AWS Lambda Serverless Computing 기술을 활용한 효율적인 딥러닝 기반 이미지 인식 서비스 시스템

이현철,이성민,김강석 한국정보처리학회 2020 정보처리학회논문지. 소프트웨어 및 데이터 공학 Vol.(이전)9 No.6

최근 딥러닝(Deep Learning) 기술의 발전에 따라 컴퓨터 비전(Computer Vision) 분야의 이미지 인식 성능이 향상되고 있으며, 또한 Serverless Computing이 이벤트 기반의 클라우드 애플리케이션 개발 및 서비스를 위한 차세대 클라우드 컴퓨팅 기술로 각광받고 있어 딥러닝과 Serverless Computing 기술을 접목하여 실생활에 이미지 인식 서비스를 사용하고자 하는 시도가 증가하고 있다. 따라서 본 논문에서는 Serverless Computing 기술을 활용하여 효율적인 딥러닝 기반 이미지 인식 서비스 시스템 개발 방법을 기술한다. 제안하는 시스템은 Serverless Computing 기반 AWS Lambda Server를 이용하여 적은 비용으로 대형 신경망 모델을 사용자에게 서비스할 수 있는 방법을 제안한다. 또한 AWS Lambda Server의 단점인 Cold Start Time 문제와 용량제한 문제를 해결하여 효과적으로 대형 신경망 모델을 사용하는 Serverless Computing 시스템을 구축할 수 있음을 보인다. 실험을 통해 AWS Lambda Serverless Computing 기술을 활용하여 본 논문에서 제안한 시스템이 비용 절감뿐만 아니라 처리 시간 및 용량제한 문제를 해결하여 대형 신경망 모델을 서비스하기에 효율적인 성능을 보임을 확인하였다. Recent advances in deep learning technology have improved image recognition performance in the field of computer vision, and serverless computing is emerging as the next generation cloud computing technology for event-based cloud application development and services. Attempts to use deep learning and serverless computing technology to increase the number of real-world image recognition services are increasing. Therefore, this paper describes how to develop an efficient deep learning based image recognition service system using serverless computing technology. The proposed system suggests a method that can serve large neural network model to users at low cost by using AWS Lambda Server based on serverless computing. We also show that we can effectively build a serverless computing system that uses a large neural network model by addressing the shortcomings of AWS Lambda Server, cold start time and capacity limitation. Through experiments, we confirmed that the proposed system, using AWS Lambda Serverless Computing technology, is efficient for servicing large neural network models by solving processing time and capacity limitations as well as cost reduction.

