A project-based learning (PBL) education model that consists of three meta models, i.e. a resource pyramid meta model, an education process meta model, and an ability mapping meta model is proposed in practice. Most features of the model support various products such as the system software, web application, IoT program, and multimedia contents. Even if PBL products differ among the PBL education classes, fair student evaluations can be achieved because the meta models are dependent only on the maturity levels of the student abilities, and not the quality of the final products. In addition, the model was constructed based on empirical data of previous PBL projects. Based on this model, we conducted PBL classes in 2018 and 2019 at a university. The projects were focused on a business web application, project mapping system, programmable drone, and dynamic movie changing system. The students increased not only their programming techniques as novice members but also their project management techniques as team leaders through the four different projects conducted during the PBL classes. As a result, the teachers were able to fairly evaluate each student even though the final products of the PBL classes differed. At the same time, the students were able to grow as project managers or program specialists that they selected.

Recently, computer systems have become increasingly complex, involving a wide range of infrastructure and software technologies in the same system. Because these complex technologies affect each other, it is difficult to identify the causes of system faults while the system is running. Faults caused by a mix of software and infrastructure issues are particularly difficult for software engineers to resolve. We therefore propose a framework for identifying the causes of system faults that targets both software and infrastructure problems. The aim of this research is to identify faults that are related to software and/or infrastructure. The framework is based on system logs, including both software and infrastructure logs. When a fault occurs while a system is running, software and infrastructure logs are first collected and then compared with a large dataset of past logs to rapidly identify the cause of the fault. After that, debug mode level logs are collected in order to identify proper bugs of source code level and/or setting values of infrastructure configuration file level. The framework is also applied to two examples of real system faults. We find that the first fault was caused by software bugs, but the second was caused by both software bugs and infrastructure problems. We have also implemented a tracing and replay tool based on the framework. We tried to detect the true causes of 17 system faults that were randomly picked up past log dataset. 96% system logs in the 17 faults occurrence operation patterns were detected the differences between fault and normal cases. Then the two concrete session error faults occurred by two causes, infrastructure problems and software bugs. In this way, the framework will help software engineers to identify the causes of faults even when they find it difficult to separate software and infrastructure issues.

Recently, large-scale computer system is important for social human life.Such large-scale system is basically ordered through a competitive bidding.A development company that won the bidding starts developing.After that, customers use the computer system on their business operations.However, customers do not use some functions.Functions proposed in a system proposal in a bid phase are not same functions that customers frequently use in an operation phase.Functions change in various phases.Therefore, we propose process metrics for system quality.The feature of the metrics is target duration from a bid phase to an operation phase.Specification shifts of functions are clear in a simple rule.The metrics quantitatively measure specification shifts.We check active or non-active of each function on each phase.The total number of change of active or non-active of all functions is an important element of the metrics.Moreover, timing of the change of active or non-active of functions is also a significant factor of the metrics.As a result of applying a real project, we found that usage of the uncompleted package software makes development process complicated.

We began implementing contest-based learning with a blend of software engineering and business management 10 years ago. At first, a project subject was assigned. However, several problems occurred: For example, the students became absorbed in programming rather than design and analysis activities. Therefore, the curriculum changed from project subjects to contest-based learning. Business management, marketing, and accounting subjects were added to the new curriculum, and students made information technology (IT) business plans using their knowledge of software engineering and business management. The IT business plans were submitted to various contests held by public newspaper companies and the federation of economic organizations in Japan. As a result, in the 10 years of the contest-based learning implementation, 20 teams have received awards in various IT business plan contests. We investigated 10 persons who had experience submitting business plans. We confirmed that contest-based learning had clearer goals, such as to win the contest prize, compared to project-based learning. Further, the abilities to solve problems and to investigate increased more in comparison with lecture-style and project-style education.

We propose a basic metric for design sheets of infrastructure in order to improve infrastructure quality. The metric is called as LOI (Line Of Item). LOI is a simple metric of infrastructure for measuring scale and quality of infrastructure, as same as LOC (Line Of Code) of software. In addition, a class diagram of design sheets is provided in UML software design technique. LOI and the class diagram are applied to a real large-scale project. As a result, the whole of structure of the design sheets was clear, and, the scale of the infrastructure was able to be measured. In addition, the 40 system faults caused by the design sheets are analyzed. Correlation coefficient between LOI and the system faults was 0.483, it is weak correlation. Moreover, we confirmed that omission of design sheets such as blank columns may cause system faults, and copy actions of sub-design sheets may lead system faults.

