• Journal of KIISE:Software and Applications
• /
• v.30 no.11
• /
• pp.1072-1079
• /
• 2003

This study is purported for suggesting a new clustering algorithm that enables incremental categorization of numerous documents. The suggested algorithm adopts the natures of the spherical k-means algorithm, which clusters a mass amount of high-dimensional documents, and the fuzzy ART(adaptive resonance theory) neural network, which performs clustering incrementally. In short, the suggested algorithm is a combination of the spherical k-means vector space model and concept vector and fuzzy ART vigilance parameter. The new algorithm not only supports incremental clustering and automatically sets the appropriate number of clusters, but also solves the current problems of overfitting caused by outlier and noise. Additionally, concerning the objective function value, which measures the cluster's coherence that is used to evaluate the quality of produced clusters, tests on the CLASSIC3 data set showed that the newly suggested algorithm works better than the spherical k-means by 8.04% in average.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.1
• /
• pp.17-25
• /
• 2018

Recently, with the development of ICT technology and the rapid spread of smart devices, a huge amount of information is being generated. The recommendation system has helped the informant to judge the information from the information overload, and it has become a solution for the information provider to increase the profit of the company and the publicity effect of the company. Recommendation systems can be implemented in various approaches, but social information is presented as a way to improve performance. However, no research has been done to utilize trust cluster information among users in the recommendation system. In this paper, we propose a method to improve the performance of the recommendation system by using the influence between the intra-cluster objects and the information between the trustor-trustee in the cluster generated in the online review. Experiments using the proposed method and real data have confirmed that the prediction accuracy is improved than the existing methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37C no.12
• /
• pp.1328-1336
• /
• 2012

Compared to terrestrial RF communications, underwater acoustic communications have several limitations such as limited bandwidth, high level of fading effects, and a large underwater propagation delay. In this paper, in order to tackle those limitations of underwater communications and to make it possible to form a large underwater monitoring systems, we propose a hierarchical underwater network architecture, which consists of underwater sensors, clusterheads, underwater/surface sink nodes, autonomous underwater vehicles (AUVs). In the proposed architecture, for the maximization of packet delivery ratio and the minimization of underwater sensor's energy consumption, a hybrid routing protocol is used. More specifically, cluster members use Tree based routing to transmit sensing data to clusterheads. AUVs on optimal trajectory movements collect the aggregated data from clusterhead and finally forward the data to the sink node. Also, in order to minimize the maximum travel distance of AUVs, an Integer Linear Programming based algorithm is employed. Performance analysis through simulations shows that the proposed architecture can achieve a higher data delivery ratio and lower energy consumption than existing routing schemes such as gradient based routing and geographical forwarding. Start after striking space key 2 times.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11b
• /
• pp.229-231
• /
• 2005

S-System 모델은 동적 시스템을 기술하기 위한 여러 모델 중의 하나로서, 높은 표현력으로 인해 다양한 분야에서 사용되어져 오고 있다. 하지만 S-System 모델이 갖고 있는 많은 매개변수는 목표 시스템을 모델링하는 데에 있어 고려해야 할 탐색 공간의 넓이를 크게 증가시키는 단점을 갖고 있으며 그로 인해 고려될 수 있는 변수의 수는 극히 적은 수로 제한되어져 왔다. 일반적인 S-System 모델의 경우, ${_n}$개의 변수로 이루어진 시스템을 모델링하는 데 결정되어져야 할 매개변수의 수는 $O(n^2)$이다. 본 논문에서는 시스템 내의 변수들을 서로간의 연관 정도에 따라 클러스터링하고. 클러스터 사이의 동적 모델링을 통해 고려하는 매개변수의 수를 O(kn) $(k{\leq}n)$으로 줄이는 방법을 제안한다. 매개변수 값의 탐색을 위해 유전자 알고리즘을 사용하며, 제안된 방법이 기존의 방법으로는 학습할 수 없었던 규모의 S-System 모델을 학습할 수 있는 가능성을 지님을 보인다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.204-207
• /
• 2015

이 연구는 공학 및 실험과학에서 활용되는 몬테 칼로 시뮬레이션 기법 중 하나인 무작위 행보 알고리즘의 성능 개선을 목표로 하였다. 이를 위해 무작위 행보 과정에서 난수 발생부와 행보 진행부를 분리하여 처리 시간을 단축하는 방안과, 문제 영역의 계산 규모를 2단계로 분할하여 시뮬레이션의 수렴 속도를 향상 시키는 방안을 제안한다. 또한 대규모 문제를 병렬처리 가능하도록 구현하고, 서로 다른 작업 분할 방식을 혼합하여 최적화를 수행 하였다. 순차 알고리즘만으로 실험한 결과 단순 구현방법과 비교해 실행시간과 에너지 소모량이 각각 18%의 성능향상을 얻었으며, 병렬 알고리즘을 8개의 노드(16코어)의 클러스터에서 실행했을 때 행 분할 방식의 성능이 블록 분할 방식보다 8% 빨라지는 것을 확인하였다.

