• Journal of Information Technology Applications and Management
• /
• 제19권2호
• /
• pp.181-196
• /
• 2012

The cost of software maintenance occupies about two thirds in the software lifecycle. However, it is not easy to estimate the cost of software maintenance because of various viewpoints about software maintenance, unclear estimation methods, and complex procedures. Until now, the cost estimation model has used compensation factors for software characteristic and environment on the basis of program size. Especially, most of existing models use maintenance rate of total software cost as a main variable. This paper suggests the software maintenance cost estimation model that uses the result of calculating real maintenance efforts. In this paper, we classify functional maintenance and non-functional maintenance as software maintenance activity type. For functional maintenance, present function point of target software is needed to evaluate. The suggested maintenance cost evaluation model is applied to a software case in public sector. This paper discusses some differences between our model and other modes.

• 한국IT서비스학회지
• /
• 제9권4호
• /
• pp.169-185
• /
• 2010

According to software growth, also software maintenance has been continuously improving. In addition, the existing concept of software maintenance process demands operational management and improvement of service task. However, when we perform maintenance task, we have several constraints about applying service based requirement to system. Therefore, in order to solve these problems, we need a study of task of process for service based maintenance. In this paper, we propose a Service based Software Maintenance Process. Proposed process based on MaRMI-RE standard for software development and maintenance and compares it with the service based representative standards. In a related works, we study activity of ITIL and identify activities and tasks for maintenance. After this, identified activities and tasks compare with activities of MaRMI-RE. And then, we derive activities and tasks of a Service based Software Maintenance Process. Finally, we validate a result by comparing the proposed process with a general service operational process.

• 디지털콘텐츠학회 논문지
• /
• 제11권4호
• /
• pp.537-543
• /
• 2010

소프트웨어의 발전에 따라 유지보수 프로세스도 지속적으로 개선되어 왔다. 또한 기존 하자유지보수 중심의 개념은 운영관리 및 서비스의 향상을 요구하고 있다. 그러나 서비스 기반의 요구사항들을 유지보수 수행에 반영하는 데에는 여러 가지 제약사항을 가지고 있다. 이런 문제들을 해결하기 위하여 서비스에 기반하여 유지보수를 할 수 있는 프로세스의 연구가 필요하다. 본 연구에서는 서비스 기반의 소프트웨어 유지보수 프로세스를 제안한다. 제안하는 프로세스는 소프트웨어 개발과 유지보수 관련 표준인 ISO12207을 기반으로 서비스 기반의 대표적 표준들과 비교한다. 관련연구로 $SM^{MM}$, ITSCMM, ITIL의 활동을 연구하여 유지보수와 관련된 활동들을 찾아내고 ISO 12207의 활동들과 비교하여 서비스 기반의 유지보수 프로세스 활동을 식별한다. 이를 통해 서비스 관리 단계를 제안한다. 서비스 관리 단계는 4개의 활동과 8개의 산출물로 정의하였다. 마지막으로 제안한 프로세스를 일반적인 서비스 수행과정과 비교하여 결과를 검증하였다.

• 산업공학
• /
• 제17권spc호
• /
• pp.122-130
• /
• 2004

Since the cost of software maintenance occupies about 50~75% in a general successful organization, the software maintenance plays an important role in software life cycle. In particular, if the managed system needs to be operated in a long term or the system is very large and complex, then the maintenance is especially more important. Software maintenance is defined as software modification activities after customer delivery, such as improvement of performance or functionality, error correction, adaptation to environmental changes, etc. In this paper, software cost estimation models are proposed, that is based on productivity of manpower in maintenance projects. In order to do this, the activities of maintenance are classified into function change, non-function change, user support and application operation. The proposed models are constructed and verified based on the real size and cost information of projects in the real world. The approach in this paper is to discriminate the heterogeneous activities in maintenance projects, and then to calculate the respective cost of each discriminated activity. By using the proposed models, the total cost of maintenance project is summed from the costs of four activities. In addition the number of conflicts between owner and order receiver about the amount of cost will be reduced and the reasonable cost estimation system will be established.

• 응용통계연구
• /
• 제23권3호
• /
• pp.573-584
• /
• 2010

This paper considers the differences in the software execution environments in the testing phase and the operational phase to determine the optimal release time and warranty period of software systems. We formulate equations for the total expected software cost until the end of the software life cycle based on the NHPP. In addition, we derive the optimal release time that minimizes the total expected software cost for an imperfect debugging software reliability model. Finally, we analyze the sensitivity of the optimal testing and maintenance design related to variation of the cost model parameters based on the fault data observed in the actual testing process, and discuss the quantitative properties of the proposed model.

