Software Project Metrics and Quality Management
Abstract- The quality of software is more of vital concernment. Software quality characteristics reflect the essence of software. Quality management of software development process is an important approach to solve the problem of software quality. There have been a lot of standards of software quality management already, such as ISO9001 and CMMI, which can help organizations to establish the system of quality management. Different from traditional industrial products, the process of software development is at the same time as the process of producing. Thus traditional methods of quality control of hardware obviously could not be adapted to quality management of software reasonably. This thesis describes a software project metrics and quality management knowledge combined with practical experience of the software project quality management to do some research.
Keywords- - Software; Project Metrics; Quality
I.METHODOLOGY AND PRACTICE OF PROJECT METRICS
A. Project Metrics
The project metrics is the specific metrics for software development projects, the purpose is to control the project of measure the size of project, project cost, project schedule, customer satisfaction and auxiliary project management.
B. Scale Metrics
Software development project size is the basis of measurement estimating software project effort, cost budgeting and planning reasonable progress of the project. Size measurement is a major reason for software project failure.[1] A good size measurement model can solve this problem. Effective software size measurement is the core elements of successful projects: effective software size
measurement can plan appropriate project plan, reasonable project plan will help manage the project effectively. Size measurement point is: estimate by the development of on-site project members; flexible use of the actual development of job data; put an end to blindly cater to customer demand "delivery inverse extrapolation".
Software size measurement helps software development teams to accurately grasp the development time, it costs distribution and defect density. There are a variety of software size estimation methods, such as: Function Point Analysis (FPA), lines of code (LOC), Delphi technique,
COCOMO model, feature point, standard component, object point, 3-D function points, De Marco's Bang metric, fuzzy logic and more specific methods.
Constructive Cost Model (COCOMO) is an accurate and easy-to-use model-based cost estimation method, it was first proposed by Boehm in 1981. The model is divided into three: Basic model, intermediate model and detailed model accord to their level of detail. The basic model is a static single-variable model with an estimated source line of code (LOC) as the independent variable function to calculate the software development effort. Intermediate model with LOC function as independent variables on the basis of the workload of the software development, then property products, hardware, personnel, projects and other aspects influencing factors to adjust the estimates of the workload.
Detailed model to include all the features of the intermediate model, but it with the various factors affecting the adjustment effort estimation also considers the impact of each step of the software engineering process analysis and design. COCOMO model has the features of estimates accurate and easy-to-use. In this model, the basic amount of the following:
1) The number of DSI (source instruction) is defined as the number of lines of code and include comment lines code. If a line has two statements, it will count as an instruction.
2) MM (unit of measure for men) said the development effort.
3) TDEV (the unit of measure for the month) means the progress of the development, it is determined by the workload.
4) 15 kinds of COCOMO model key consideration are affecting the software workload factors and by defining multiplication factor to accurately estimate the workload of the software.
C. Process Measurement
1) The Performance of Project Process
Process metrics measure all aspects of the software development process, the aim of the future performance of the forecasting process reduce the deviation of the results of the process, the behavior of the software process is management by objectives, process control and process evaluation continuous improvement to provide a quantitative basis. Process metrics and software development process is related to the strategic significance closely. Software process quality will directly affect that the quality of software products are good or bad, measurement and evaluation process can improve the process maturity and product quality.
Instead of measure and assess the quality of software products provide the necessary feedback and evidence to improve the quality of software process. Process metrics are related to the maturity of the software process closely.
2) The Process Metrics of Project Process Management
William A.Florac, Robert E.Park and Anita D.Carleton described the relationship between process management and project management in the Practical Software Measurement: Measuring for Process Management and Improvement. Production of products is based on three elements of software project team: product requirements, project plans and defined software process. Metrics in project management will be used to:
(1) Identify and describe needs.
(2) Prepare plans to achieve its objectives.
(3) Implementation plan.
(4) Track the implementation status and progress of work based on the goals of the project plan.
Process management can also use the same data and associated metrics to control and improve the software process by itself. This means that software organizations can use to construct and maintain a measure of a common framework of activities to provide data the two management capabilities of process management and project management. Course of the project management is an important step in the course of the project management, it include defined process, planning measure, the implementation of software process, application metrics, process control and improve the process, the planned metrics and metric, but also measure the core content of the course of the project. The planned measure to establish the understanding have been identified of the defined project process, products, processes, resources related matters and attributes, the collection and use of metrics are integrated into the project for process performance tracking requirements. Application metrics will be executed through the process measurement data obtained in the course of the project, as well as product-related data by product measures used to control and improve the course of the project.
3) Project Process Metrics
The project process metrics mainly includes three aspects, Maturity metrics are including organizational metrics, resource metrics, training metrics, document the standardized measure, data management, analysis metrics and process quality metrics; management metrics (such as milestone management metrics, risk measurement, work flow, measure, control measure and manage database metrics) are including project management, quality management measures (such as the measure of quality review, quality testing metrics and quality assurance measure. etc.) and configuration management measures (such as a design change control measure, version management control measure); Third, life cycle metrics are the main measure of problem definition, including requirements analysis metrics, design metrics, manufacturing, metrics and maintenance measure .
4) The Process of Project Process Measurement
A measure of the course of the project is implemented in accordance with the already well-defined process of measurement, which will enable the course of the project to measure job controllability and traceability, thereby increasing the effectiveness of the measure. The general process of software process measurement includes: the confirmation process issues; collect process data; analysis of process data; the interpretation process data; reporting process; process proposals; implementation actions; implementation of the supervision and control. The process quality of the measurement process ensures that the project process measurement data and questions about software process improve the course of the project implementation.
(1) The project needs are the base of measure software quality.
(2) Project quality can not be guaranteed if only to meet clearly defined needs, and did not meet the implied demand.
(3) Does not follow the standard definitions for the development of rules, project quality can not be guaranteed.