http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
알베르 라페의 영화의 논리 - 앙드레 바쟁과의 비교를 중심으로
안상원 ( Ahn¸ Sangwon ) 한국미학회 2021 美學 Vol.87 No.3
영화학자로서 알베르 라페는 앙드레 바쟁과 동시대 인물이라고 할 수 있다. 영화 이미지의 사실적 느낌을 각자의 출발점으로 삼았다는 점에서 두 이론가는 공통점을 가지고 있지만, 그들의 이론이 궁극적으로 지향하는 바는 매우 달랐다. 바쟁이 영화 매체를 통한 리얼리티의 구현에 경도되어 있었던 반면, 라페는 실사 이미지를 통한 인위적 질서의 구축에 주목했다. 라페의 영화론에서는 실사 이미지가 현실의 두께를 제시하는 실존의 구현이 되어야 하며, 궁극적으로 영화는 이미지의 물질성에 의해 서사가 결정되는 정황의 예술이 된다. 따라서, 바쟁이 리얼리즘 영화론의 대표적인 이론가로 분류되는 반면, 라페는 물질적 결정론을 통해 해석되어야 하는 일련의 이미지로서 영화를 간주했기에, 전통적인 사실주의와 형식주의의 대립을 넘어설 수 있는 하나의 길을 제시한 이론가로서 평가될 수 있을 것이다. As a film theorist, Albert Laffay was a contemporary of André Bazin. These two theorists had something in common in that they took as their starting point the impression of reality of the cinema, but they proceeded respectively toward quite a different conclusion. While Bazin was preoccupied with the revelation of the real on the screen, Laffay paid attention to the construction of an artificial order through the realistic images. In Laffay's film theory, photographic/filmic images should be the embodiment of the existence that represents the thickness of reality, and ultimately the art of conjuncture in which the narrative is determined by the materiality of the image. Thus, while Bazin is considered as a representative theorist of film realism, Laffay could be regarded as a theorist who suggested a path beyond the traditional opposition of realism and formalism, because he took the cinema as a series of looks which should be interpreted through the determinism of images to be a moral discourse.
Analysis of Requirements for New Aids-to-Navigation Service: Focusing on Ship Operator
Sangwon Park(Sangwon Park),Min-Ji Jeong(Min-Ji Jeong),Yunja Yoo(Yunja Yoo) 한국항해항만학회 2022 한국항해항만학회지 Vol.46 No.5
With digitalization accelerated by the 4th industrial revolution, the marine traffic environment is also expected to change rapidly. The purpose of this study was to derive a new service for Aids to Navigation (AtoN) to respond to a new marine traffic environment and suggest service requirements from the perspective of the ship’s operator. For this purpose, a survey of ship operators was conducted and IPA analysis was performed. Marine traffic intelligence and information services and Image-based ocean environment information services were deduced as high-priority services from the perspective of ship operators. Thus, the derived results can serve as basic data for developing new AtoN services and determining policy directions.
Mechanism Design of Open Innovation for Smart City
Sangwon Lee(이상원),Dongoun Choi(최동운),Sungbum Park(박승범),Jumin Lee(이주민),Hyunsup Ahn(안현섭) 한국엔터테인먼트산업학회 2014 한국엔터테인먼트산업학회 학술대회 논문집 Vol.2014 No.11
요즘, 혁신이라는 것은 기업의 미래를 위해 왜 중요한지에 대한 의심의 여지는 없다. 혁신은 기업의 성장과 이익을 유지하는데 필수적인 요소로 받아들여지고 있기 때문이다. 하지만, 단순한 혁신은 기업이 운영하는데 있어서, 위험성이 증가하고 있다. 개방형 혁신은 혁신의 새로운 개념으로 학계의 주의를 끌고 있다. 우리는 폐쇄형 혁신과 개방형 혁신의 개념에 대해 조사하고, 이른바 스마트 도시 설계를 위한 주요 고려사항을 살펴본 후, 개방형 혁신을 통한 스마트 도시 설계를 위한 주요 방법론과 비즈니스 사례에 대해서도 연구하고자 한다. Nowadays, there is no inquiry left about why innovation is crucial for the future of the enterprises. It has been accepted its necessity to sustain growth and profitability for enterprises. But, the simple innovation becomes riskier for the companies to run by them. The new term, Open Innovation, appears as an attractive alternative and has captured the attention of the academics as a new breed of innovation. We research on the concept of Innovation, Closed Innovation and Open Innovation, and some considerations to design Smart City. And then, we study on several famous methodologies and their business cases.