• 한국경영과학회지
• /
• 제18권3호
• /
• pp.129-149
• /
• 1993

The high level of activity in the development and maintenance of computer software makes the scheduling of software projects an importnat factor in reducing operating costs and increasing competitiveness. Software activity is labor intensive. Scheduling management of hours of software work is complicated by ther interdependencies between the segments of work, and the uncertainties of the work itself. This paper discusses issues of scheduling in software engineering management, and presents a modular decomposition model for software project scheduling, taking advantage of the facility for decomposition of a software project into relatively independent work segment modules. Modular decomposition makes it possible to treat scheduling as clustering and sequencing in the context of integer programming. A heuristic algorithm for the model is presented with some computational experiments.

• 정보처리학회논문지D
• /
• 제13D권2호
• /
• pp.241-250
• /
• 2006

유지보수는 소프트웨어 생명 주기에서 대부분의 비용이 소요되는 중요한 활동이다. 기존에 소프트웨어 유지보수 활동에 소요되는 비용의 예측을 위하여 다양한 소프트웨어 메트릭이 제안되었다. 본 논문에서는 객체지향 소프트웨어의 유지보수성과 기존에 제안된 메트릭과의 관계를 파악하기 위해서 수행된 실험의 결과를 소개한다. 실험에서는 메트릭으로서 LCOM, RFC, DAC, LOC를 사용하였으며 클라이언트/서버 및 웹 기반 시스템을 대상으로 유지보수 활동 시 실제로 소요되는 시간을 측정하였다. 본 실험을 통해서 메트릭과 유지보수 비용과의 관계에 대한 기존의 일반적인 주장을 실제 시스템의 유지보수 활동을 통해서 확인한 결과 기존의 주장과 달리 일부 시스템의 경우에는 기존의 주장을 뒷받침하지 못할 수 있음을 확인하였다. 구체적으로 말하면 소프트웨어 아키텍처, 설계 패턴 등이 적용되는 개발된 최근의 소프트웨어에서는 메트릭과 유지보수 활동과의 관계에 대한 추가적인 많은 연구가 필요함을 확인하였다.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2008년도 춘계학술대회 논문집
• /
• pp.659-664
• /
• 2008

The Railroad Safety Act's regulation reads as follows. "The Minister of Construction and Transportation may qualify and authorize the product to guarantee performance and safety of parts, machine, and device used in Railway fields." Another regulation reads as follows."“The guidelines about targets, standards, and procedures of Quality and Authority in first provision are decided as Ministry of Construction and Transportation Decree." The software used in rail cars and facilities is considered as a railway product. Therefore, it is qualified and authorized for acquiring the safety of rail cars and facilities. The software businesses shall again a Quality and Authority for applying a software to the rail cars and facilities. This paper regulates some guidelines that are needed to develop a software. The procedures that a software developer performs are divided by plan, requirement, design, implementation, and maintenance. The procedures that a software verification person performs are classified by verification plan, requirement verification, design verification, implementation verification, testing verification, maintenance verification, and safety activity. The entire processes and detailed activities to develope and verify a software are described as new guidelines.

• 한국IT서비스학회지
• /
• 제16권2호
• /
• pp.45-60
• /
• 2017

The study suggests a contract model of application operation through case study of A bank's IT outsourcing application contract based on workload. The IT outsourcing order form has a problem in that the scope of work is ambiguous due to the integration of operation and maintenance. In this study, application operation and maintenance were separated by referring to application operation history provided in ISO/IEC15504-5 standard. The scope of the IT outsourcing service was clarified by organizing the definition and detail activities of the application operation business. Application operation contract method has generally applied estimation method by the number of input manpower and period by agreement between buyer and client. As there is no activity to calculate the number of input manpower based on the operational work history and based on the standard workload per activity. In this case is not guaranteed due to the simple agreement between the contractors. In this paper, we propose an application operating cost estimation model that measures the size of the operating software using function point analysis that is the basis of application operation tasks. In order to verify the validity of the application operation cost model, we verified the correlation between the application size and the labor cost through regression analysis using SPSS.

• 한국컴퓨터정보학회논문지
• /
• 제21권5호
• /
• pp.119-126
• /
• 2016

Software maintenance becomes extremely difficult, especially caused by multiple versions in project-based or customer-oriented software development methodology. For reducing the maintenance cost, reengineering to software product line can be a solution to the software which either is a family of products nevertheless little different functionalities or are customized for each different customer's requirement. At an initial stage of the reengineering, the most important activity in software product line is feature extraction with respect to commonality and variability from the existing system due to verifying functional coverage. Several researchers have studied to extract features. They considered only a single version in a single product. However, this is an obstacle to classify the commonality and variability of features. Therefore, we propose a method for systematically extracting features from source code and its change history considering several versions of the existing system. It enables us to represent functionalities reflecting developer's intention, and to clarify the rationale of variation.