Shinobu Saito
Software Engineering Project
Distinguished Research Engineer
(As of November 2018)
A mediator between the inside and outside of NTT
In addition to research activities in my own field, I also dispatch information on our R&D activities as a distinguished research engineer. This role is required to disseminate the importance and effectiveness of NTT’s R&D results inside and outside NTT. SIC, to which I belong, and the NTT Group are making efforts in R&D in the software development field, aiming to establish a novel methodology for providing high quality software systems and solutions with both a short lead-time and lower cost. I believe that this methodology could be one of advantages for our group in the global market. From my viewpoint, an ideal example of this is the TPS (Toyota Production System) developed by Toyota Motor Corporation. The TPS has become widely known as an excellent methodology for producing vehicles with high quality and lower costs. TPS has also been very effective in promoting the Toyota brand. Similarly, I think that NTT can demonstrate that, since we have such a novel methodology, we can develop and deliver excellent software systems. In short, I think we can do the same thing as Toyota.
System comprehension: technologies to understand how software is used
Our current research topic is to restore knowledge and know-how for developing software systems. The set of technologies (e.g., methods and tools) related to this topic is known as “system comprehension.” These technologies aim to recognize when software systems are used, who uses them, and how and where they are used. There are many cases where usage of software systems, in general, changes over time. These changes in usage are different from what the software engineers expected when they developed the systems. When smartphone applications, for example, are updated to a new version, it is likely that the look and feel of the screens change. This change sometimes makes users, who used the old version, give the applications low ratings. This unfavorable situation results from users’ lack of knowledge of the usages of these applications. To avoid such situations, software engineers need to investigate usage comprehensively. It can be said that similar situations occur in large-scale software system development projects in corporations.
When a user uses a software system, the system generally records data on what times it was used, how it was used, or what part of it was not used. System comprehension technologies analyze the data and then help software engineers understand not just the software systems themselves but the user’s behavior and situation. These technologies enable us to restore valuable information for developing software systems that users find easy to use and rate highly.
New value of IT experts in the AI era.
To determine users’ needs and requirements, methodologies such as Design Thinking and User Experience are attracting attention in the field of software development. On the other hand, to prevent the above-mentioned unfavorable situations, we need to not only collect requirements and needs, but also investigate how to use running software systems comprehensively. However, existing investigation methods have been very human-intensive and time-consuming. For example, one of these methods is “shadowing,” in which an investigator stands behind a user of a software system and observes how it is used. Depending on the target software and number of users, it may be necessary to conduct a survey that requires several months. Therefore, we focus on system comprehension technologies. We think that these technologies could contribute to understanding how to use software systems accurately and rapidly. Currently, in the field of software development, AI and robotics have been attracting attention. Those technologies have already been utilized for the automation of software system development and promotion of autonomy of software system operations. On the other hand, in order to comprehend software systems and the people (e.g., users and customers) related to them, many research issues still remain. I think that system comprehension capabilities could be one of the new values of IT experts in the future.
Profile
Shinobu Saito
Software Engineering Project
Distinguished Research Engineer
Career background
2011
- Joined NTT DATA Corporation
2007
- Completed a doctoral course for Ph.D. (Engineering).
2012
- Assigned to NTT DATA Intellilink Corporation Solution Business Division.
2015
- Assigned to NTT DATA Corporation R&D Headquarters.
- Joined NTT Software Innovation Center.
2016 to 2018
- University of California, Irvine
- Visiting researcher at Institute for Software Research.
Currently belongs to the Software Innovation Center. Having engaged in software engineering, in particular, research and development of upstream processes.
Research presentation
Paper（Journal） :
Shinobu Saito, Yukako Iimura, Aaron K. Massey, and Annie I. Antón, Discovering undocumented knowledge through visualization of agile software development activities, Requirements Engineering, September 2018, Vol. 23, Issue 3, pp. 381-399（2018.9）
Paper（International conference） :
Shinobu Saito, Yukako Iimura, Aaron K. Massey, and Annie I. Antón, "How Much Undocumented Knowledge is there in Agile Software Development?: Case Study on Industrial Project Using Issue Tracking System and Version Control System," Proceeding of the 25th IEEE International Requirements Engineering Conference (RE 2017), Sep. 4-8, 2017, Lisbon, Portugal, pp. 194-203. (2017.9)