• 한국정보과학회논문지:소프트웨어및응용
• /
• 제36권4호
• /
• pp.273-282
• /
• 2009

소프트웨어 프로젝트 데이터를 이용한 각종 분석 예측 모델 생성시 직면하는 문제 중 하나는 데이터에 포함된 결측값이며 이에 대한 효과적인 방안은 결측값 대치 법이다. 대표적인 결측값 대치법인 K 최근접 이웃 대치법은 대치과정에서 결측값을 포함하는 인스턴스의 관측정보를 활용하지 못한다는 단점이 있다. 본 연구에서는 이러한 단점을 극복하기 위해 K 최근접 이웃 대치법과 최대 우도 추정법을 결합한 새로운 소프트웨어 프로젝트 수치 데이터용 결측값 대치법을 제안한다. 또한 결측값 대치법의 정확도를 비교하기 위한 새로운 측도를 함께 제안한다.

• Journal of Information Technology Applications and Management
• /
• 제13권4호
• /
• pp.237-255
• /
• 2006

This study was exploratively performed to demonstrate the management behavior of the public software project in korea. In this study, 167 auditing reports on the public software project which were carried out during $1999{\sim}2003$, were employed and surveyed. Survey reveals that quality management was the most neglected in process of software project management relative to other 12 management areas. And configuration management, risk management, time management, and scope management follow up in order in its vulnerability in software project management.

• Journal of Information Technology Applications and Management
• /
• 제11권4호
• /
• pp.103-120
• /
• 2004

Identifying validated risk factors in software risk management is imperative for project managers. Although validated risk lists were provided by previous researchers, risk list associated with software project performance areas was not provided as yet. This paper represents a first step toward understanding risk lists by various project performance areas (time, cost, and quality) to help project managers alleviating the possibility of software project failure. Four simultaneous exploratory surveys were conducted with 29 experienced software project managers. Three different risk factor ranking sets for each project performance area were compared with, the risk ranking, which was provided without clarifying specific project performance areas. The risk lists and their corresponding perceived importance were different from previous research results. This implies that identifying risk factors for specific project performance areas can provide additional information for project managers. We concluded by discussing implications of our finds for both research and improving risk management practice.

• Journal of Information Technology Applications and Management
• /
• 제11권2호
• /
• pp.15-27
• /
• 2004

Most software projects inevitably involve various types and degrees of uncertainty. Without proper risk assessment and coordination, software projects can easily run out of control and consume significant additional resource. Thus, risk management techniques are critical issues to information system researchers. Previous empirical studies of U.S. software firms support the adoption of development standardization and user requirement analysis techniques in risk-based software project management. Using data collected from software projects developed in Korea during 1999-2000, we conduct a comparative study to determine how risk management strategies impact software product and process performance in countries with dissimilar IT capabilities. In addition, we offer an alternative conceptualization of residual performance risk. We show that the use of residual performance risk as an intervening variable is inappropriate in IT developing countries like Korea where the role of late stage risk control remedies are critical. A revised model is proposed that generates more reliable empirical implications for Korean software projects.

• 한국컴퓨터정보학회논문지
• /
• 제23권12호
• /
• pp.65-72
• /
• 2018

In this paper, we have studied how to reuse common data link software by applying software product line engineering. Existing common data link software performed different stages of design, implementation, and testing without sharing the accumulated knowledge of different developers. In this situation, developers agreed that sharing the assets of each project and reusing the previously developed software would save human and time costs. Even with the initial difficulties, the common Data Link is a continually proposed project in the defense industry, so we decided to build a product line. The common data link software can be divided into two domains. Among them, the initial feature model for the GUI software was constructed, and the following procedure was studied. Through this, we propose a plan to build a product line for core assets and reuse them in newly developed projects.