클라우드 컴퓨팅 기반 사회과 협동학습의 원리와 구조 설계

박기범(Kee-Burm Park) 한국사회과교육연구학회 2015 사회과교육 Vol.54 No.1

본 연구는 클라우드 컴퓨팅 기반 사회과 협동학습 모형의 개발에 목적이 있다. 클라우드 컴퓨팅은 기존의 웹 사이트 방식과는 달리 가상의 공간에서 사용자의 컴퓨팅 작업을 지원하는 시스템이다. 구성주의적 교육패러다임은 21세기에도 지속되고 있으며, 협동학습이 효과적인 방법론으로 간주된다. 전통적 협동학습의 패러다임은 교실 환경을 근간으로 논의되었다. 그러나, 오늘날 네트워크의 발전으로 빠르게 확산되는 클라우드 환경에서는 교실환경이 아닌 클라우드 환경에 기반 한 협동학습의 설계가 필요하다. 본 연구는 협동학습모형들의 공통된 특징과 클라우드 기반 컴퓨팅의 장점에 터하여 클라우드 기반 협동학습을 설계하였다. 구성주의에 기반 한 협동학습의 공통된 특징은 ①긍정적 상호의존성, ②개별 책임성, ③동등한 참여, ④동시적 상호작용, ⑤열린 지식 구성으로 정리할 수 있다. 이러한 협동학습의 주요 특징을 클라우드 컴퓨팅 환경의 장점과 연결하면 ①‘열린 지식 환경’, ②학습공동체 내에서 ‘효과적인 상호작용’, ③‘진화하는 지식 구성’을 클라우드 기반 협동학습의 원리로 도출할 수 있다. 이에 터한 클라우드 기반 협동학습 모형은 ①과제 확인, ②클라우드 컴퓨팅 환경 선정, ③클라우드 영역 할당, ④과제관련 정보 수집, ⑤개별 지식 구성 및 소통, ⑥조별 지식 구성 및 소통, ⑦최종 결과물 확정 및 동료 평가의 단계로 설정될 수 있다. 클라우드 컴퓨팅의 투명성, 평등성은 학습자의 긍정적 상호작용을 도울 수 있으며 효과적이고 창의적인 결과를 산출하도록 한다. 이는 지식 생산자로서의 창의적인 시민 역량의 함양에 기여한다. The purpose of this study is the design of cloud computing-based collaborative learning model for social studies. Unlike traditional website methods, the cloud computing is a system that supports the user’s computing experience in a virtual space. Cloud computing can support effectively collaborative learning to direct the constructivism. Constructivism educational paradigm is being continued in the 21st century, cooperative learning is regarded as an effective methodology. Through the media and networks of the 21st century, constructivismoriented cooperative learning can be effectively executed. Common features of collaborative learning based on constructivism are ① positive interdependence, ② individual accountability, ③ equal participation, ④ simultaneous interaction, ⑤ open knowledge construction. By connecting these key elements of cooperative learning and characteristics of cloud computing environments, ① ‘open knowledge environment’, ② ‘effective interaction within learning communities’, ③ ‘evolving knowledge construction’ can be derived as a key principle of learning for cloud-based collaboration. This study was designed for cloud-based collaborative learning based on the core principles of cooperative learning and key elements of cloud-based computing. Cloud-based collaborative learning can be set to ① identification of challenges ② selecting cloud computing environment, ③ cloud area allocation, ④ challenges related information gathering, ⑤ individual knowledge construction and communication, ⑥ group knowledge building and communication, ⑦ final results presentation and peer review. Transparency and equality of cloud computing can support the positive interaction of learners. By doing this, we can make the effective and creative output.

메타-학습: 이론적 토대, 주요 원리와 모형, 그리고 컴퓨터-지원 협력학습의 적용

