Qualitative Quality Analysis of Process Factors Based on Software Design Review Record Affecting Software Reliability  
Author Takashi Sato

 
Co-Author(s) Shigeru Yamada

 
Abstract It is important to reliably conduct reviews and tests to assure the quality of software. In particular, reviewing design specifications raises the degree of completion of documents at each process step and leads to securing the quality of the final product. In addition, if review is done sufficiently, it would lead to productivity improvement by reducing backtrack work done to correct defects detected during development. For our quality improvement policy, the standard value of review efforts, which is one of our process metrics, has been implemented, helping the review process to be carried out reliably. As a result, the number of defects detected after shipment have decreased, attaining a consistent level of good quality. However, this does not mean that the more hours spent on review, the higher the quality becomes. If the process secures a certain number of review efforts and proceeds steadily, review efforts do not tend to increase more than that. There is a limit to the way that increasing the review efforts more assures the quality for an organization that has a mature development process. Therefore we considered what factors other than review efforts assure review quality. We also considered ways to measure the product metrics of a review, obtained from the information obtained from the text of a review record, that have an influence on design quality. In a study on past reviews, there were a few cases of quality engineering approaches done through human factor analysis [1] [2] and process evaluation and improvement approaches [3] [4], but there seemed to be few cases of methods using product metrics with review records. In this paper, we analyzed the factors that influence design quality from the metrics obtained from the information of a review record.

 
Keywords software quality analysis, design review, product metrics, review record, text mining
   
    Article #:  22117
 
Proceedings of the 22nd ISSAT International Conference on Reliability and Quality in Design
August 4-6, 2016 - Los Angeles, California, U.S.A.