designed a project-based learning program that can strengthen those competencies. The contents of the study for developing a Program for Cultivating Christian Community Competence are as follows. 1. Extract the concept and sub-elements of Christian community competence. 2. Identify the principles, process, components of Project based Learning. 3. Develop a project based learning program to strengthen Christian community competencies. First, this study examined the meaning and subcategories of community competencies presented in the Korean national curriculum(2015 Revised Curriculum) and future core competencies announced by the OECD DeSeCo. This researcher also re-conceptualized them based on the concepts and categories of Christian community competencies that were studied by various researchers. Next, this study examined the principles, process and components of a project-based learning to enhance Christian community competencies. The principles of Project-based learning were aimed at solving a complicated problem related to life by cooperating with relationship, investigating method to solve problems, fucosing on learner, keep giving feedback. The processes of Project-based learning were five procedures: “choosing a topic”, “planning project”, “starting project”, “doing project activities”, “finishing project(feedback project)”. In addition, in terms of the components, this study compared the competency elements of project based learning with the sub-elements of Christian community competencies. After that, a school-level program was designed according to project-based learning named “The Village Project” and this study presented example project the <Walk into the world>. It was applied to Christian Alternative Elementary School. Through this study, the following concluding thoughts were found. First, Christian education and competency-based education can be interpreted in a similar context in that they aim to enable them to practice and utilize what they have learned with a real life tendency. Second, Christian community competencies were a complex ability. It was difficult to learn through several short-term classes. Therefore, a program designed for Christian community competency education was needed. Third, “The village project” program expects students to cultivate Christian community capabilities to recognize community problems and actively participate in expanding the kingdom of God. To strengthen Christian community competencies, this researcher would like to make the following suggestions: first, since this study is just a plan for a program, it is necessary to examine the development of community competency after actual implementation. Second, this program should be conducted at the school level. Third, the evaluation of this program should also focus on the process of solving complex problems. Forth, this program needs a leeway for the outcomes, because this program focuses on the harmony and participation of students and the growth of initiative in the community. Finally, Christian alternative schools should pay attention to the ever-changing national curriculum. Also efforts to reconstruct core competencies from a Christian point of view should continue.

A Study on the Application of Story Based Teaching in an Ideological and Political Theory Course in Chinese University

양검홍 세한대학교 일반대학원 2022 국내박사

Abstract This study aimed to develop a story based teaching suitable for ideological and political theory courses in Chinese universities and confirm its teaching application ef ect. Therefore, this study investigates the relevant theories and models of various story-based teaching methods and then formulates the theoretical frame of story based teaching methods for ideological and political theory courses in Chinese universities. After expert interviews verify the theoretical frame, it constitutes the es ential teaching frame. On this basis, the teaching method developed is implemented in a Chinese university to demonstrate the ef ect on the participation of students' learning engagement and learning at itude and analyze the students' perception of teaching. The results of this study are as fol ows. First, through literature review and expert interviews, the story basis course frame of ideological and political theory courses in Chinese universities is mainly combined with the nar ative teaching design model of Lauritzen and Jaeger (19 7) and the story form model of Egan(1986). There are thre parts of teaching objective set ing, story based teaching implementation, and teaching mat ers ne ding at ention. The story based teaching implementation stage is 1) selecting context, 2) inquiries, 3) the goals filter, 4) explorations, 5) culminations and as es ment. Second, after the practical implementation of the developed teaching method in Chinese universities, it has be n confirmed that this teaching method af ects students' learning engagement and at itude,as wel as their perceptions of the course. In addition, the experimental group's student perception, study engagement, and at itude were significantly improved compared with the control group. Therefore, the conclusions of this study are as fol ows: First, the combination of story based teaching methods and ideological and political theory courses in Chinese universities is appropriate. Second, story based courses promote university students' learning engagement, at itude, and perception. The theoretical frame of “the story based ideological and political theory course” derived from this study wil inspire other similar systems in universities, such as general political education courses, to apply story based teaching methods.