• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2004.05a
• /
• pp.211-211
• /
• 2004

최근 인터넷 비즈니스의 증가로 고객 수요를 정확하게 예측하여 적절한 상품을 추천하는 추천 시스템의 개발 및 사용이 활발해지고 있다. 현재의 추천 시스템은 주로 내용 기반 추천(Content based Filtering)과 협력적 추천(Collaborative Filtering)을 사용하고 있으나 이러한 추천 시스템으로는 추천의 이유나 배경의 설명이 곤란하며, 시시각각 변하는 사용자의 의도를 파악하고 적절하게 응답하기에는 부족한 면이 있다.(중략)

• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.1009-1012
• /
• 2005

e-러닝의 활용도와 역할이 커져 가면서, 온톨로지(Ontology)나 컨셉맵(Concept Map)을 이용하여 e-러닝의 학습효과를 높이는 방안들이 연구되고 있다. 그러나 아직까지 e-러닝에 온톨로지나 컨셉맵 개념을 적용한 연구 사례는 미미한 수준이며, 이들간의 연계에 대한 고려 없이 별개의 대상으로 다루어져 왔다. 본 연구는 온톨로지와 컨셉맵의 상호 연관 관계와 각각의 장점들을 분석하여 학습에 있어서 시너지(Synergy)를 가져올 수 있는 새로운 e-러닝 시스템 구축 방안을 제안한다. 제안 시스템에서 온톨로지와 컨셉맵 간의 연계는 컨셉맵 에이전트에 의해 이루어진다. 컨셉맵 에이전트는 학습자의 수준이나 관심영역(주제와 범위)에 맞게 온톨로지로부터 추출한 학습 콘텐츠를 재구성해 준다. 학습자는 제안 시스템의 사용자 인터페이스를 통해 자신이 이해하고 있는 지식을 컨셉맵 형태로 표현할 수 있고, 컨셉맵 에이전트에게 요청하여 제공 받은 모범답안 컨셉맵과 자신이 표현한 컨셉맵을 비교하여 학습자가 스스로 자기 평가를 할 수 있다. 본 e-러닝 시스템이 제공하는 이러한 새로운 형태의 학습 환경은 학습자가 학습 지식에 대해 보다 체계적으로 접근하여 효과적으로 학습할 수 있게 해준다. 또한, 학습에 있어서 컨셉맵을 이용하므로 학습 형태의 특성상 보다 원천적으로 암기 위주의 학습에서 탈피하여 구성주의적인 학습을 가능하게 한다.

• Journal of KIISE:Databases
• /
• v.34 no.6
• /
• pp.493-502
• /
• 2007

Due to a growing interest in searching and expressing knowledge effectively, knowledge management methods such as Topic Map are becoming more important. Topic Map organizes knowledge that is full of intricate relations, so maintaining and managing Topic Map consistently is very essential. TMCL and other constraint languages have limits as they can check simple constraints but can not support complex constraints like dependence constraints. Current constraint checking systems operating at the application level are also showing an inferiority in performance. In this paper, we extend TMCL based on the characteristics of other constraint languages in the information system field and related fields. We build and propose an RDBMS-based Topic Map constraint checking system to support the extended constraint language effectively. This new system handles complex types of constraints like dependency constraint as well as basic Topic Map constraints present in the TMCL. As the system examines each constraint it uses templates to generate queries for effective checking and overall shows a higher performance level than current systems.

• Information Systems Review
• /
• v.10 no.2
• /
• pp.1-19
• /
• 2008

The purpose of this research is to schematize and suggest the new trends of study and changing aspects of science and technology hidden in a bibliographical phenomenon of documentation to researchers and policy-makers all through the Knowledge Map. The field of study to be analyzed in this research is the Access Network field. The reason why this field has been selected as the main target of study is that the Access Network field is economically important and characterized by its wide sphere where a variety of fields are interconnected. In addition, it is important to measure the applied as well as fundamental aspects of technology by using bibliographical method and technique. Knowledge Map successfully visualizes the inter-relations of the keywords and sub-fields of Access Network. The importance of visualizing methods in the convincing presentation of results has not been sufficiently understood in the past. Knowledge Map opens a new opportunity for cartography of science and information visualization. The Knowledge Map results have produced a great deal more than statistical artifact. We aimed to exploit the visualization effect of the Knowledge Maps to the aid of searchers in Access Network domain, and the results are quite encourging.

• Journal of The Korean Association of Information Education
• /
• v.13 no.2
• /
• pp.215-224
• /
• 2009

The aim of teaching writing is not to memorize a single piece of knowledge but to acquire higher thinking that could be related to other fields using learners' various reading experiences and schemata. Among several strategies or techniques to write well, it is recommended to use a mind-map technique. A mind-map software applies the mind-map technique into computer for learners to study autonomously. However, the mind-map software is not common in educational fields yet and there are only few studies on the effect of it. Therefore, this study analyzed various curricula to extract the elements of writing and developed a mind-map software program which was applied and tested. As a result, the writing program using the mind-map software was effective in promoting learners' motivation and lowering anxiety. It also improved learners' competence of writing such as understanding, analytic thinking, creativity, ability to organize, and expression.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.31-39
• /
• 2023

In this paper, we propose an Internal/External Knowledge Distillation (IEKD), which utilizes both external correlations between feature maps of heterogeneous models and internal correlations between feature maps of the same model for transferring knowledge from a teacher model to a student model. To achieve this, we transform feature maps into a sequence format and extract new feature maps suitable for knowledge distillation by considering internal and external correlations through a transformer. We can learn both internal and external correlations by distilling the extracted feature maps and improve the accuracy of the student model by utilizing the extracted feature maps with feature matching. To demonstrate the effectiveness of our proposed knowledge distillation method, we achieved 76.23% Top-1 image classification accuracy on the CIFAR-100 dataset with the "ResNet-32×4/VGG-8" teacher and student combination and outperformed the state-of-the-art KD methods.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.5
• /
• pp.961-966
• /
• 2012

In this paper, we analyze a main patent positioning to analyze data based on patent application as stage of data collection to organize knowledge map of LED through patent application. We also analyze the present condition of patent application for technology sector. We propose basic data to make knowledge map through the analysis of technical distribution for applicant by each country.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.379-384
• /
• 2006

본 논문에서는 과학기술 정보분석을 위한 방법론 중의 하나인 지식맵을 기반으로 유망연구영역을 탐색하는 방법론을 다룬다. 이를 위해 정보보안 분야를 대상으로 지식맵을 작성하고, 또한 허브탐색을 통한 정보보안 분야의 유망연구영역을 제안하였다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.1
• /
• pp.1-6
• /
• 2013

In this paper, we review an outline as premise for environmental analysis of green IT which has basic stage through the world in order to make knowledge map of green IT. We can analyze genuine environment for green IT based on this premise, and we expect that we can also use to analyse power of competition based on intellectual property after performing these environmental analysis.

• Product Safety
• /
• s.210
• /
• pp.17-17
• /
• 2011

지식경제부는 지난 5월 26일(목) "산업기술로드맵"을 공유 기획하는 웹시스템인 "MAE(매)"를 민간에 공개했다.