Search

This paper points out various relationships between Design Knowledge and Software Engineering. After introduction of human design, the relationships between industrial Software Engineering is discussed, then further details of human design knowledge are revealed with discussions on humanistic aspects of design.

An approach towards developing a Knowledge Based Software Engineering (KBSE) tool by merging a conventional CASE tool with the expert system technology is introduced. This is found useful in assisting human designers. Experimental expert systems CREATOR2 and CREATOR3 are introduced and applied to the design of switching software. The CREATOR2 has the following features: representing software design knowledge, composed of design product knowledge and design process knowledge , using frame technology; and integrating knowledge based reasoning techniques with a SDL CASE tool. CREATOR3 is an extension of the CREATOR2 system. It enables one with additional design schemas for splitting a process, adding events, etc., and additional representation power, such as using pictorial elements and designers' comments in the frame representation. This leads to a uniform modeling and advanced reasoning environment for software design. Experiments on designing switching software are reported.

The World Wide Web (Web) allows people at remotely located sites to communicate and share their ideas using a common communication protocol. A common use of the Web system is running a client application, using a browsing tool, by pointing to a local or proxy server to browse data written in the hypertext format that contains anchors that address other URLs. In this paper a new application of the Web system for sharing knowledge-based systems and groupware development activities is introduced. An architecture for a Web-based Distributed Expert System (Ex-W-Pert System) is proposed and an implementation of the proposed architecture in groupware design is demonstrated. The resources and knowledge bases are distributed and can be accessed through the internet.

The paper reports on results of a study aiming at establishing a fundamental basis for automating design of any kind of software. Considering the final object, an automatic design learning human designer has been taken. An excellent software organization with high maturity has been taken as the expert, and the hierarchical work process is the knowledge model. For detailing, the major operations are made by hierarchical detailing. Namely, a software design may be reduced hierarchically to various design rules, which are parent and children relationship of a human concept created during expansion of a piece of design to more detailed form. To know the inside of human mental operations generating a design rule, a more detailed study was made and it was found that a design rule may be further reduced hierarchically to some fundamental human mental operations, called micro design rules. They represent basic operations during a design. After reporting them, a discussion on further lower level structure used during them, which are dictionary type definitions is made.

In the automatic programming area, users still fail to find available software components that match their needs faster than developing them again. In this paper, a new information retrieval technique to help in retrieving the software components from repositories based on different levels of accuracy (i.e., exact match, match, and similar) is presented. The software components had been classified into classes according to their functionality, and then they had been stored with what we call a 'data dictionary' in a repository. The exact match is not required to retrieve a required software component from this repository using our retrieving technique. An example of how to apply this technique in the automatic programming area to produce a program code list automatically is given. The feature of learning the structure of the new software component exists.

Programming errors are explained from the viewpoint of human error. Two kinds of programming error are identified: building-in errors during design and checking-out errors following the design. By separating error problems, key points for reducing building-in errors can be reduced to the questions of how to make the building-in rate small and how to make checking-out perfect. Error problems during design and testing are explained. Suggestions for reducing errors are given. >

This paper reports on an automatic software design system used for a switching control system. The design knowledge is gained from diagrams produced by a CASE tool during the initial human design, is stored in each corresponding expert system unit, and is reused for reproducing the design. The design procedure consists of splitting a call process, inserting tasks and detailing. Various considerations for implementing the system in an intelligent CASE tool are discussed.

This paper describes a design procedure with documentation for embedded software systems aiming at high quality. It starts from process level design. A system is decomposed to several orthogonal processes by data flow division. Making each block or process mono-functional and repeating the division, the system is reduced to a hierarchical cluster of Extended Finite State Machines. The design proceeds by converting an abstract level message sequence chart to multi-process message sequence charts, to a state diagram for one trace, and finally to a state diagram including all traces. Each state transition route is finally converted to source code. A design of an embedded system is thus partitioned to many small step designs each accompanied with the interfacing documents. Through careful, rigorous and repetitive documentation and check, the overall error rate is decreased. Experiments based on applying this method in developing skills of software design and education are reported.