Earticle

오픈소스 소프트웨어는 대가 없이 활동하는 자발적인 개발자들에 의해 개발되며 그 소스 또한 대가 없이 공개된다. 오픈소스 소프트웨어 프로젝트는 계약이나 의무관계가 없는 비공식적인 네트워크에 의존하고 있으므로 위태로운 조직처럼 보이기도 한다. 하지만 텍이나 아파치 그리고 리눅스 프로젝트에서 보듯이 매우 성공적인 프로젝트가 존재하며 이들이 개발하는 소프트웨어의 품질은 전유 소프트웨어가 따라오지 못한다. 성공적인 오픈소스 프로젝트의 공통점은 명료한 거버넌스 모델을 가지고 있다는 것이다. 본 연구는 오픈소스 소프트웨어 프로젝트의 거버넌스를 다룬다. 오픈소스 거버넌스의 이론적 근원을 Coase(1937)의 거래비용에 두고 이로부터 조직의 거버넌스 관점에서 최근의 오픈소스 거버넌스 이론에까지 이어지는 이론적 고리를 연결한다. 또한 오픈소스 거버넌스의 모델의 실제를 검토하고 이로부터 시사점을 찾아 적용하는 문제를 다룬다.

This paper examines theoretical backgrounds of the governance models of open source projects. We found as the theoretical root of governance models the transaction cost theory that could be traced back to Coase(1937). We have tried to build a theoretical stream flowing from the transaction cost theory to the theory of governance models of open source projects, via network governance theory and virtual organization theory. This paper also reviews two types of governance models of open source software projects in practice: the meritocratic governance models and the benevolent dictator governance models, and characterizes their features. Based on the characterization of the governance models, we made a suggestion to the Korean TeX Users Group to build a governance model in accordance with the cathedral-type and meritocratic governance model.

지난 몇 년간 공공 정보화 사업의 기능점수 고정단가 방식의 한계점과 문제점을 조사하고자 해 외 주요 선진 국가들이 적용하고 있는 기능점수 기반의 생산성 적용 사업대가 산정 방식을 검토하고, 다양한 소프트웨어 사업의 특성별 개발 생산성에 영향을 미치는 요인들을 실증적으로 분석하였다. 기 존 국내외 관련 문헌연구와 해외 주요 선진 국가들의 사업대가 산정 가이드를 참조하여 개발 생산성 에 영향을 미치는 7개의 독립변수(개발유형, 기능점수 예측규모, 인력 참여형태, 개발언어 유형 및 개 수, 개발 방법론 유형, 개발 생명주기 모형, 개발 플랫폼)와 종속변수인 개발 생산성(1FP당 투입 M/ M, 총 기능점수/총 투입인력)간 상관관계에 관한 가설수립과 통계 분석을 수행한 결과 기능점수 예측 규모, 개발언어 유형, 개발 방법론, 개발 플랫폼의 유형에 따라 개발 생산성 차이가 있는 것으로 나타 났다. 매년 기능점수 단가 상승 요구 및 단가 신뢰성 이슈에 부담을 줄이고 대가 산정의 유연성 확보 를 위해서는 본 연구 결과를 활용한 생산성 기반의 대가산정 체계로의 적극적 검토가 필요할 것으로 생각된다. 또한 앞으로 보다 다양한 분석을 통해 국내 공공 소프트웨어 개발 생산성을 진단하고 및 제 고 방안을 마련하고, 개발 생산성을 고려한 적정 사업기간과 투입인력을 산정하는 등 국내 소프트웨 어 개발 경쟁력 향상에 기여될 수 있기를 기대한다.

This study aims to improve the shortfalls and limitations of the function point per fixed pricing methodology of public software projects. The calculation methodologies used by developed countries for function point development productivity of software project pricing has been evaluated. In addition, the purpose of this research pertains to the analysis of empirical factors that affect the development productivity of various aspects of the software industry. Between the 7 independent variables that affect development productivity(type of development, function point scale, labor participation, type and number of development language, type of development methodology, development life-cycle model, development platform) and dependent variables (M/M per 1FP, total function point/total # of employees), it was discovered that development productivity was affected by the correlation hypotheses and the results of the statistical analyses of the function point scale, type of development language, development function and development platform. Existing domestic literature and software project pricing guides utilized by developed countries were used as references. In order to alleviate the burden of responsibility related to the increase in unit price for function points and securing the flexibility of cost calculations, serious consideration should be given to the pricing methodologies outlined in this research. Going forward, the development productivity for domestic software projects should be assessed via rigorous analyses, resulting in a plan for process improvement. It is hoped this research can contribute to the efficacy of the calculating the adequate project period and personnel with due consideration to the development productivity and an improved competitiveness for domestic software developers.