We propose a process refactoring technique and we show experience reports. The process refactoring is based onprocess complexity and activity priority lists. Values of process complexity are calculated by additional processes. The priority list presents priority levels of all activities that do not finish. In an industrial large-scale project, the process refactoring was executed. The period of the project was 2 years and 6 months. Four times process refactoring were executed. As a result, customers' satisfactions were high although several customers' requests were not realized. In addition, vendor's atisfactions also were high. The vendor felt glad that resources about all activities were sufficiently prepared by the process refactoring. In addition, we confirmed that process refactoring was useful to narrow all customers' requests down into actual indispensable requests.

We have developed an interactive function named “student voice” on p-HInT system. The p-HInT is an educational environment including game terminals in order to improving large-scale lectures. The “student voice” function consists of three element functions; a sending messages function like twitter tools, “OK-understand!” and “?-Not understand!” buttons, and a red-card function. The “student voice” function was adapted to lectures of 10 subjects. In addition, we have also developed an analysis tool based on twitter logs (student voice logs) and lecture video. We tried to improve lectures of System Development theory subject based on results of analysis in the tool. As a result, teacher was able to communicate interactively with students, and the number of pushing “?-Not understand!” button decreased. 68% students felt more motivation and more interests in lectures.

New trends in software development, such as agile software development, have a difficulty in conventional document-based management. Executable software has a higher priority than development documents such as detail design documents and formal bug reports. If managers depend on development reports in order to determine the project progress and the product quality, they will miss the opportunity of determining the progress and quality in agile software development. Therefore, we proposed a project reliability growth model for determining the project state without development documents. This model is based on conventional software reliability growth models. The parameters related to bugs are replaced with communication topic parameters. The concept and procedure of the model are the same as those of the software reliability growth model. By applying this model to open source projects, it is possible to detect a significant change in the project state without development documents.

We propose a way of extracting micro processes that influence software complexity. Purposes of extracting the micro processes are (1) clarifying what micro processes influence software complexity, and (2) preventing software complexity from increasing by the micro processes. In the extraction way, at first, modules in which micro processes occurred are detected by interviews to developers of ¿known projects¿. Because relationships between complexity and occurrences of micro processes are analyzed, micro processes influencing complexity are clarified in ¿known projects¿. After that, we evaluate whether the micro processes similarly influence complexity of modules of the other ¿known¿ projects. As a result, three micro processes influencing complexity have been detected. After constructing difference models based on the relationships, the difference models have been adapted to ¿unknown¿ projects such as open source projects. Modules in which the micro processes might occur have been detected. In addition, we developed a programming editor including a function of automatic detection of occurrence of the micro processes. Using the editor, we tried to prevent source code complexity from increasing.

In large scale software projects, developers make much software during long term. The source codes of software are frequently revised in the projects. The source codes evolve to become complex. Measurements of software complexity have been proposed, such as module coupling and logical coupling. In the case of the module coupling, even if developers copy pieces of source codes to a new module, the module coupling can not detect relationship between the pieces of the source codes although the pieces of the two modules have strong coupling. On the other hand, in the logical coupling, if two modules are accidentally revised at same time by a same developer, the logical coupling will judge strong coupling between the two modules although the modules have no relation. Therefore, we proposed a visualization technique and software complexity metrics for software evolution. A basic idea is that modules including strong module coupling should have strong logical coupling. If a gap between a set of modules including strong module couplings and a set of modules including strong logical couplings is large, the software complexity will be large. In addition, our visualization technique helps developers understand changes of software complexity. As a result of experiments in open source projects, we confirmed that the proposed metrics and visualization technique were able to detect high risky project with many bugs.

Extracting requirements for large-scale software development has become increasingly complicated because many users from different organizations should collaborate with each other. Although use cases and UML diagrams for analyzing requirements are powerful tools, users often have difficulty understanding them. Therefore, we tried to extract detailed requirements via an empirical approach when we developed a plagiarism detection tool for students' reports. Key points of this approach are recording empirical data in an experiment and using an incomplete prototype based on component requirements. By comparing empirical data with output of the incomplete prototype, the detailed requirements are incrementally determined without additional efforts by the users. As a result of this approach, accuracy of the detection tool's output has reached 71%. In addition, we propose a framework for applying an empirical approach to requirements to adapt this approach to various developments.

In many software development projects, people tend to repeat same mistakes due to lack of shared knowledge from past experiences. Generally, it is very difficult to manually find out valuable phenomena from huge data. Invisible context, which cannot be known directly from software documents or formal reports, is an important factor to these difficulties. We propose a new method to find contexts based on analysis to email archives in a project repository. In this method, we first apply natural language processing to extract keywords from email messages. Next, similarities among the messages are calculated based on the extracted keywords, and the messages are classified into clusters according to the similarities. The clustering result can be presented with other information such as code growth graph or schedule charts. This method is implemented as an extension to the Project Replayer, a tool to review past project data. Pilot analysis confirms that a researcher could grasp important contexts of failures in actual projects using the Project Replayer.