박휴용(Hyu-Yong Park) 숭실대학교 영재교육연구소 2022 Global Creative Leader Vol.12 No.2

본고는 기존의 인본주의적 이론에 입각한 경험 학습을 대신하는 메타-학습의 개념과 특징 및 유형과 학습이론으로서의 가능성을 논의하고자 하였다. 이를 위해 본고는 세 가지 연구문제를 제시하였다: i) 경험 학습의 기본 원리와 그 특징은 무엇이며, 인간의 학습환경과 학습 양식은 어떻게 변하고 있는가?; ii) 메타-학습의 개념과 기본 구조는 무엇이고, 어떤 원리와 모형이 있는가?; iii) 메타-학습이 실제적으로 적용되는 컴퓨터-지원 협력 학습의 원리와 고려해야할 요인은 무엇인가? 본고의 핵심적 주장을 정리하면 다음과 같다. 첫째, 경험 학습은 대면적이고 개별적이며 맥락적인 경험을 주목하여 주로 인지적 학습을 설명하는데 주력하였지만, 메타-학습은 비대면적, 탈맥락적, 간접적, 그리고 매체활용 학습에 유용하다는 것이다. 둘째, 메타-학습은 직관적/통찰적 학습과 기술기반 학습에 활용도가 높기 때문에 다양한 정보통신 기기들이 생활화된 오늘날 이러한 기술적 도구들(컴퓨터, 스마트/유비쿼터스 기기, PPT, 각종 앱 등)에 익숙한 학습자들에게 특히 유용하다는 것이다. 셋째, 메타-학습의 주요 원리(모형)는 외적 강화학습, 분산적 학습, 연합적 학습, 그리고 자기-적응적 학습을 통해 설명될 수 있다. 넷째, 메타-학습이 구체적으로 구현된 컴퓨터-지원 학습은 실제 세계를 최대한 재현한 디지털 학습플랫폼에서 다양한 기술적 매체를 활용하여 개별적이면서도 다중적이고 자기-적응적 보완 학습을 촉진하는 과정이라는 것이다. 마지막으로 본고는 메타-학습과 관련한 주요한 논란점에 대해 논의하였고, 이러한 논의가 인본주의적 학습과 기계학습의 공통된 원리인 메타-학습에 대한 학계의 추가적 연구를 촉발할 수 있을 것이다. This study aims to discuss the concepts, characteristics, and principles (models) of meta-learning, as a strategic model for machine learning, which might supplement humanistic experiential learning theory. To address this issue, this paper suggested three kernel research questions: i) What are the basic principles and traits of experiential learning and how are the learning environments (spaces) and modalities changing in the technology-oriented society?, ii) What are the notions and basic structures of meta-learning and what principles and models can be adopted as the meta-learning practices?, iii) What is the actual case of meta-learning employed in a technology-based learning milieu and what are the seminal principles and necessary tools for it? This paper’s biggest arguments are as follow: First, experiential learning has been solely interested in contact, personal, and aspects of contextual experiences for cognitive learning, while meta-learning has special attention on non-contact, indirect, beyond-context, and a mediated learning. Second, since meta-learning is especially useful for intuitive and technology-based learning situations, it helps much the learning practices of technology-friendly learners. Third, the meta-learning principles (models) may include meta-reinforcement learning, distributed learning, federated learning, and self-adaptive learning. Fourth, meta-learning has been actualized in the computer-supported collaborative learning, which highlights the individual, mediated, and self-adaptive learning that utilizes technological tools diversely in a digital learning platform. This paper’s suggestive arguments on meta-learning may promote active discussions in academic fields about the link between humanistic experiential learning and machine learning.