Function point elements are metrics to measure the size of software projects. This article investigates the relationship among function point elements using principle component analysis. Principle component analysis reveals the relationship of different function point elements such that they may be measuring the same attribute of a software project. Therefore, principle component analysis brings out the influence of a function point element over each other. Principle component analysis can help to integrate the function point elements.

The concern of this study is to identify software risks and controls in the software development lifecycle. The aim of this study is to rank the software risks factors according to their importance and occurrence frequency based on the data source. The survey questionnaire is used to collect data and method of sample selection referred to as ‘snowball’ and distribution personal regular sampling was used. The seventy six software project managers have participated in this study who works in the Palestinian software development. Fifty software risk factors in all phases SDLC and thirty risk management techniques were presented to respondents. The results show that all risks in software projects were significant and important in software project manager's perspective. However, the ranking of the importance of the risks is assigned according to it: Analysis, planning, maintenance, design, and implementation. In addition, the top ten software risk factors in software development are selected and used for further analysis such as: Risk13, Risk 14, Risk15, Risk16, Risk11, Risk18, Risk12, Risk50, Risk19, and Risk 9. The concern of this paper the top ten controls are used to model its relationship with the risk, such as: C29, C30, C20, C27, C21, C19, C28, C25, C26, and C23. Software risks can be modelled empirically with risk management control techniques. We recommended applying more studies in software risk management practices with real world companies and building tools to identification and analysis software risks based on quantitative and intelligent techniques.

The risk of software projects is measured in terms of cost that is needed to abate the risk. Traditional practice to measure the risk of software projects uses risk exposure; however, risk measure cannot quantify the risk beyond the expected value of cost. Software project managers are keen to quantify the risk based on a certain probability which is beyond the expected value of the cost. This research work presents a model to measure the risk based on certain probability beyond the expectation. A case-study validates that proposed model shows an improvement in the measurement of risk of real software projects compared to the actual risk of software projects.

Software development process is vulnerable to different risk factors. Several works were done in order to identify these factors and to introduce new ones. On the other hand, little research has been conducted around investigating the relationship between these risk factors. In this paper, a detailed set of software risk factors were examined. It was found that the relationship between these factors is a (cause- effect) relationship. As a result, a causality table and a cause-effect diagram are introduced to illustrate this causality relationship.

This paper presents a risk classification scheme for software projects that utilizes the dependence of risk events to occur on other risk events and the statistical dependence of impacts of risk events. The proposed risk classification scheme helps to avoid the under and over estimations of the overall impact of dependent risk events. Further, the proposed risk classification scheme protects against under and over counting of the impacts of statistically dependent risk events of software projects. The proposed risk classification scheme utilizes the Software Engineering Institute’s risk taxonomy and shows that different classes of the risk taxonomy are independent having statistically independent impacts while different attributes of a taxonomy class are independent having statistically dependent risk impacts.

The development process of the software system is influenced by multi factors, thus the risk control of software projects is of vital importance. Targeting at the features including multi-factor, multilayer, fuzziness and uncertainty during the process of risk assessment of software projects, this paper analyzed the risk measurement of software projects and proposed a multilevel risk measurement algorithm of software projects based on grey variable weight clustering model. This algorithm firstly analyzed the influencing factors of the risk control of software projects, establishing a multilevel risk measurement system for software projects; next, establishing the whitenization weight functions of different measurement indexes and different measurement grades of software projects risks respectively on the basis of the grey system theory, and under the condition of considering the degree of importance of the indexes of different levels, the corresponding risk measurement algorithm of software projects based on grey variable weight clustering model was established; then the paper presented the detailed execution steps of this algorithm and discussed the detailed process to achieve the algorithm. Finally, the algorithm was verified through the case study.

Software project management has emerged as a new discipline with wide-ranging ideas and across-the-board insights for effectively managing key areas of software projects. The remarkable work of the software project managers, professionals and researchers across the globe have resulted in substantial improvements in this field. Likewise, software project failure rate has decreased considerably due to the use of effective software project management tools and techniques by the software houses. Now more efficient, robust and quantitative measures are being practiced in the areas of software requirement gathering, analysis, design, architecture, development, quality assurance, integration, deployment and support. A number of metrics are used by the requirement engineers, system analysts, software engineers, team leads, software project managers and other professionals to successfully manage, execute and complete the software projects. As the software industry moves towards a more mature state, the need for employing more effective tools, techniques and benchmarks for managing software projects has become indispensable to minimize the negative risk factors and improved adherence to quality assurance. Particularly, requirement metrics are useful in identifying risks of a project by locating errors in the requirements document. These metrics validate the gathered requirements against the actual requirements by evaluating whether the requirements are complete or not. A range of metrics are used for measuring the requirements e.g., volatility metrics check changes in the requirements, traceability evaluates links among the requirements within a document and requirements completeness metrics verifies whether the specified requirements are complete or not. Multiple metrics are recommended to be used to assess the health of a software project to ensure overall quality as a single metric cannot suffice. In this paper, we explore how different metrics relating to different areas of software project, especially in requirement engineering, can be useful to manage the software projects for knowledgeably. The focus of this research study is to evaluate and highlight the importance of various performance metrics and propose additional metrics for requirement gathering and management.

