Kunio Noguchi

Priority assessment of software process requirements from multiple perspectives

Abstract The improvement of software processes involves the collection of software process requirements from various groups of stakeholders – each having their own perception of the process. Often stakeholders specify similar requirements in different ways. This results in the fact that different perceptions are related to each other. An effective technique is needed to incorporate the relationships between various stakeholders’ process requirements. This paper presents a Correlation-Based Priority Assessment framework (CBPA), which prioritizes software process requirements gathered from multiple stakeholders by incorporating inter-perspective relationships of requirements. This research uses a relationship matrix to analyze the impacts between requirements and to facilitate the integration. Higher stakeholder satisfaction is achieved by increasing the priorities of the software process requirements whose satisfaction can increase the satisfaction of other requirements from multiple perspectives.

A quantitative approach for setting technical targets based on impact analysis in software quality function deployment (SQFD)

Target setting in software quality function deployment (SQFD) is very important since it is directly related to development of high quality products with high customer satisfaction. However target setting is usually done subjectively in practice, which is not scientific. Two quantitative approaches for setting target values: benchmarking and primitive linear regression have been developed and applied in the past to overcome this problem (Akao and Yoji, 1990). But these approaches cannot be used to assess the impact of unachieved targets on satisfaction of customers for customer requirements. In addition, both of them are based on linear regression and not very practical in many applications.In this paper, we present an innovative quantitative method of setting technical targets in SQFD to enable analysis of impact of unachieved target values on customer satisfaction. It is based on assessment of impact of technical attributes on satisfaction of customer requirements. In addition both linear and non linear regression techniques are utilized in our method, which certainly improves the existing quantitative methods which are based on only linear regression.

Priority assessment of software requirements from multiple perspectives

The development of complex software systems involves collecting software requirements from various stakeholders. Often stakeholder perceptions conflict during the requirements elicitation phase. An effective technique to resolve such a conflict is needed. We presented a framework that prioritizes software requirements gathered from multiple stakeholders by incorporating inter-perspective relationships, which is not addressed by existing priority assessment techniques. We use a relationship matrix to analyze the impact between requirements and facilitate the integration process which assesses their priorities based on their relationships from multiple perspectives. It allows the development team to resolve conflicts effectively and concentrate their valuable time and resources on the critical few requirements from multiple perspectives that directly contribute to high customer satisfaction.

Requirement Acquisition, Analysis, and Synthesis in Quality Function Deployment

One of the most important problems in product design using quality function deployment is how to capture true customer re quirements. In this research, we develop systematic methods, processes, and techniques for eliciting complete, consistent, non- redundant, and true customer requirements from multiple perspectives. It covers from requirements acquisition, requirements extraction to requirements synthesis both within and across perspectives. In our framework, acquisition of requirements starts with identification of key quality concerns from individual perspectives. It is followed by elicitation, analysis, and extraction of both functional and non- functional requirements using various techniques, such as structured interviews, functional decomposition, quality attribute analysis, and use case analysis. The derived requirements are synthesized into required quality requirements using generic rules. In the synthesis, we must make sure to detect and resolve mutual exclusive requirements, remove ambiguity in requirements, and delete duplicate ones.

QFD application in software process management and improvement based on CMM

Capability Maturity Model (CMM) from Software Engineering Institute has been used successfully by many organizations for software process improvement. However, there exists a disconnection between business goals and maturity levels. A new framework using Quality Function Deployment (QFD) is developed to deal with this problem. This framework serves three purposes: (1) it provides a connection between business requirements and CMM; (2) it proposed a methodology for the priority assessment of requirements from multiple perspectives; and (3) it helps identify a set of software process improvement actions based on business requirements and CMM.

Business-oriented software process improvement based on CMM using QFD

The Capability Maturity Model (CMM) from the Software Engineering Institute has been used successfully by many organizations for Software Process Improvement (SPI). Despite its great success, there exists a disconnect between business goals and maturity levels. A new framework using Quality Function Deployment (QFD) has been developed to deal with this problem. This framework serves three purposes: (i) it establishes a connection between requirements from the business and Key Process Areas (KPAs) in CMM; (ii) it proposes a methodology for the priority assessment of requirements from multiple perspectives; and (iii) it helps identify a set of software process improvement actions based on business requirements and KPAs. Copyright

Technical Target Setting in Time-Stamped Quality Function Deployment

Abstract Quality Function Deployment (QFD), one of the major quality systems, has been applied to develop numerous products (Akao, 1990), including software systems, to improve their quality. Target setting is a complicated and important task in product development using Quality Function Deployment (QFD). The targets are extremely important since they directly affect customer satisfaction. We need to develop methods of setting target values of engineering attributes based on their technical trend to achieve the quality goals in terms of satisfaction of customer requirements for a given timeframe so that by the time of completion of a project the values of technical attributes are competitive. Two quantitative approaches for setting target values: benchmarking and primitive linear regression were developed and used in the past, both of which are based on linear regression and not very practical in many applications. In this paper, we present an innovative quantitative approach to set technical targets in QFD to enable assessment of the impact of unachieved target values on customer satisfaction. In addition, both linear and non-linear regression are utilized in our method; hence, we need not be confined to straight-line relationships when an exponential function, a power function, or a polynomial rather than a straight line may better represent the data obtained. We further develop a time-stamped target setting method by incorporating the technical trend and time. By incorporating the market trend and the time of delivery of the product into the target setting process, we can set targets that provide a competitive edge for our product over the competitors products and a high level of customer satisfaction.

Target Setting for Technical Requirements in Time-Stamped Software Quality Function

Technical requirements are hard to determine in software development. They are often specified subjectively in practice. Poorly determined technical requirements often lead to poor customer satisfaction, cost overrun, and delay in schedule, and poor quality. Quality Function Deployment (QFD) is one of major engineering methods used to elicit customers needs and transforms them into technical requirements in industry. It has been applied to develop numerous products, including software systems, to improve their quality. However, target setting for technical software requirements is a complicated and challenging task in product development in Software Quality Function Deployment (SQFD). Current methods for target setting for technical software requirements do not consider their technical trends for a given timeframe. As a result, by the time of completion of a project the target values of technical requirements may not be competitive any more. In this paper, we first discuss benchmarking, primitive linear regression and target setting based on impact analysis to set targets for technical requirements in Software Quality Function Deployment (SQFD). We then develop a method of target setting for technical requirements by incorporating timeframe and the technical trend. It can help us to assess impact of both under-achieved and over-achieved targets. By incorporating the technical trend and the time of delivery of the product into target setting process, we can set targets for technical requirements that provide a competitive edge for our product over the competitors products and a high level of customer satisfaction.