A Development of Adaptive Learning System Considering Learners’ Knowledge Level

이재무,김민희 한국지식정보기술학회 2018 한국지식정보기술학회 논문지 Vol.13 No.6

The biggest advantage of computer based education is providing differential learning considering each learner’s situation. Recently, adaptive learning system to support individualized learning has been actively studied. This study is about the adaptive learning system considering learners' individual achievement. This paper examines an adaptive learning system applying van Hiele’s learning theory in order to support differential leveled learning in geometrical figures for elementary students. We classified learners’ level of knowledge into three levels; lower level, middle level and upper level through the online test. We provide the leveled appropriate courses for the geometrical figures learning. A PHP script language is used to implement for learning algorithm and to manipulate the database. Finally, a Flash is used to present for adaptive learning object. The learning achievement of our proposed adaptive learning system will be increased because the learning system is constructed using the figure learning teaching strategies proposed by proven van Hiele’ theory. Futhermore, the adaptive learning system considering learners’ knowledge level will increase not only learning effects but also learning motivation by providing proper leveled learning and content. This study contributed to realize an adaptive learning system that takes into account the learner's knowledge level. And we showed how mathematical theory can connect with an adaptive learning system.

외국어 학습에서의 컴퓨터 및 모바일 기기의 교육 효율성에 대한 연구

주치운 한국컴퓨터정보학회 2023 韓國컴퓨터情報學會論文誌 Vol.28 No.5

This study aims to show that "Computer-assisted language learning (CALL)" and "Mobile-based language learning (MALL)" actually influence education, deviating from the traditional "drill and practice" method in foreign language education and learning due to the development of information and communication technology (IT). Specifically, for first-year college students who have relatively poor English skills and do not feel enough motivation for English learning, I will produce educational video content using multimedia authoring tools and upload it to the e-learning system. Video content is configured to be accessed and utilized through various media such as computers, smartphones, tablets, laptops, etc. Ultimately, an exploration of educational value behind the utilization of IT devices in English language Teaching(ELT) and the Second Language Acquisition (SLA) theory behind effective instructional use of such technology are presented. That is to say, the effectiveness of language learning using information and communication technology (IT) is introduced. The article closes by suggesting how to use computers and mobile media for 'Flipped Learning'. 본 연구는 정보통신기술(IT)의 발전으로 외국어 교육과 학습에서 기존의 전통적인 ‘훈련과 연습(drill and practice)’의 방법에서 벗어난 ‘컴퓨터 보조언어학습(Computer-assisted learning: CALL)’과‘모바일 지원 언어학습(Mobile-based Language Learning: MALL)’이 실제로 어학 교육에 미치는 영향을 분석한다. 연구 대상은 비교적 영어 능력이 상대적으로 우수하지 않고, 또한 영어학습에 대한 동기부여를 충분히 느끼지 못하는 전문대학 1학년 학생들을 대상으로 진행하였다. 연구 방법은 멀티미디어 저작 도구를 이용하여 디지털 콘텐츠를 제작하여 이러닝(e-learning) 시스템에 등재하여 교육보조재료로 한 학기 동안 활용한다. 동영상 콘텐츠는 컴퓨터, 스마트폰, 테블릿, 노트북등 다양한 매체로 접근하여 활용할 수 있도록 구성한다. 동영상 콘텐츠에 대한 설문조사를 통하여 콘텐츠의 구성, 학습에 도움 여부, 보완해야 할 부분, 콘텐츠 접근 빈도, 학습경로 등을 분석한다. 분석의 결과를 통하여 정보통신기술을 활용한 어학 학습의 효과, 교육용 콘텐츠의 개선해야할 부분 및 교육에 어떻게 활용해야 하는지에 대해 제언한다. 궁극적으로는 ‘플립러닝((Flipped Learning)’을 위한 컴퓨터 및 모바일 매체에 대한 효용성 분석과 활용방안을 제언한다.

A Protocol Model of S3 Computing Designed for Learning Community Platform of College Teachers

Liang Jia,Liuhong Yan 보안공학연구지원센터 2016 International Journal of Future Generation Communi Vol.9 No.7

S3 Computing requires distributed computing performed by social network platform has high scalability and security. Protocol models meeting the requirements of S3 Computing not only ensure the correctness and robustness of distributed computing, but also reduce risks introduced by involvement of nodes with low reputation in computing. These models safeguard the data collections and computations performed on platform of teacher’s learning community for social researches. This paper constructs a protocol model entitled which adapts platform of teacher’s learning community and meets the requirements of S3 Computing. This protocol model is the key step of implementing distributed computations on learning community platform.

A Co-Evolutionary Computing for Statistical Learning Theory

Sung-Hae Jun 한국지능시스템학회 2005 INTERNATIONAL JOURNAL of FUZZY LOGIC and INTELLIGE Vol.5 No.4

Learning and evolving are two basics for data mining. As compared with classical learning theory based on objective function with minimizing training errors, the recently evolutionary computing has had an efficient approach for constructing optimal model without the minimizing training errors. The global search of evolutionary computing in solution space can settle the local optima problems of learning models. In this research, combining co-evolving algorithm into statistical learning theory, we propose an co-evolutionary computing for statistical learning theory for overcoming local optima problems of statistical learning theory. We apply proposed model to classification and prediction problems of the learning. In the experimental results, we verify the improved performance of our model using the data sets from UCI machine learning repository and KDD Cup 2000.

