• Journal of the Korea Computer Industry Society
• /
• v.5 no.1
• /
• pp.101-108
• /
• 2004

As the dependency to network system and demands of efficient storage systems rapidly grows in every networking filed, the current trends initiated by explosive networked data grow due to the wide-spread of internet multimedia data and internet requires a paradigm shift from computing-centric to data-centric in storagesystems. Furthermore, the new environment of file systems such as NAS(Network Attached Storage) and SAN(Storage Area Network) is adopted to the existing storage paradigm for Providing high availability and efficient data access. We describe the design issues and system components of SANiqueTM, which is the cluster file system based on SAN environment. SANiqueTM has the capability of transferring the user data from the network-attached SAN disk to client applications directly We, especially, present the protocol and functionality of the global buffer manager in our cluster file system.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.689-691
• /
• 2018

의존 구문 분석은 자연어 이해 영역의 대표적인 과제 중 하나이다. 본 논문에서는 한국어 의존 구분 분석의 성능 향상을 위해 Deep Bi-affine Network 와 스택 포인터 네트워크의 앙상블 모델을 제안한다. Bi-affine 모델은 그래프 기반 방식, 스택 포인터 네트워크의 경우 그래프 기반과 전이 기반의 장점을 모두 사용하는 모델로 서로 다른 모델의 앙상블을 통해 성능 향상을 기대할 수 있다. 두 모델 모두 한국어 어절의 특성을 고려한 자질을 사용하였으며 세종 의존 구문 분석 데이터에 대해 UAS 90.60 / LAS 88.26(Deep Bi-affine Network), UAS 92.17 / LAS 90.08(스택 포인터 네트워크) 성능을 얻었다. 두 모델에 대한 앙상블 기법 적용시 추가적인 성능 향상을 얻을 수 있었다.

• Transactions of Materials Processing
• /
• v.27 no.4
• /
• pp.227-235
• /
• 2018

The flow stresses have been identified prior to a numerical simulation for predicting a deformation of materials using the experimental or analytical analysis. Recently, the flow stress models considering the temperature effect have been developed to reduce the number of experiments. Artificial neural network can provide a simple procedure for solving a problem from the analytical models. The objective of this paper is the prediction of flow stress on the fiber metal laminate using the artificial neural network. First, the training data were obtained by conducting the uniaxial tensile tests at the various temperature conditions. After, the artificial neural network has been trained by Levenberg-Marquardt method. The numerical results of the trained model were compared with the analytical models predicted at the previous study. It is noted that the artificial neural network can predict flow stress effectively as compared with the previously-proposed analytical models.

• Journal of Communications and Networks
• /
• v.18 no.6
• /
• pp.938-947
• /
• 2016

A mobile ad hoc network (MANET) is a collection of mobile nodes communicating in an infrastructure-less environment without the aid of a central administrating authority. Such networks entail greater dependency on synergy amongst the nodes to execute fundamental network operations. The scarcity of resources makes it economically logical for nodes to misbehave to preserve their resources which makes secure routing difficult to achieve. To ensure secure routing a mechanism is required to discourage misbehavior and maintain the synergy in the network. The proposed scheme employs a partially distributed dynamic model at each node for enhancing the security of the network. Supplementary information regarding misbehavior in the network is partially distributed among the nodes during route establishment which is used as a cautionary measure to ensure secure routing. The proposed scheme contemplates the real world scenario where a node may exhibit different kinds of misbehavior at different times. Thus, it provides a dynamic decision making procedure to deal with nodes exhibiting varying misbehaviors in accordance to their severity. Simulations conducted to evaluate the performance of the model demonstrate its effectiveness in dealing with misbehaving nodes.

