• Journal of KIISE:Computing Practices and Letters
• /
• v.10 no.2
• /
• pp.156-173
• /
• 2004

Business process should support the execution of collaboration process with agility and flexibility through the integration of enterprise inner or outer applications and human resources from the collaborative workflow view. Although the dependency of enterprise activities to the automated system has been increasing, human role is as important as ever. In the workflow modelling this human role is emphasized and the structure to control human error by analysing decision-making itself is needed. Also, through the collaboration of activities agile and effective communication should be constructed, eventually by the combination and coordination of activities to the aimed process the product quality should be improved. This paper classifies human errors can be occurred in collaborative workflow by applying GEMS(Generic Error Modelling System) to control them, and suggests human error control method through hybrid based modelling as well. On this base collaborative workflow modeling tool is designed and implemented. Using this modelling methodology it is possible to workflow modeling could be supported considering human characteristics has a tendency of human error to be controlled.

• Journal of KIISE:Software and Applications
• /
• v.26 no.11
• /
• pp.1288-1297
• /
• 1999

오늘날 많은 워크플로우 응용들은 마감 시간과 같은 시간 제약 조건들의 효율적인 관리를 요구하고 있다. 이러한 시간 제약성 워크플로우의 가장 중요한 목적들 중의 하나는 마감 시간을 만족하는 워크플로우 인스턴스의 수를 최대화하는 것이다. 시간 제약성 워크플로우 처리의 성능을 개선하기 위해서는 효과적인 액티비티들의 마감 시간 할당 방법이 요구된다. 이에 본 논문에서는 액티비티의 체재 시간과 워크플로우 제어 구조들의 특성을 고려하여 효과적인 마감 시간 할당 방법인 PST를 제안한다. PST는 기존의 방법들과 달리 워크플로우 제어 흐름을 분석하여 액티비티의 마감 시간을 할당하기 때문에 과부하에서도 높은 성능을 지원할 수 있다. 부가적으로 기존의 방법들과 성능 평가를 통해 본 논문에서 제안된 방법의 효용성을 검증한다.Abstract Nowadays, many workflow applications are requiring efficient time management such as deadline. With time-constrained workflows, one of the most important goals is to maximize the number of workflow instances that satisfy the deadline. In this regard, effective deadline assignment mechanisms should be developed to achieve high performance of time-constrained workflow processing. We propose in this paper our novel deadline assignment method called PST based on both an activity's sojourn time and the characteristics of workflow control structures. Because the method allocates the deadline to activities from analyzing the control flow of workflow, it can support good and stable workflow performance even under the high workload. In addition, we validate the usefulness of our proposed method through various experiments.

• Annual Conference of KIPS
• /
• 2010.04a
• /
• pp.808-811
• /
• 2010

최근 네트워크를 구성하는 기반 시설의 성능이 향상됨에 따라 대량의 트래픽에 대한 네트워크 침입탐지 시스템의 성능을 향상시키기 위한 연구가 진행되고 있다. 대규모 네트워크에서는 단일 시스템으로 네트워크 내의 모든 트래픽을 분석하는 것이 불가능하므로 병렬 구조 NIDS를 도입하여야 하는데, 이를 위해서는 병렬 구조를 이루는 각 NIDS 노드로의 부하 분산이 필요하다. 플로우 기반 부하 분산 기법은 이러한 부하 분산 기법 중 하나로, TCP 세그먼트의 재조합으로 인해 발생하는 통신 오버헤드를 줄일 수 있어 효율적이다. 본 논문에서는 네트워크 트래픽의 특성과 각 노드의 성능을 고려하여 플로우 기반 부하 분산이 효율적으로 이루어질 수 있는 방안을 제안한다.

• Proceedings of the CALSEC Conference
• /
• 2004.02a
• /
• pp.261-265
• /
• 2004

□목적 ·플랫폼에 독립적인 기업간 상호운영성 방법론의 획득 ·기업간 워크프로우 통합에 필요한 관련 기술 및 표준의 분석 ·기업간 정보시스템 통합 환경을 위한 시스템 구조 제공 ·협업 프로세스의 구축 및 실행 방안 제공 ·XML 기반 기업간 상호운영성의 획득에 필요한 기술적 사항 제공(중략)

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.8
• /
• pp.2370-2387
• /
• 2000

This paper proposes a new design strategy that can overcome limitations of two existing approaches, focusing on performance and scalability issues among various architectural issues which must be considered in designing workflow systems. In the proposed approach, we suggest to introduce 'delegation model' which corresponds to the high-level strategy utilizing the fundamental characteristics of mobile agents, after establishing a mobile agent-based workflow system architecture focused on performance and scalability. We point out three major issues that should be considered in this approach and propose a 3-tier mobile agent-based workflow system architecture that effectively reflects these considerations. Also, we show that the new approach can provide better performance and scalability than existing ones - approaches based on the client-server paradigm and other approaches based on mobile agent paradigm represented by DartFlow system - by evaluating performance of the proposed architecture through the GSPN simulation.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.8 no.3
• /
• pp.541-548
• /
• 2007

