• Bulletin of the Korean Chemical Society
• /
• 제34권9호
• /
• pp.2680-2684
• /
• 2013

Protein kinases (PKs) are an important source of drug targets, especially in oncology. With 500 or more kinases in the human genome and only few kinase inhibitors approved, kinase inhibitor discovery is becoming more and more valuable. Because the discovery of kinase inhibitors with an increased selectivity is an important therapeutic concept, many researchers have been trying to address this issue with various methodologies. Although many attempts to predict the activity and selectivity of kinase inhibitors have been made, the issue of selectivity has not yet been resolved. Here, we studied kinase selectivity by generating predictive models and analyzing their descriptors by using kinase-profiling data. The 5-fold cross-validation accuracies for the 51 models were between 72.4% and 93.7% and the ROC values for all the 51 models were over 0.7. The phylogenetic tree based on the descriptor distance is quite different from that generated on the basis of sequence alignment.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.597-598
• /
• 2015

In order to improve the reliability of cost estimation results using CBR, there has been a continuous issue on similarity measurement to accurately compute the distance among attributes and cases to retrieve the most similar singular or plural cases. However, these existing similarity measures have limitations in taking the covariance among attributes into consideration and reflecting the effects of covariance in computation of distances among attributes. To deal with this challenging issue, this research examines the weighted Mahalanobis distance based similarity measure applied to CBR cost estimation and carries out the comparative study on the existing distance measurement methods of CBR. To validate the suggest CBR cost model, leave-one-out cross validation (LOOCV) using two different sets of simulation data are carried out. Consequently, this research is expected to provide an analysis of covariance effects in similarity measurement and a basis for further research on the fundamentals of case retrieval.

• Journal of Information Processing Systems
• /
• 제14권4호
• /
• pp.948-960
• /
• 2018

Many mobile sensing frameworks have been developed to help researcher doing their mobile sensing research. However, energy consumption is still an issue in the mobile sensing research, and the existing frameworks do not provide enough solution for solving the issue. We have surveyed several mobile sensing frameworks and carefully chose one framework to improve. We have designed an adaptive sampling module for a mobile sensing framework to help solve the energy consumption issue. However, in this study, we limit our design to an adaptive sampling module for the location and motion sensors. In our adaptive sampling module, we utilize the significant motion sensor to help the adaptive sampling. We experimented with two sampling strategies that utilized the significant motion sensor to achieve low-power consumption during the continuous sampling. The first strategy is to utilize the sensor naively only while the second one is to add the duty cycle to the naive approach. We show that both strategies achieve low energy consumption, but the one that is combined with the duty cycle achieves better result.

• 한국멀티미디어학회논문지
• /
• 제10권12호
• /
• pp.1655-1662
• /
• 2007

Although the PKl-based user authentication solution has been widely used, the security of it can be deteriorated by a simple password. This is because a long and random private key may be protected by a short and easy-to-remember password. To handle this problem, many biometric-based user authentication solutions have been proposed. However, protecting biometric data is another research issue because the compromise of the biometric data will be permanent. In this paper, we present an implementation to improve the security of the typical PKI-based authentication by protecting the private key with a fingerprint. Compared to the unilateral authentication provided by the typical biometric-based authentication, the proposed solution can provide the mutual authentication. In addition to the increased security, this solution can alleviate the privacy issue of the fingerprint data by conglomerating the fingerprint data with the private key and storing the conglomerated data in a user-carry device such as a smart card. With a 32-bit ARM7-based smart card and a Pentium 4 PC, the proposed fingerprint-based PKI authentication can be executed within 1.3second.

• 유기물자원화
• /
• 제13권2호
• /
• pp.65-73
• /
• 2005

생활폐기물 종량제 도입에 따라 음식물류폐기물 처리가 점차 사회적 문제로 대두되면서 '98년 9월에 3개 부처(환경부, 농림부, 보건복지부)공동으로 '음식물쓰레기 5개년 자원화 기본계획'을 수립하는 등 폐기물 처리체계와 구별하여 음식물류폐기물 관리체계를 확립하고 자원 순환형 관리구조로 변화를 모색하게 되었다.

• ETRI Journal
• /
• 제43권5호
• /
• pp.775-786
• /
• 2021

To translate in real time, a simultaneous translation system should determine when to stop reading source tokens and generate target tokens corresponding to a partial source sentence read up to that point. However, conventional attention-based neural machine translation (NMT) models cannot produce translations with adequate latency in online scenarios because they wait until a source sentence is completed to compute alignment between the source and target tokens. To address this issue, we propose a reinforced learning (RL)-based attention mechanism, the reinforced attention mechanism, which allows a neural translation model to jointly train the stopping criterion and a partial translation model. The proposed attention mechanism comprises two modules, one to ensure translation quality and the other to address latency. Different from previous RL-based simultaneous translation systems, which learn the stopping criterion from a fixed NMT model, the modules can be trained jointly with a novel reward function. In our experiments, the proposed model has better translation quality and comparable latency compared to previous models.