• Journal of the Korean Physical Society
• /
• v.73 no.10
• /
• pp.1589-1595
• /
• 2018

Similarity index measures the topological proximity of node pairs in a complex network. Numerous similarity indices have been defined and investigated, but the dependency of structure on the performance of similarity indices has not been sufficiently investigated. In this study, we investigated the relationship between the performance of similarity indices and structural properties of a network by employing a two-state random network. A node in a two-state network has binary types that are initially given, and a connection probability is determined from the state of the node pair. The performances of similarity indices are affected by the number of links and the ratio of intra-connections to inter-connections. Similarity indices have different characteristics depending on their type. Local indices perform well in small-size networks and do not depend on whether the structure is intra-dominant or inter-dominant. In contrast, global indices perform better in large-size networks, and some such indices do not perform well in an inter-dominant structure. We also found that link prediction performance and the performance of similarity are correlated in both model networks and empirical networks. This relationship implies that link prediction performance can be used as an approximation for the performance of the similarity index when information about node type is unavailable. This relationship may help to find the appropriate index for given networks.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2002.05a
• /
• pp.239-242
• /
• 2002

멀티미디어 환경에서 많은 패킷 스위치 네트워크로부터 발생되는 트래픽은 burstiness성질이 상당히 넓은 범위의 time scale상에 존재하고, 트래픽 특성이 self-similar현상을 보이고 있다는 것이 알려지고 있다. 본 논문에서는 Shared-Buffer를 이용한 ATM 스위치 buffer관리방법을 적용하여 입력포트에 self-similar 트래픽이 들어올 때 출력포트의 long-range dependency의 변화를 분석하며 아울러 큐잉 지연, 셀 손실률 그리고 이용도(utilization)등이 self-similar 트래픽에 미치는 영향을 분석할 것이다.

• Journal of the Korea Safety Management & Science
• /
• v.7 no.4
• /
• pp.219-227
• /
• 2005

Three general algorithms for evaluating the reliability for complex bridge system are proposed. These methods, such as Keystone, Boolean, Network algorithms are powerful and effective to derive an reliability expression for many practical complex systems. The combination approach of RBD and FTA proposed in this paper provides an effective way to evaluate the functional dependency for applications of FMEA.

• Journal of Information Management
• /
• v.36 no.4
• /
• pp.199-215
• /
• 2005

Experts participating in the knowledge expert network externalize their implicit knowledge by providing information or writing reports. Almost all the members of the network share externalized knowledge and the network facilitate the dissemination and diffusion of knowledge. Individuals reproduce another implicit knowledge by internalizing shared knowledge through the network and re-created knowledge is externalized, establishing knowledge circulation. In this paper, we analyze the expert groups of the Global Network of Korean Scientists & Engineers(KOSEN, www.kosen21.org), the Korea's No. 1 science and engineering knowledge expert community, with the application of the theory of policy network proposed by Marsh & Rhodes. According to the principal standards of policy network classification such as the number of participants, interaction among participants, consistency, distribution of resources and dependency, we categorize the KOSEN expert groups as closed policy network and opened issue network, and divide closed policy network into core community and periphery community.

