• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.07a
• /
• pp.277-280
• /
• 2019

본 논문에서는 화학분야 특허 문서에 존재하는 도표(TABLE) 데이터를 인공지능 기술을 활용하여 자동으로 추출하고 정형화된 형태로 가공하는 방법을 제안한다. 특허 문서에서 도표 데이터는 실시예에서 실험결과나 비교결과를 간결하고 가시적으로 표현하기 위하여 주로 사용되나, 셀의 속성을 정의하는 헤더부분과 수치가 표현되는 값 부분의 경계가 모호하여 구조화하는데 어려움이 있다. 본 논문에서 제안하는 방법은 소량의 학습데이터를 구축하고 기계학습을 통해 도표에 존재하는 셀의 속성을 예측하고, 예측된 속성을 토대로 조성과 물성 정보를 자동으로 구분하여 추출하는 방법을 제시한다. 제시된 방법을 활용하여 화학 분야 조성물 특허의 도표데이터에 시뮬레이션 결과 각 항목별 98.17%의 속성 예측 정확도를 나타내었으며 기존 규칙기반 연구보다 작업난이도, 예측정확도에서 우수한 성과를 보인다.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2015.07a
• /
• pp.280-281
• /
• 2015

해사영어는 SMCP(표준해사통신용어)를 기반으로 하는 특수목적영어의 한 분야로 콘텐츠 개발을 비롯한 학습과 훈련을 지원할 수 있는 해사영어훈련기관을 통해 효과적인 학습과 훈련이 제공되어 진다면 교육생을 단기간에 그 구사 능력을 상당한 수준에 이르게 할 수 있다. 이 연구에서는 우리나라 해양분야의 국제 경쟁력을 제고하고 새로운 일자리 창출을 위해 필수적인 해양산업 종사자들의 영어구사 능력개발을 위한 해사영어훈련센터(METC)의 필요성과 설립방안을 제시한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.549-550
• /
• 2021

사용자에게 맞춘 개인화된 제품과 서비스를 제공하는 기술의 발전으로 개인화의 수요는 점점 늘어날 것으로 전망하고 있다. 또한 개인 맞춤형으로 전문 스포츠 선수화, 족부 장애우를 위한 정형 제화 등 전문적인 기능 중심의 개인화나 패션을 위한 스타일 중심의 개인화 등 개인 맞춤 제작 신발을 제작할 때 기존의 아날로그적인 방식으로 발 변인을 측정했을 때 각 변인에 대해 기준점이 명확하지 않아서 재현성이 떨어진다. 따라서 본 논문에서는 자를 이용해 간단히 측정 가능한 기본적인 발 변인 이용하여 다른 변인들을 학습하고 딥러닝을 이용해 추정하는 방법에 대해 서술한다. 이를 위해 20개의 발 변인을 휙득 하였고 그 중 6개의 기본적인 발 변인을 이용해 14개 변인을적합 방지를 위해 Dorpout을 적용해 학습하고 학습한 데이터를 이용해 학습하지 않은 데이터를 테스트해 각 변인별 결과를 보여준다.

• The Journal of the Korea Contents Association
• /
• v.21 no.6
• /
• pp.701-715
• /
• 2021

The drastic advance of recent deep learning technologies is heavily dependent on training datasets which are essential to train models by themselves with less human efforts. In comparison with the work to design deep learning models, preparing datasets is a long haul; at the moment, in the domain of vision intelligent, datasets are still being made by handwork requiring a lot of time and efforts, where workers need to directly make labels on each image usually with GUI-based labeling tools. In this paper, we overview the current status of vision datasets focusing on what datasets are being shared and how they are prepared with various labeling tools. Particularly, in order to relieve the repetitive and tiring labeling work, we present an interactive smart image annotating system with which the annotation work can be transformed from the direct human-only manual labeling to a correction-after-checking by means of a support of automatic labeling. In an experiment, we show that automatic labeling can greatly improve the productivity of datasets especially reducing time and efforts to specify regions of objects found in images. Finally, we discuss critical issues that we faced in the experiment to our annotation system and describe future work to raise the productivity of image datasets creation for accelerating AI technology.