We propose a new Web browser supporting asynchronous communication with Web server computers like Ajax approach. When users access Web applications on our browser, the users can receive similar benefits from Ajax even if the Web applications do not adopt Ajax approach. The browser has a transferable function from normal HTML sources to Ajax codes

We can receive large information from various Web sites using the Internet. The Web sites are important means to collect the required information in researching, learning, and even in commerce activities. Importance of Web browsers grows increasingly as Web sites increase. Web browsers are powerful tools to refer to Web sites. However, due to necessary of synchronous communication with Web servers, operationality of Web pages on Web browser is not better than operationality of desktop applications. Therefore, we propose a new Web browser for Ajax approach with asynchronous communication model. Our Web browser can improve the operationality that is equivalent to the operationality of desktop applications without revising program codes of the Web applications. A feature of the browser is partial updating of a Web page even if the Web application does not adopt Ajax approach. As a result of experiments, we have confirmed improvement of operationality of "Yahoo auction sites" on our Web browser when the load of the Web server increases

Recently, agile software development methods are popular because software should be developed in a short period. However, conventional project management techniques are often not adaptable to such new development methodologies. Especially, progress and quality control of the document-based management does not fit the agile software development. Therefore, a new project management support tool has been developed based on a new communication model. The new model focuses on relationships between communication and project states. The basic idea of the model is that good communication leads to high quality products and good progress of projects. Based on scattering ratio of communication on time series, the tool can calculate project state as a whole. From results of trial experiments, we confirmed that the strong correlation coefficient between the scattering ratios of communication and the product qualities. Furthermore, the tool was applied to open source projects. The projects' features were clarified in communication analysis using the tool.

Software development project managers have already known that adding manpower to a late software project makes it later. This is a famous sentence in "Mythical man-month" written by Brooks (1975). The managers also recognize that even if two man-month workloads are assigned to two developers, the development period is not one month. However, the managers want to see when the two developers finish the two man-months workload. Of course, the managers don't think that the development period reduces from two months to one month. If the development period in the two developers is shorter than two months, the managers will decide to add the new developer to the work. Therefore, we propose a new workload estimation model based on communications among shared works. The model includes the Concept Model of UML of a software development project. The classes and relationships in the Concept Model present the kinds of works and relationships among the works, respectively. Sharing works causes increments of communication workload. The communication workload is derived from the relationships among the classes in the Concept Model. In case studies using the model, we were able to reconfirm quantitatively the famous sentence of "Mythical man-month".

One of the major problems in object-oriented software projects is the lack of management's ability to comprehend and control the development progress of a project. This is because traditional phases of software development are not appropriate for object-oriented development. The project manager's "road map" is likely to be different with different phases, different milestones, and different checkpoints. This paper proposes a new framework which gives us a guideline for generating software process with relevant milestones for object-oriented development methods. The framework provides algorithms for identifying development phases and baseline products based on relationships among activities and products of the development method. In addition, the framework defines a software process model to manage a development progress in which milestones are established at the end of each phase in order to check the baseline product and establish goals for the following phase. Results of the application of the proposed framework show that the framework can generate a software process customized for well-known object-oriented development methods in a systematic way.

Many of the non conventional software development methodologies (such as object-oriented analysis methodology) and tools (such as visual programming environment) have been applied in real life projects. These projects have been started without sufficient previous training given to the developers. An increment in the productivity has been seen as the projects progress. This paper proposes a simulation model for software development which can deal with variances of developers' productivity during software development. As the proposed model takes into account the developer's learning curve, it can be used to compute a developer's productivity and the quantity of gain to the developer's knowledge in executing an activity. The proposed model has been applied to four typical scenarios in our case study. The results show that it is highly practicable. An outline of a project planning prototype which is based on the proposed model is presented. The prototype can be used to make project plans which take the developer's learning curve into consideration.

One of the major problems in object-oriented software projects is the lack of management ability to comprehend and control the development progress of a project. This is because traditional phases of software development are not appropriate for object-oriented development. The project manager's "road map" is likely to be different with different phases, different milestones, and different checkpoints. The paper proposes a new framework which gives one a rigorous guideline for generating software process with relevant milestones for various kinds of software development methods, especially for object-oriented development methods. The framework provides algorithms for identifying development phases and baseline products based on relationships among activities and products of the development method. In addition, the framework defines a software process model to manage a development progress in which milestones are established at the end of each phase in order to check the baseline product and establish goals for the following phase. Results of the application of the proposed framework show that the framework can generate a software process customized for well-known object-oriented development methods in a systematic way.