• Journal of Contemporary Eastern Asia
• /
• v.15 no.1
• /
• pp.41-63
• /
• 2016

Energy policy is known to have higher path dependency among policy fields (Kuper and van Soest, 2003; OECD, 2012; Kikkawa, 2013) and is a critical component of the infrastructure development undertaken in the early stages of nation building. Actor roles, such as those played by interest groups, are firmly formed, making it unlikely that institutional change can be implemented. In resource-challenged Japan, energy policy is an especially critical policy area for the Japanese government. In comparing energy policy making in Japan and Germany, Japan’s policy community is relatively firm (Hartwig et al., 2015), and it is improbable that institutional change can occur. The Japanese government’s approach to energy policy has shifted incrementally in the past half century, with the most recent being the 2012 implementation of the “Feed-In Tariff Law” (Act on Special Measures Concerning Procurement of Renewable Electric Energy by Operators of Electric Utilities), which encourages new investment in renewable electricity generation and promotes the use of renewable energy. Yet, who were the actors involved and the factors that influenced the establishment of this new law? This study attempts to assess the factors associated with implementing the law as well as the roles of the relevant major actors. In answering this question, we focus on identifying the policy networks among government, political parties, and interest groups, which suggests that success in persuading key economic groups could be a factor in promoting the law. Our data is based on the “Global Environmental Policy Network Survey 2012-2013 (GEPON2)” which was conducted immediately after the March 11, 2011 Great East Japan Earthquake with respondents including political parties, the government, interest groups, and civil society organizations. Our results suggest that the Feed in Tariff (FIT) Law’s network structure is similar to the information network and support network, and that the actors at the center of the network support the FIT Law. The strength of our research lays in our focus on political networks and their contributing mechanism to the law’s implementation through analysis of the political process. From an academic perspective, identifying the key actors and factors may be significant in explaining institutional change in policy areas with high path dependency. Close examination of this issue also has implications for a society that can promote renewable and sustainable energy resources.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2001.10a
• /
• pp.165-196
• /
• 2001

DNA chip 또는 microarray는 다수의 유전자 또는 유전자 조각을 (보통 수천내지 수만 개)칩상에 고정시켜 놓고 DNA hybridization 반응을 이용하여 유전자들의 발현 양상을 분석할 수 있는 기술이다. 이러한 high-throughput기술은 예전에는 생각하지 못했던 여러가지 분자생물학의 문제에 대한 해답을 제시해 줄 수 있을 뿐 만 아니라, 분자수준에서의 질병 진단, 신약 개발, 환경 오염 문제의 해결 등 그 응용 가능성이 무한하다. 이 기술의 실용적인 적용을 위해서는 DNA chip을 제작하기 위한 하드웨어/웻웨어 기술 외에도 이러한 데이터로부터 최대한 유용하고 새로운 지식을 창출하기 위한 bioinformatics 기술이 핵심이라고 할 수 있다. 유전자 발현 패턴을 데이터마이닝하는 문제는 크게 clustering, classification, dependency analysis로 구분할 수 있으며 이러한 기술은 통계학과인공지능 기계학습에 기반을 두고 있다. 주로 사용된 기법으로는 principal component analysis, hierarchical clustering, k-means, self-organizing maps, decision trees, multilayer perceptron neural networks, association rules 등이다. 본 세미나에서는 이러한 기본적인 기계학습 기술 외에 최근에 연구되고 있는 새로운 학습 기술로서 probabilistic graphical model (PGM)을 소개하고 이를 DNA chip 데이터 분석에 응용하는 연구를 살펴본다. PGM은 인공신경망, 그래프 이론, 확률 이론이 결합되어 형성된 기계학습 모델로서 인간 두뇌의 기억과 학습 기작에 기반을 두고 있으며 다른 기계학습 모델과의 큰 차이점 중의 하나는 generative model이라는 것이다. 즉 일단 모델이 만들어지면 이것으로부터 새로운 데이터를 생성할 수 있는 능력이 있어서, 만들어진 모델을 검증하고 이로부터 새로운 사실을 추론해 낼 수 있어 biological data mining 문제에서와 같이 새로운 지식을 발견하는 exploratory analysis에 적합하다. 또한probabilistic graphical model은 기존의 신경망 모델과는 달리 deterministic한의사결정이 아니라 확률에 기반한 soft inference를 하고 학습된 모델로부터 관련된 요인들간의 인과관계(causal relationship) 또는 상호의존관계(dependency)를 분석하기에 적합한 장점이 있다. 군체적인 PGM 모델의 예로서, Bayesian network, nonnegative matrix factorization (NMF), generative topographic mapping (GTM)의 구조와 학습 및 추론알고리즘을소개하고 이를 DNA칩 데이터 분석 평가 대회인 CAMDA-2000과 CAMDA-2001에서 사용된cancer diagnosis 문제와 gene-drug dependency analysis 문제에 적용한 결과를 살펴본다.