• 교육시설 논문지
• /
• 제27권2호
• /
• pp.69-78
• /
• 2020

Interest in the school environment design is increasing as the project to restructure school space and to become common in school through smart devices. However, no research has been presented on the current status of studies on school environmental design, and the need for analysis of recent research trends in school environmental design is required. In response, this study aims to provide basic data on school environmental design by analysing the trends of domestic studies of school environmental design since 2010 and to present the direction of research on future school environmental design. A total of 312 papers related to school architecture and spatial design are set for study in the results of academic information search on school environmental design. Based on this, frequency analysis and cross-sectional analysis are conducted based on the classification table according to the year of issue, research subject, research purpose, and research method. The statistical significance of the research subjects was secured in accordance with the year of issue and subject to study, and accordingly the key research key words were derived. It is meaningful that through this study, the current status of school environmental design research over the last 10 years can be analyzed and the direction for the future school environmental design research has been suggested.

• 한국콘텐츠학회논문지
• /
• 제20권6호
• /
• pp.556-578
• /
• 2020

실시간으로 발생하는 뉴스 기사로부터 이슈를 분석하기 위한 다양한 연구가 진행되어 왔다. 하지만 범주에 따라 계층적으로 이슈를 분석하는 연구는 많이 진행되지 않았고, 계층적 이슈 분석을 위한 기존의 연구에서 제안하는 방식 또한 뉴스 기사 증가에 따라 군집화 속도가 느려지는 문제점이 있다. 따라서 본 논문에서는 준 실시간으로 뉴스 기사의 이슈를 분석하는 계층적·점증적 군집화 방식을 제안한다. 제안하는 군집화 방식은 샴 신경망을 이용한 가중 코사인 유사도 측정 모델 기반의 k-평균 알고리즘을 이용한 단어 군집 기반 문서 표현 방식을 통해 뉴스 기사를 문서 벡터로 표현한다. 그리고 문서 벡터로부터 초기 이슈 군집 트리를 생성하고, 새로 발생한 뉴스 기사를 해당 이슈 군집 트리에 추가하는 점증적 군집화 방식을 제안함으로써 뉴스 기사의 계층적 이슈를 준 실시간으로 분석한다. 마지막으로, 본 논문에서 제안하는 방식과 기존 방식들과의 성능평가를 통해 제안하는 군집화 방식이 정확도 측면에서 기존 방식 대비 NMI 지표 기준 0.26 정도 성능이 향상되었고, 속도 측면에서 약 10배 이상의 성능이 향상됨을 입증하였다.

• 공학교육연구
• /
• 제7권4호
• /
• pp.22-31
• /
• 2004

이미 여러 분야에서 모집단의 실향에 대한 질적 평가 방법 중 우수하다고 정립이 된 조사기법인 '초점그룹(focus group)'을 공학교육의 학습 성과의 성취도 평가에 이용하려면, 본 기법에 대한 정확한 이해를 바탕으로 우리 교육 실정과 공학 교육의 특징을 살펴 적용하여야 한다. 본 연구는 '초점그룹'에 대한 소개와 함께, 공학 교육의 학습 성과 평가에 활용할 때 제기될 수 있는 문제점들을 파악하고 대안을 제시함으로서, 학습 성과 평가의 새로운 기법으로써 '초점그룹'을 활용하고자 하였다. 조사 기획, 실행, 분석 단계에서 일어날 수 있는 문제점들을 정리하고 가능한 해결 방안에 대해 연구하였다. 실제 공학교육에 적용을 위한 실행에 이용할 수 있도록 조사 지침서의 간단한 예도 연구 제시하였다.

• 지식경영연구
• /
• 제18권4호
• /
• pp.287-306
• /
• 2017

This research focus on the effect of developers' participation structure in knowledge creation and knowledge sharing activities in open source software development projects. Based on preferential selection theory, hypotheses of relationship between a developers' concentration of knowledge creation/sharing activities and collaboration performance was derived. To verify the hypotheses, we use the Gini coefficient in the commit contribution of the developers (knowledge creation) and the centralization index in the repository issue network (knowledge sharing network). Using social network analysis, this paper calculates centralization index from developers in the issue boards in each repository based on data from 837 repositories in GitHub, a leading open source software development platform. As a result, instead of all developers creating and sharing knowledge equally, only a few of developers creating and sharing knowledge intensively further improve the performance of the open collaboration. In other words, a few developers predominantly providing commit and actively responding to issues raised from other developers enhance the project performance. The results of this study are expected to be used by developers who manage open source software project as a governance strategy, which could improve the performance of open collaboration.