• 인터넷정보학회논문지
• /
• 제19권2호
• /
• pp.43-48
• /
• 2018

Various artifacts that are produced as software R&D project progresses contain research plan, research report, software requirements and design descriptions, etc. When conducting a software R&D project, it is necessary to confirm that the developed system has implemented its research requirements well. However, various research results make it difficult to design appropriate tests. So, there is a practical need for us to comprehensively handle the planning, execution, and reporting of software test for finding and verifying information related to the research. In this paper, we propose a useful method for software test process in R&D project which supports model based software testing. The proposed method supports automation of test design and generation of test data by explicitly separating each step of System Under Test (SUT). The method utilizes the various models representing the control flow of the function to extract the information necessary for testing the system. And it supports a systematic testing process based on TMMi and ISO 29119. Finally, we show the validity of the method by implementing a prototype with basic functionality to generate test data from software behavioral models.

• 정보처리학회논문지D
• /
• 제9D권2호
• /
• pp.259-266
• /
• 2002

소프트웨어 개발시 중요하게 제기되는 문제점으로 소프트웨어 생명주기의 초기단계에서 개발에 투입될 노력과 비용을 추정하는 능력이다. 측정된 소프트웨어 규모인 기능점수 (FP, Function Point)로부터 프로젝트 개발에 소요되는 노력과 비용을 추정하는 회귀모델들이 제안되었다. 그러나 이들 제안된 모델들은 사용한 표본의 크기가 작고, 과거에 개발된 프로젝트를 대상으로 하고 있다. 1990년대에 개발된 789개 소프트웨어 개발 프로젝트들에 투입된 개발노력에 이들 모델을 적용한 결과 결정계수 $R^2$가 0.53 이하로 데이터의 변동을 설명하였다. 본 논문은 프로젝트 인도율(PDR)을 이용하여 동질성을 갖는 그룹으로 분류하여 개발노력을 추정하는 모델을 제안하였다. 제안된 모델은 랜덤한 잔차 분포를 갖고 있고, 대부분의 PDR 범위에서 0.99이상의 결정계수로 데이터의 변동을 설명하였다.

• 소프트웨어공학소사이어티 논문지
• /
• 제26권4호
• /
• pp.77-92
• /
• 2013

소프트웨어 프로젝트 계획 생성 과정은 (1)프로젝트를 수행하기 위한 작업 구조(WBS)를 작성하고, (2)각 작업에 필요한 공수를 예측한 뒤, (3)작업에 인력을 할당하여, (4)전체 일정을 예측하는 과정으로 이루어진다. 프로젝트의 규모가 커질수록 가능한 작업 구조, 공수, 인력 할당의 조합의 수가 급격히 많아지며 이에 따라 프로젝트 계획 생성 과정의 복잡도가 매우 높아지게 된다. 따라서 이를 지원하기 위한 프로젝트 계획 생성 지원 기법이 필요하다. 본 연구에서는 실무 전문가 그룹과의 논의를 통해 소프트웨어 프로젝트 계획 생성 지원 기법에서 고려해야 할 여러 실무 요구사항들을 도출했다. 도출된 실무 요구사항을 고려하여 개발된 프로젝트 계획 생성 지원 도구 APP(Automatic Project Planner)는 개발 조직의 과거 지식 데이터를 활용한 공수 예측을 지원하며, 실무 이슈가 고려된 자동 인력 할당을 제공한다. 본 도구를 통해 합리적이고 현실적인 프로젝트 계획의 기반을 마련할 수 있다.

• Asia pacific journal of information systems
• /
• 제13권2호
• /
• pp.199-217
• /
• 2003

Based on the IS literature on software project risk management, we developed an integrated model to investigate the risk factors and risk management factors in project development. We also analyzed the interrelation between project risk and project risk management. The questionaries are collected from 83 project leaders. We tested reliability and validity of the measure and analyzed the obtained data. The results support our risk-based hypothesis that shows the importance of risk management in reducing project risks and improving performance of project and process.

• Asia pacific journal of information systems
• /
• 제16권3호
• /
• pp.143-158
• /
• 2006

Identifying risk factors in software risk management is imperative for project managers. The purpose of this paper is to provide software project risk factors validated by statistical analysis, and thus to help project managers alleviating the possibility of software project failure. Factor analysis with data collected from 264 Korean project managers and consultants identified 12 categories and 46 risk factors. T-test results showed that project managers and participants had statistically different perception on 3 risk factors among those 46 risk factors. We concluded by discussing implications of our findings and future research directions.