번역 생성기와 문장 판별기 기반의 문장 데이터 증강 방법
SangWon Lee,WonIk Choi 한국차세대컴퓨팅학회 2021 한국차세대컴퓨팅학회 학술대회 Vol.2021 No.05
본 논문에서는 번역 생성기 (translation generator)와 문장 판별기 (sentence discriminator) 기반의 문장 데이터 증강 방법을 제안한다. 번역 생성기는 다른 언어로의 반복 번역을 통해 원본 문장과 비슷한 의미를 가진 변형문장을 생성하여 데이터의 수를 늘린다. 하지만 이러한 방식으로 생성된 문장의 일부는 학습에 방해가 된다. 이에 문장 판별기는 학습에 방해되는 문장을 판별하기 위하여 Convolutional Neural Network (CNN)와 Bidirectional Long Short-Term Memory (Bi-LSTM)를 병렬로 병합한 딥러닝 모델로 원본문장을 학습하고 변형문장 중에서 학습에 방해되는 문장을 판별하고 제거하여 데이터 증강 문장을 출력한다. 본 논문의 데이터 증강 방법으로 데이터를 증강한 후 5가지의 딥러닝 모델로 테스트를 진행한 결과, 데이터의 크기에 따라 최대 9.28%의 성능 향상을 보였다.
Stress Corrosion Cracking of Heat Exchanger Tubes in District Heating System
( Sangwon Cho ),( Seon-hong Kim ),( Woo-cheol Kim ),( Jung-gu Kim ) 한국부식방식학회(구 한국부식학회) 2019 Corrosion Science and Technology Vol.18 No.2
The purpose of this paper is to present failure analysis, of the heat exchanger tube in a district heating system. SS304 stainless steel is used, as material for the heat exchanger tube. The heat exchanger operates in a soft water environment containing a small amount of chloride ions, and regularly repeats operation and standstill period. This causes concentration of chloride ions on the outer surface of the tube, as well as repeat of thermal expansion, and shrinkage of the tube. As a result of microscopic examination, cracks showed transgranular as well as branched propagation, and many pits were present, at the initiation point of each crack. Energy disperstive spectroscopy analysis showed Fe and O peak, as well as Cl peak, meaning that cracks were affected by Cl ion. Failure of the tube was caused by chloride-induced stress corrosion cracking by thermal stress, high temperature, and localized enrichment of chloride ions.
Requirement-Oriented Entity Relationship Modeling
Sangwon Lee,Kyung-shik Shin 한국데이타베이스학회 2010 Journal of information technology applications & m Vol.17 No.3
Most of enterprises depend on a data modeler during developing their management information systems. In formulating business requirements for information systems, they widely and naturally use the interview method between a data modeler and a field worker. But, the discrepancy between both parties would certainly cause information loss and distortion that lead to let the systems not faithful to real business works. To improve or avoid modeler-dependant data modeling process, many automated data design CASE tools have been introduced. However, since most of traditional CASE tools just support drawing works for conceptual data design, a data modeler could not generate an ERD faithful to real business works and a user could not use them without any knowledge on database. Although some CASE tools supported conceptual data design, they still required too much preliminary database knowledge for a user. Against these traditional CASE tools, we proposed a Requirement-Oriented Entity Relationship Model for automated data design tool, called ROERM. Based on Non-Stop Methodology, ROERM adopts inner systematic modules for complete and sound ERD that is faithful to real field works, where modules are composed of interaction modules with a user, rules of schema operations and sentence translations. In addition to structure design of ROERM, we also devise detailed algorithms and perform an experiment for a case study.