JPH02245827A - System for obtaining request specification knowledge in system design - Google Patents

Abstract

PURPOSE:To obtain efficient and flexible request specification knowledge by clarifying vague and obscure request specification knowledge and expressing it in graph structure through the use of a visual type request specification knowledge expression language while an interview is given. CONSTITUTION:When a user 10 himself describes request specification knowledge on a screen in graph structure by using a language system following the methodology of system design, a monitoring means 111 monitors it and monitor information is transmitted to a knowledge base means 142. When the analyzed result of monitor information by the knowledge base means 142 is decided that the support of the description work of the user 10 or the interpolation of request specification knowledge are required, an interview means 12 operates, an advice and a question are generated based on the knowledge base and they are displayed in the screen. Thus, the user himself can clearly grasp vague and obscure request specification knowledge which the operator 10 presents, and request specification knowledge can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a method for acquiring requirements specification knowledge in system design. Concerning methods for extracting and acquiring specification knowledge (information).

[Background Art] Requirement specification definition in system design and system design specification definition are recognized as important design needs for determining the functions and structure of the system to be designed. Figure 2 shows the results of the literature (R.

Ba1zer, et al.
Chnology in the 1990's: Us
ing a New Paradaigm, IEE
E Computer, Mol.

16, No, 11. pp, 39-45.198
3. ) is an example of a system design process that includes requirements specification definition disclosed in . In the current system design process that includes requirements specification definition, as shown in Figure 2 (b), the requirements specification definition and actual system design work are connected by extremely vague requirements specification descriptions.
Changes in requirement specifications, changes in system design specifications, verification, etc. are handled as separate processes.

In response to this situation, as mentioned in the above literature, some kind of new framework is required, and Figure 2 (
A more consistent system design process is being proposed, as shown in a). That is, formal specifications are created in the early stages of the process, and mechanical optimization, including coding, is introduced in the later stages to obtain formal development and complete programs. According to this, requirements specification definition and actual system design work are connected through a formal requirements specification description, creating a process in which changes in requirements specifications, changes in system design specifications, verification, etc. are handled as an integrated process. becomes.

For the requirements specification definition at the initial stage, an assistant system has been proposed to support the requirements specification designer, and at present it is not just an abstract framework, but several support systems are also being specifically developed and used. It looks like this.

[Problems to be Solved by the Invention] However, the following problems have been pointed out in such support systems for defining requirements specifications.

(1) Requirements specification knowledge, which is information that defines requirements specifications, is
The requirements specification designer is supposed to collect system requirements specification knowledge from the provider or the development client (hereinafter simply referred to as the user) by some means, and no means is provided for the collection process. It has not been.

(2) Requirement specification knowledge changes easily as user needs change, but no means are provided to flexibly deal with such changes.

(3) There are some areas in which the methodology for defining requirements specifications is incompatible with the methodology for designing the target system, but no means are provided to resolve this.

(4) In the first place, the requirements specification knowledge submitted by users is ambiguous and vague, and in many cases, users themselves do not truly understand the contents of their own requirements.

Therefore, even if a support system is used, the specifications can only be defined to the same extent as the support system.

Due to these problems (1) to (4), it is currently still difficult to define requirements specifications that truly reflect user requirements.

An object of the present invention is to provide a method for acquiring requirements specification knowledge in system design that is capable of extracting and acquiring efficient and flexible requirements specification knowledge by solving the problems of the prior art described above. be.

[Means for Solving the Problems] The method for acquiring requirements specification knowledge in system design of the present invention is a method for acquiring requirements specification knowledge used in system requirements specifications performed at the initial stage of system design. , which can be used by users themselves, configures requirements specification knowledge as a visual knowledge expression language based on the target system design methodology, and provides an expression means for expressing requirements specification knowledge as a graph structure based on this language system. , monitors the user's writing work of requirements specification knowledge based on the above language system performed on the display screen via the interface,
and monitoring means for generating the monitoring information.