• Journal of Gifted/Talented Education
• /
• v.20 no.3
• /
• pp.831-846
• /
• 2010

The Purposes of this study were to compare the level of study habits and test anxiety between gifted middle-school students and non-gifted and to find out the correlation between study habits and test anxiety in two groups. The total participants of this study were 437 middle school students. One hundred eighty three students (127 boys, 56 girls) belonged to gifted group who were enrolled in Cyber Education Center for Math & Science Gifted Students of KAIST in Daejeon. And two hundred fifty four (128 boys, 126 girls) were non-gifted group who were from the middle school in Seoul City and Gyeonggi province. The results revealed that the level of study habits of gifted middle school students was higher than that of non-gifted. And gifted group felt lower level of test anxiety than non-gifted group. Additionally gifted boys showed significantly higher level of study skills application behavior than gifted girls.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.739-741
• /
• 2022

개인교수형, 반복학습형, 학습용 게임형, 시뮬레이션형 프로그램등의 교육콘텐츠등의 우수사례인 장애인식교육 이러닝센터와 Simcity 시리즈를 교육 목적, 교육 대상, 관련 교과목, 전달 매체, 전체 분량, 프로그램 및 학습 내용의 구조를 바탕으로 분석하여 교육 자료가 필요한 교수자들이 활용할 수 있는 방안을 모색하였다.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.191-206
• /
• 2022

Recently, it is a de-facto approach to utilize a pre-trained language model(PLM) to achieve the state-of-the-art performance for various natural language tasks(called downstream tasks) such as sentiment analysis and question answering. However, similar to any other machine learning method, PLM tends to depend on the data distribution seen during the training phase and shows worse performance on the unseen (Out-of-Distribution) domain. Due to the aforementioned reason, there have been many efforts to develop domain-specified PLM for various fields such as medical and legal industries. In this paper, we discuss the training of a finance domain-specified PLM for the Korean language and its applications. Our finance domain-specified PLM, KB-BERT, is trained on a carefully curated financial corpus that includes domain-specific documents such as financial reports. We provide extensive performance evaluation results on three natural language tasks, topic classification, sentiment analysis, and question answering. Compared to the state-of-the-art Korean PLM models such as KoELECTRA and KLUE-RoBERTa, KB-BERT shows comparable performance on general datasets based on common corpora like Wikipedia and news articles. Moreover, KB-BERT outperforms compared models on finance domain datasets that require finance-specific knowledge to solve given problems.

• Journal of the Korean Society for Library and Information Science
• /
• v.30 no.2
• /
• pp.127-148
• /
• 1996

The purpose of this study is to assert the fact that School Library Media Center(SLMC) is the most important body which influences the whole process of loaming in school and at the same time it functions as a very essential infrastructure of acquiring loaming in the society. The writer analysed the studies on learning outcomes across a range of educational conditions and settings. She found that loaming outcomes are influenced positively by the educational conditions and settings that accomodate the student diversity and individualization, and loaming materials for their individual needs. It means that the outcomes of effective teaming are influenced by using teaming materials of individual student guided by teachers and teacher librarians. In case, school education furnishes desirable programs of SLMC and use it properly, students canhave enough experiences what and how they learn by using library materials during their school days. As school days are in their early days of their lifetime, they are easy to carry their way of loaming with them even after they finish school. The members of society who are accustomed to use library materials during their school days will have loaming needs for their better life. The writer wants to call it the loaming needs of society that school produces. A teaming society is composed of two factors One is the loaming needs of the people and the other is the environment to meet them. SLMC produces loaming needs, and it meets to the needs of student learning. Consequently it can be said SLMC is the infrastructure for loaming society. The author pointed out that the Educational Reformation Draft of current Government does not mention about learning materials prerequisite to enhance student loaming outcomes, especially in relation to their creativity. She concluded educators of every level of school including government officials in charge of education have to think and do something about reformation of School Library Media Center for the fulfillment of the goal of Korean Educational Reformation Draft announced by the Educational Reformation Committee Draft announced by the Educational Reformation Committee May 31, 1995.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.228-230
• /
• 2022

딥러닝 기반의 인공지능 기술이 발달함에 따라 이상 탐지 방법에도 딥러닝이 적용되었다. 네트워크 트래픽으로부터 요약 및 집계된 Feature 를 학습하는 방법과 Packet 자체를 학습하는 등의 방법이 있었다. 그러나 모두 정보의 제한적으로 사용한다는 단점이 있었다. 본 연구에서는 Http Request에 대한 사전학습 기반의 효과적인 이상 탐지 방법을 제안한다. 사전학습에 고려되는 토큰화 방법, Padding 방법, Feature 결합 방법, Feature 선택 방법과 전이학습 시 Numerical 정보를 추가하는 방법을 소개하고 각 실험을 통해 최적의 방법을 제안한다.

• Journal of The Korean Association of Information Education
• /
• v.22 no.1
• /
• pp.123-130
• /
• 2018

The purpose of this study is to investigate the process in which children center students are interested in SW education through the change of perception of SW education. Since the students in the regional Community Child Center are different in grade and age, they have suggested curriculum and teaching-learning process for each level. As a result, the students' attitude and satisfaction with SW education increased and it was found that the education program for the underprivileged is desperately needed.