Accurate estimation of software development effort needs to consider numerous factors in software projects such as project size, programming language, application type, team size, development platform and so on. Software methodology is one of the most effective parameters in this field. Basically, software methodologies are used in software projects in order to manage activities, artifacts, roles and disciplines. Researchers believe that methodologies can make project managers enable to deal with uncertain and complex nature of software projects. In spite of having noticeably advantages, using the methodologies may lead to inaccurate development effort estimation because there is no analytical evidence to clarify the real effect of using methodology on development effort. Underestimating the amount of effort required for utilizing methodology can have negative influences throughout the project. In this paper, an analysis was performed to draw the real situation of software projects in term of using methodologies. Furthermore, the effect of using methodologies on development effort was investigated to show how methodologies can affect the software development effort. A real and large dataset including 5052 software projects from various areas was used in this study. The results showed high interest to employ methodologies in recent software projects as well as high and significant impact of methodologies on development effort.

The opportunity cost decision of selecting a software project is dependent, among other things, on how a given project best meets a company’s business goals and overall competitive strategy. Remaining competitive in the agile software market today requires selecting only those projects that position a business strategically in the market place and that render it competitive over time. Using an industrial case study, we demonstrate the role of the Complex Adaptive Systems Software Engineering (CASSE) framework in supporting value-based project selection. We apply Actor Object Dependencies (AOD) analysis and Functional Points (FP) sizing techniques to predict the overall project value before incurring any actual costs of implementing such a project. The overall contribution of this work therefore lies in demonstrating that alternative engineering approaches that analyze AODs can enhance how we select software projects and how we plan project schedules optimally so as to increase business value derived on projects.

A number of different factors are thought to influence the efficiency of the software development process. These include programming languages, use of formal methodologies, CASE tools, etc. Moreover, efficiency in the context of software development has traditionally been measured as the ratio of functionality, either lines of code or functions points, and the effort expended. This is a unidimensional measure and ignores factors such as quality and elapsed time. This study utilizes Data Envelopment Analysis (DEA) to develop a multidimensional measure of efficiency in the context of software development and then examines how efficiency varies with other influencing factors such as the use of CASE tools, use of methodologies, team size, application type, etc. using a decision tree based model.

Techniques and models for mitigating risk in software development projects classified into three categories–namely, qualitative, quantitative, and intelligent approaches. This paper is to review the quantitative and intelligent risk models in software risk management for software development projects. Indeed, this area needs more effort from scholars and researchers in quantitative and intelligent risk models to mitigate risks. As future work, we will use these hybrid models of quantitative and intelligent for mitigating software risks in cloud computing such as neural network, genetic algorithm and others artificial intelligence techniques.

As a practice-oriented discipline, Software Engineering is most effectively taught by using a variety of hands-on experiences. Team-based projects where students can practice their technical and soft skills are a key feature of many software engineering courses aimed to prepare students for the realities of industrial software development. This paper surveys the objectives of software development projects in undergraduate courses and summarizes four roles in which the instructor can participate in student projects: a customer, an architect, a team manager, and a mentor. The paper is concluded with case studies illustrating the nature of these roles.

An exertion capitalized especially in an IoT software project, is probably one of the most imperative and analysed phenomena in current stage in real time system software management. The purpose of the value of this variable when commencing IoT software projects permits to map adequately any immediate actions. To the extent that estimation and prediction are apprehensive, there are tranquil some unsettled problems and faults. To obtain high-quality results, it is vital to take into contemplation any previous IoT software projects. Estimating the exertion with a high rating of reliability is a dilemma which has not yet been solved and yet the project executive has to deal with it ever since the foundation. In this work, the concept of constructive cost model (COCOMO) 2.0 experiment for modelling of IoT software projects and performance of developed models are compared with the RMSE (root mean squared error) along with MAE (mean absolute error) for a deep analysis of COCOMO 2.0. The techniques are run on Analogy-based estimation for building the model structure for IoT software and formulae are premeditated over mat-lab software. The valuation benchmarks are based on RMSE and MAE. The outcome shows that the COCOMO 2.0 has the best feat and could be used for all types of IoT software projects for other IoT manufacturers in the market.

