Hideto Ogasawara

Software quality analysis and measurement service activity in the company

It is very important to improve software quality using program analysis and measurement tools and the SQA (Software Quality Assurance) method at the appropriate points during the process of development. In many development departments, there is often not enough time to evaluate and use the tools and the SQA method or to accumulate the know-how for effective use. This paper describes the support activity of a software quality analysis and measurement service which is performed by our laboratory team within the company as a third-party independent staff group. We call this activity HQC (High Quality software creation support virtual Center). The purpose of this activity is as follows. First we improve the software quality engineering process in the development department, for example, we help them to increase efficiency of review or testing. To accomplish this, we use program static analysis tools to detect fault-prone software components. Next we assist in starting their own self-improvement process. In addition, we provide service activity to many development departments concurrently. We have been making progress with these activities, and some development departments have begun to establish improvement processes themselves.

Experiences with program static analysis

Conventionally, software quality has been measured mainly by the number of test items, the test coverage, and the number of faults in the test phase. This approach of relying heavily on testing is not satisfactory from a quality assurance viewpoint. Since software is becoming larger and more complex, quality must be assured from the early phases, such as requirements analysis, design and coding. Code reviews are effective to build in software quality from the coding phase. However, for a large-scale software development, there are limitations in covering all the programs. The advantage of using static analysis tools is the capability to detect fault-prone programs easily and automatically. We describe the effective use of a static analysis tool, and show the effectiveness of the static analysis technique.

Proposal and practice of software process improvement framework - Toshiba's software process improvement history since 2000

For the effective promotion of software process improvement (SPI) activities in a large organization, it is necessary to establish an organizational structure and a deployment method for promotion and to develop training courses, support tools, and other materials. Even if an organizational promotion system is established, the SPI activities of each development department cannot be promoted effectively without an SPI community. To promote SPI activities throughout Toshiba Group, we organized a Corporate Software Engineering Process Group in April 2000. We also proposed an SPI framework. This framework extracts the features of organizations in which improvement activities were successful in the 1990s and combines them systematically. For over 10 years, we have been constructing Toshibas SPI framework based on the SPI framework that we proposed and promoting process improvement activities using it. As a result, maturity levels were improved in many development departments, and the applicability of the SPI framework has been confirmed. Copyright

Quality Control Techniques for Constructing Attractive Corporate Websites: Usability in Relation to the Popularity Ranking of Websites

Content and usability are important in the context of efforts to raise the attractiveness of a corporate website. Regarding the content, there must be information indispensable for a corporate website. Regarding the usability, it is necessary to establish a usability test method. This paper describes the results of an investigation of the Japanese-language corporate websites of Japanese companies in order to specify the content necessary for a corporate website. Furthermore, a scenario-based usability test technique was applied and the relationship between the popularity of a website and its usability was examined. Based on the results obtained, we propose a framework for constructing a corporate website by tailoring the WBS (Work Breakdown Structure) and checklist for static quality assurance, and usability and function test for dynamic quality assurance.

Application of the 0-1 Programming Model for Cost-Effective Regression Test

This paper reports an application of the 0-1 programming model to the regression testing plan for an industrial software. The key idea is to formulate a testing plan as a 0-1 programming problem (Knapsack problem). The empirical study shows that the 0-1 programming method can produce a cost-effective testing plan in which all potential regressions are found at only 22% of the cost of running all test cases.

Status of SPI activities in japanese software – a view from JASPIC

For the effective promotion of software process improvement (SPI) activities in the word, it is very important to establish a community beyond various social/organizational barriers. Like the auto-mobiles industry, to share various knowledge/experiences is to evolve one’s own industry. To promote SPI activities in Japan, we established Japan SPI Consortium (JASPIC) in October 2000. In this paper, we describe the status of SPI practice in Japan through our experience in various activities in JASPIC, analyze current issues from software engineering point of view, and make some proposals for future action.

Application of Mahalanobis-Taguchi Method and 0-1 Programming Method to Cost-Effective Regression Testing

To enhance the cost effectiveness of regression testing, this paper proposes a method for prioritizing test cases. In general, a test case can be evaluated from various different points of view, therefore whether it is worth it to re-run should be discussed using multi criteria. This paper shows that the Mahalanobis-Taguchi (MT) method is a useful way to successfully integrate different evaluations of a test case. Moreover, this paper proposes to use the 0-1 programming method together with the MT method in order to take into account not only the priority of a test case but also its cost to run. The empirical study with 300 test cases for an industrial software system shows that the combination of the MT method and the 0-1 programming method is more cost-effective than other conventional methods.

How to Effectively Promote the Software Process Improvement Activities in a Large-Scale Organization

In effective promotion of SPI (Software Process Improvement) activities in a large-scale organization, it is necessary to establish an organizational structure and a deployment method for promotion, develop training courses and support tools, etc. In order to promote the SPI activities throughout Toshiba group, we organized an SPI promotion project in April 2000. This paper discusses the problems encountered in the promotion of SPI activities and presents the solution to those problems. Moreover, the actual results are explained. As a result, it was found that the solution we developed can be used to effectively promote SPI activities. Further, some development departments reached a higher maturity level more quickly than envisaged in the CMU/SEI maturity profile.

A Topic Model and Test History-Based Test Case Recommendation Method for Regression Testing

In the regression testing, an oversight of a regression is a serious problem to be avoided. A test engineer usually selects test cases to rerun for a regression testing. While the selection is a useful expert decision, there is also a risk of missing some important test cases. To support a more effective regression testing, this paper focuses on the following two kinds of data: 1) the similarity of test cases in terms of their topics, and 2) the test history. Then, the paper proposes a hybrid method for recommending test cases in two steps by using the above data. As the thirst step, it recommends test cases which are highly-similar to the manually-selected ones. Then, as the second step, the method recommends the remaining test cases in decreasing order of priority computed by using the test history. The usefulness of the proposed method is proved through an empirical study using a set of regression test data for an industrial product.

A Test Case Recommendation Method Based on Morphological Analysis, Clustering and the Mahalanobis-Taguchi Method

This paper focuses on the content of test cases, and categorizes test cases into clusters using the similarity between test cases, their degree of similarity is obtained through a morphological analysis. If there are two similar test cases, they would test the same or similar functionalities in similar but different conditions. Thus, when one of them is run for a regression testing, the remaining one should be run as well, in order to reduce a risk of overlooking regressions. Once a test engineer decides to run a set of test cases, the method proposed in this paper can recommend adding similar test cases to their candidate set. The proposed method also considers the priorities of recommended test cases by using the Mahalanobis-Taguchi method. This paper reports on an empirical study with an industrial software product. The results show that the proposed method is useful to prevent overlooking regressions.