• The KIPS Transactions:PartA
• /
• v.10A no.6
• /
• pp.635-642
• /
• 2003

The radical development of recent network technology provides the basic foundation which can establish a high speed and cheap cluster system. It is a general trend that conventional cluster systems are built as the system over a fixed level based on stabilized and high speed local networks. A multi-distributed web cluster group is a web cluster model which can obtain high performance, high efficiency and high availability through mutual cooperative works between effective job division and system nodes through parallel performance of a given work and shared memory of SC-Server with low price and low speed system nodes on networks. For this, multi-distributed web cluster group builds a sub-cluster group bound with single imaginary networks of multiple system nodes and uses the web distributed shared memory of system nodes for the effective data transmission within sub-cluster groups. Since the presented model uses a load balancing and parallel computing method of large-scale work required from users, it can maximize the processing efficiency.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.6
• /
• pp.237-247
• /
• 2008

A virtual screening is the process of reducing an unmanageable number of compounds to a limited number of compounds for the target of interest by means of computational techniques such as molecular docking. And it is one of a large-scale scientific application that requires large computing power and data storage capability. Previous applications or softwares for molecular docking such as AutoDock, FlexX, Glide, DOCK, LigandFit, ViSION were developed to be run on a supercomputer, a workstation, or a cluster-computer. However the virtual screening using a supercomputer has a problem that a supercomputer is very expensive and the virtual screening using a workstation or a cluster-computer requires a long execution time. Thus we propose a service-based virtual screening system using Grid computing technology which supports a large data intensive operation. We constructed 3-dimensional chemical molecular database for virtual screening. And we designed a resource broker and a data broker for supporting efficient molecular docking service and proposed various services for virtual screening. We implemented service based virtual screening system with DOCK 5.0 and Globus 3.2 toolkit. Our system can reduce a timeline and cost of drug or new material design.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.11a
• /
• pp.863-866
• /
• 2010

무선 센서 네트워크의 규모가 커짐에 따라 대규모의 센서 네트워크의 활용에 대한 기대가 커지고 있다. 기존 IEEE 802.15.4 를 이용한 네트워크의 규모에 따른 한계를 극복하기 위해 듀얼 라디오 센서 네트워크에 대한 많은 연구가 진행되고 있다. 듀얼 라디오 센서 네트워크는 단일 라디오를 내장한 전통적인 센서 노드(싱글 라디오 노드)에 듀얼 라디오 인터페이스(IEEE 802.11g, IEEE 802.15.4)를 내장한 센서(듀얼 라디오 노드)를 포함하여 클러스터를 구성한다. 본 논문에서는 이러한 듀얼 라디오 환경에서 효과적으로 동작하는 라우팅 프로토콜을 제안한다. 듀얼 라디오 노드는 네트워크 상에서 클러스터 헤드 역할을 하고 상위 계층을 이루며, 일반 노드는 클러스터 멤버 역할을 하는 하위 계층을 이룬다. 본 논문에서는 각 계층의 네트워크가 나타내는 특징을 이동성이 적거나 높은 네트워크의 특징으로 대입하여, 상위 계층에는 Pro-Active 프로토콜, 하위 계층에는 Re-active 프로토콜이 효과적임을 설명한다. 이를 바탕으로 듀얼 라디오 센서 네트워크에 적합한 라우팅 프로토콜로써 Pro-Active 프로토콜을 대표하는 DSDV 와 Re-Active 프로토콜을 대표하는 AODV 를 조합한 하이브리드 라우팅 프로토콜을 제안한다. 제안한 프로토콜의 성능을 확인하기 위해 센서네트워크 테스트베드를 구성하여 라우팅 복구 시간과 패킷 전송 안정성에 대한 성능을 입증한다.

• Journal of Korea Port Economic Association
• /
• v.32 no.4
• /
• pp.41-56
• /
• 2016

China is pushing a policy called China's Western Development in order to balance the continent's development level. They are connecting the rich natural resources of the mid-western areas and the capital of the eastern coastal areas of China. Since Korea depends on China the most for trade, it is necessary for Korea to always consider the trade scale and pattern changes of China. Response to changes in the transport and logistics system between the two countries is also demanded. This study aims to analyze the import and export trends between Korean and Chinese provinces of each item from the trade scale point of view. China's global trade routes reorganization and China's western development are considered and the corresponding direction is presented accordingly. The study also suggests strategies to expand the import and export scale with China based on the analysis.

• KIISE Transactions on Computing Practices
• /
• v.21 no.1
• /
• pp.13-18
• /
• 2015

Computational science simulations have been used to enable discovery in a broad spectrum of application areas, these simulations show irregular demanding characteristics of computing resources from time to time. The adoption of virtualized high performance cloud, rather than CPU-centric computing platform (such as supercomputers), is gaining interest of interests mainly due to its ease-of-use, multi-tenancy and flexibility. Basically, provisioning a virtual cluster, which consists of a lot of virtual machines, in a real-time has a critical impact on the successful deployment of the virtualized HPC clouds for computational science simulations. However, the cost of concurrently creating many virtual machines in constructing a virtual cluster can be as much as two orders of magnitude worse than expected. One of the main factors in this bottleneck is the time spent to create the virtual images for the virtual machines. In this paper, we propose a novel technique to minimize the creation time of virtual machine images and improve I/O performance of the provisioned virtual clusters. We also confirm that our proposed technique outperforms the conventional ones using various sets of experiments.