프로젝트 관리 성숙도(Project Management Maturity Model)란 프로젝트 관리 관련 업무 성숙도를 측정하는 것으로, 프로젝트 성공을 위해 관련 있는 업무가 어느 정도 수준으로 진행되고 있는가를 평가하는 목적으로 주로 사용된다. 본 논문에서는 프로젝트 관리 성숙도를 활용하여 프로젝트 관리 성숙도가 범위, 일정, 그리고 원가에 미치는 영향을 연구 하였다. 연구 결과로 프로젝트 관리 성숙도가 낮으면 범위 변경률이 커지고, 일정이 지연되며, 비용 또한 초과되는 것으로 나타났다. 본 연구의 성과를 토대로 기업이 프로젝트 관리 성숙도 향상을 위해 투자와 노력을 해야 한다는 것을 알 수 있다.

소프트웨어 개발 프로젝트 초기단계에 수립한 일정 및 비용 등의 계획은 수행 중 발생하는 많은 불확실성을 반영하고 있지 않기 때문에 진행되고 있는 프로젝트는 초기에 계획한 일정과 비용에 많은 차이를 보이는 경우가 발생한다. 이에 따라 본 연구에서는 진행 중인 소프트웨어 개발 프로젝트의 상황을 분석하고 예측할 수 있는 방법으로 일정과 비용을 통합 관리할 수 있는 획득가치관리(EVM: Earned Value Management)를 활용하는 방안을 제시하였다. 획득가치관리(EVM)을 적용하기 위해 일정과 비용의 통합관리하기 위해 예상되는 투입 인력과 비용 등을 감안하여 작업분할 구조(WBS)와 연계하여 계획을 수립하고 측정하는 방법을 구축 중인 프로젝트의 사례를 통하여 적용 방안을 제시하였다. 사례 분석한 각각의 프로젝트 진행 단계에 따라 예측한 값이 실제 계획 대비 일정과 비용에 대한 예측 오차 변화 추이를 분석하였으며, 향후 진행되는 프로젝트의 일정 및 비용 등의 계획을 체계적으로 수립하고, 진행되는 프로젝트의 일정 및 비용 등에 대해 분석하고 예측하는데 활용할 수 있도록 연구하였다.

Rapid advancements in the field of information communication technology (ICT) have enabled people to collaborate through the Internet. This study focuses on open-source software development, which develops complex software or web services autonomously. To verify the antecedents of the participation of developers for successful open-source software development, we examine the relationship of the leadership roles of the project leader, such as knowledge contribution, interactions, and quality control with the participation of individual developers. For data collection, we develop a web crawler using the Python programming language and collect 518 repositories from GitHub, which is one of the leading platforms for developing open source software. We find that the participation of developers is closely associated with leaders who actively contribute knowledge, interact with the developers, and ensure strict quality control. Furthermore, the software development phase moderates the relationship between quality control and the participation of individual developers.

소프트웨어 프로젝트 관리과정의 활동에서 프로젝트 관리자의 중요한 임무는 소프트웨어의 크기와 인적 노력 등을 예측하는 것이다. 최근 소프트웨어 개발에 주로 사용되는 제 4세대 언어(4GL)와 데이터베이스 환경에서 개발되는 프로젝트에 대한 크기를 예측하는 모델은 불행하게도 연구가 미비한 실정이다. 본 논문에서는 4GL로 개발되는 프로젝트 개발 초기 단계에서 수집한 메트릭스를 이용하여 소프트웨어 크기를 예측하는 추정 모델을 제안 한다. 제안된 방법은 상대오차(MRE)를 최소화시키는 방법으로 개발 초기과정에서 얻어지는 측도들의 이상치에 덜 민감한 특성을 가지고 있다. 본 논문에서 제안된 모델에 대하여 적합도와 예측력의 성능을 테스트하기 위하여 데이터 ？을 I,II 2개로 나누어 실험하였다. 실험결과, 추정된 모델의 적합도와 예측력은 제안된 MRE 추정방법이 데이터 셀 I,II 모두에서 전통적인 방법 LS, RLS보다 우수하게 나타났다.

An important task for any software project manager is to be able to predict and control project size. Unfortunately, there is comparatively little work that deals with the problem of building prediction methods for software size in fourth-generation languages and database projects. In this paper, we propose a new estimation method for estimating for software size based on minimum relative error(MRE) criterion. The characteristic of the proposed method is insensitive to the extreme values of the observed measures which can be obtained early in the development life cycle. In order to verify the performance of the proposed estimation method for software size in terms of both quality of fit and predictive quality, the experiments has been conducted for the dataset Ⅰ and Ⅱ, respectively. For the data set Ⅰ and Ⅱ, our proposed prediction method was shown to be superior to the traditional method LS and RLS in terms of both the quality of fit and predictive quality when applied to data obtained from actual software development projects.