Engaged Learning 수업 모델을 활용한 교양필수 교과목 연구-컴퓨팅적사고와 알고리즘 중심으로

노은희,조문기 한국교양교육학회 2024 교양교육연구 Vol.18 No.4

본 연구는 Engaged Learning 수업 모델을 활용한 S 대학의 교양 필수 ‘컴퓨팅적사고와 알고리즘’수업 사례를 다룬다. 연구 대상은 2023년 1학기 1학년 총 188명이다. 수업은 ‘문제 정의 - 아이디어도출 - 해결 방안 적용⋅확인’ 3단계의 수업 원리를 적용하여 2학점으로 설계하였다. 연구 방법은SPSS를 사용해 t-검증을 하였으며 분반별 분석을 수행하였다. 연구 결과, Engaged Learning 수업 모델을적용한 강의 내용의 수준, 수업 만족, 프로젝트 진행에서 분반별로 유의미한 차이를 보였다. 문예창작,영화예술, 법대 학생들로 구성된 분반이 다른 전공 계열 학생들보다 모두 낮은 만족도를 보였고, 소프트,AI 융합학과 학생들이 Engaged Learning 적용 강의 내용 수준에 대해 가장 만족하였고, 전자공학과IT융합 학생들의 Engaged Learning을 활용한 수업 방식과 Engaged Learning 수업 모델을 적용한 프로젝트 수업에서 만족도가 가장 높았다. Engaged Learning 활용 수업에 있어 학생 참여도는 긍정 84.1%,부정은 0.5%를 보였고, Engaged Learning 수업 모델을 활용한 수업 방식이 강의식 수업보다 수업내용을이해하는데 긍정 73.8%, 부정은 7.50%로 나타났다. 컴퓨팅적사고와 알고리즘 교과목 운영은 EngagedLearning 수업 모델을 적용하여 수업을 운영하는 것이 강의식 수업보다 효과가 있음을 보였다. 향후Engaged Learning 수업 모델은 다른 교과목 수업에서도 활용이 확대될 수 있을 것으로 기대된다. This study deals with a case study of a required Liberal Arts class called “Computational Thinking and Algorithms” atS University using the Engaged Learning class model. The study targets 188 first-year students in the first semester of 2023. The class was designed as a two-credit course by applying the three-stage teaching principle of problem definition - ideageneration - solution application - confirmation. The research method was a t-test using SPSS and quartile analysis. Theresearch results showed significant differences by quartile in the level of class content, class satisfaction, and project progressionapplying the Engaged Learning class model. The quartiles composed of students from the Literary and Creative Arts, FilmArts, and Law departments all showed lower satisfaction than the students from the other majors, while the students fromthe Software and AI integration department were the most satisfied with the level of lecture content applying Engaged Learning. However, the students from the Electronics and IT integration department were the least satisfied with the level of lecturecontent applying Engaged Learning as well as the project progression. The level of student participation in the EngagedLearning-based class was 84.1% positive and 0.5% negative, indicating that the Engaged Learning-based class model wasmore satisfactory to the students than the lecture-based class model. The percentage of affirmative and negative responsesto the Engaged Learning model was 73.8% and 7.50%, respectively, indicating that students were more likely to understandthe course content than in a lecture-style class. We further found that applying the Engaged Learning lesson model to themanagement of the Computational Thinking and Algorithms subject was more effective than lecture-style teaching. It isexpected that the Engaged Learning lesson model will be expanded to be used in other subject classes in the future.

Multi-dimensional Task Offloading using Deep Learning for Vehicular Edge Computing Networks: A Survey

S. Syed Abuthahir,J. Selvin Paul Peter 대한전자공학회 2024 IEIE Transactions on Smart Processing & Computing Vol.13 No.1

Vehicular networks must support connection ubiquity and high levels of services for a large number of vehicles. In vehicular networks, mobile edge computing (MEC) is considered a viable technique, utilizing computing resources at the edge of wireless access networks. This survey examines effectual task offloading systems forvehicular edge computing (VEC) networks. Several deep learning methods have been suggested to enable task offloading schemes. Therefore, task offloading optimality is a main research topic in fog computing. The main aim of this work is to offer readers an overview of the journey from the task offloading concept to its mathematic problem formulation. This survey introduces fog computing and the process of task offloading that emulates numerous aspects of optimization. Subsequently, a number of machine learning and deep learning methodologies are employed in task offloading,and the challenge of fog computing is discussed. This paper provides a detailed statistical analysis from papers published between 2019 and 2022.

LMS 데이터를 활용한 온라인 러닝의 학습 행동 및 효과에 관한 연구

전병호 (사)디지털산업정보학회 2023 디지털산업정보학회논문지 Vol.19 No.2

On-line learning has been adopted as a major educational method due to the COVID-19 pandemic. Students and faculties got accustomed to on-line educational environment as they experienced it during the COVID-19 pandemic. Development of various technologies and social requirement for educational renovation lay groundwork for on-line learning as well. Therefore, on-line learning or blended learning will be likely to go on after the end of COVID-19 pandemic and it is necessary to prepare the guidelines for effective utilizing on-line learning. The primary purpose of this study is to examine the learning behaviors and the learning effects by using LMS data. Learning behaviors were measured in terms of learning time and access frequency for pre-recorded video lectures targeting computer-practice classes. The results of empirical analysis reveal that frequency was the significant predictor of course achievements but learning time was not. The findings of empirical analysis will provide insights that the effective planning and designing on-line classes based on learning behaviors are key to enhancing learning effects and learner’s satisfaction.