The workflow, one of core features for modern business operation, models common elements and their relationships in business processes. The research is focused on how to adapt the workflow model to the healthcare examination which becomes popular these days. Especially, we intend to have fully automated process for healthcare examination by providing customers with PDA with RFTD chip. In order to realize this goal, this paper proposes the design of software architectures the class structures, and the definition of messages used for the system interactions. In other word, the paper shows the analysis of healthcare application in terms of service scenario and the design of workflow engine ant ubiquitous healthcare application under the Object-Oriented programming environment. the The result of the study may also enable the unmanned healthcare examination as well as ubiquitous based healcare examination sooner or later.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.21 no.6
• /
• pp.535-541
• /
• 2008

As the concurrent engineering concept has emerged along with the support of optimization techniques, lots of endeavors have been made to apply optimization techniques to actual design problems for a holistic decision. Even if the range of design problems which the optimization is applicable to has been extended, most of ship designs still remain in an iterative approach due to the difficulties of seamless integration of all related design activities. In this approach, an entire design problem is divided into many sub-problems and carried out by many different disciplines through complicated internal interactions. This paper focuses on preliminary ship design process. This paper proposes a process centric integrated framework as the first step to establish a workflow based design process management system. The framework consists of two parts; a schedule management part to support a manager to monitor current progress status and adjust current schedule, and a process management part to assist a design to effectively perform a series of design activities by following a predefined procedure. Overall system are decomposed into modules according to the target to be managed in each module. Appropriate interactions between the decomposed modules are designed to achieve a consistency of the entire system. Design process model is also designed on a thorough analysis of actual ship design practice. The proposed framework will be embodied using a commercial workflow package.

• Proceedings of the Korea Contents Association Conference
• /
• 2016.05a
• /
• pp.53-54
• /
• 2016

빅데이터란 대용량 데이터 활용 및 분석을 통해 가치 있는 정보를 추출하고, 이를 바탕으로 대응 방안 도출 또는 변화를 예측하는 기술을 의미한다. 그리고 빅데이터 분석에 활용되는 데이터인 페이스북과 같은 소셜 데이터, 유전자 발현과 같은 바이오 데이터, 항공망과 같은 지리정보 데이터들은 대용량 네트워크로 구성되어 있다. 네트워크 클러스터링은 서로 유사한 특성을 갖는 네트워크 내의 데이터들을 동일한 클러스터로 묶는 기법으로 네트워크 데이터를 분석하고 그 특성을 파악하는데 폭넓게 사용된다. 최근 빅데이터가 다양한 분야에서 활용되면서 방대한 양의 네트워크 데이터가 생성되고 있고, 이에 따라서 대용량 네트워크 데이터를 효율적으로 처리하는 클러스터링 기법의 중요성이 증가하고 있다. MCL(Markov Clustering) 알고리즘은 플로우 기반 무감독(unsupervised) 클러스터링 알고리즘으로 확장성이 우수해 다양한 분야에서 활용되고 있다. 하지만, MCL은 대용량 네트워크에 대해서는 많은 클러스터링 연산을 요구하며 너무 많은 클러스터를 생성하는 문제를 갖는다. 본 논문에서는 네트워크 압축을 기반으로 한 클러스터링 알고리즘을 제안함으로써 MCL보다 클러스터링 속도와 정확도를 향상시켰다. 또한, 희소행렬을 효율적으로 저장하는 CSC(Compressed Sparse Column) 자료구조와 MapReduce 기법을 제안한 클러스터링 알고리즘에 적용함으로써 대용량 네트워크에 대한 클러스터링 속도를 향상시켰다.

• Journal of Platform Technology
• /
• v.11 no.4
• /
• pp.71-79
• /
• 2023

In this paper, we analyze the characteristics of machine learning workloads and, based on them, propose a distributed in-memory caching technique to improve the performance of machine learning workloads. The core of machine learning workload is model training, and model training is a computationally intensive task. Performing machine learning workloads in a Kubernetes-based cloud environment in which the computing framework and storage are separated can effectively allocate resources, but delays can occur because IO must be performed through network communication. In this paper, we propose a distributed in-memory caching technique to improve the performance of machine learning workloads performed in such an environment. In particular, we propose a new method of precaching data required for machine learning workloads into the distributed in-memory cache by considering Kubflow pipelines, a Kubernetes-based machine learning pipeline management tool.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.3
• /
• pp.125-138
• /
• 2016

The service identification in service-oriented developments has been conducted by based on workflow, goals, scenarios, usecases, components, features, and patterns. However, the identification of service by semantic approach at the business value view was not detailed yet. In order to enhance accuracy of identifying business service, this paper proposes a method for identifying business service by analyzing syntax and semantics in XL-BPMN model. The business processes based on business scenario are identified, and they are designed in a XL-BPMN business process model. In this business process model, an unit business service is identified through binding closely related activities by the integrated analysis result of syntax patterns and properties-based semantic similarities between activities. The method through XL-BPMN model at upper business levels can identify the reusable unit business service with high accuracy and modularity. It also can accelerate more service-oriented developments by reusing identified services.