In addition, it formalizes, stores and manages the requirements specification knowledge expressed in the above languages, absorbs changes in requirements specification knowledge, analyzes the above monitoring information, and supports users' writing work or requirements specifications. A knowledge base means for storing a knowledge base for generating advice and questions for supplementing knowledge, and a knowledge base means for storing a knowledge base for generating advice and questions for supplementing knowledge, and analyzing the above monitoring information based on the above knowledge base and writing work by the user according to the analysis result. The interview means generates advice and questions for providing support or supplementing requirement specification knowledge based on the knowledge base, and displays the generated advice and questions on a screen.

[Users use a language system to describe requirements specification knowledge in a graph structure on the screen through the action co-interface. Since the above language system is based on the methodology of target system design, the definition of requirements specifications expressed based on this language system is compatible with that methodology.

When the user himself/herself describes requirements specification knowledge in a graph structure on the screen using the above-mentioned language system, the writing work is monitored by the monitoring means, and the monitoring information is generated and sent to the knowledge base means.

If the results of the analysis of monitoring information using the knowledge base method determine that it is necessary to support the user's writing work or supplement the requirements specification knowledge, the interview method will be used to provide advice to support or supplement the user's knowledge. and questions are generated based on the knowledge base and displayed on the screen.

Therefore, the user proceeds with the writing work based on the above language system in accordance with the support or advice displayed on the screen. As a result, ambiguous and ambiguous requirements specification knowledge submitted by users can be clearly understood by the users themselves, and as a result, requirements specification knowledge can be efficiently extracted from users.

By repeating the above-mentioned interactive process, the final requirements specification knowledge expressed in the above-mentioned language system is acquired.

The acquired requirements specification knowledge is formalized and stored and managed in a knowledge base means. This allows requirements specification knowledge to be managed and changes in requirements specification knowledge to be appropriately responded to.

[Examples] Examples of the present invention will be described below with reference to FIG. 1, FIG. 3, and FIG. 9. Here, the subject is the acquisition of requirement specification knowledge in the Combi-1 Comunication System (hereinafter abbreviated as CC8).

Note that the expression language used in this system other than the graph structure is English, and in particular, the language displayed on the display screen is English itself, so as a general rule, drawings should be left in English. However, although it is not a screen display, drawings related to the screen display are also written in katakana.

FIG. 3 is a table showing the descriptive elements of the requirements specification knowledge expression language, which is a means for expressing requirements specification knowledge intended for use by users themselves.

This language is customized based on the requirements specification expression method provided by the CC8 design methodology. In this embodiment, in order to express requirement specification knowledge as a graph structure, it is configured as a visual knowledge expression language (hereinafter referred to as GRKRL) mainly consisting of icons as visual description elements.

Each element is a component for visually expressing requirement specification knowledge, and each element is given a name and an icon corresponding to the name is determined.

In the illustrated example, names are roughly divided into three types: objects as objects, resources as resources, and links indicated by arrows connecting elements. Objects are further divided into standard objects, terminal object group objects, command objects, and response objects. Also,
Resources are divided into two types: functional resources and system resources.

In addition, each element has internal attributes for describing detailed information on requirements specification knowledge.
)have. This internal attribute is further defined by the default domain (D
omain). Users express requirements specification knowledge as a graph structure based on this language system.

FIG. 4 shows an example of expression of the requirement specification knowledge described in this way. The content relates to a request to execute a logic simulation using CC8.

The outline is as follows. /Logic simulator as a stem resource 41
Two files are input. These two input files are the test pattern data file T.
PDF42 and circuit data file CDF43
are generated by system resources called a generator 44 and a computer (WSloo) 45. Test case data (TCD) 441 is input to the generator 44, and command F (Get) 451 is input to the computer 45. Ru.

The logic simulator 41 is a computer (VP 900
0) 46 system resource auxiliary resources (A-
RESOURCE) and outputs two results. One of the results is the result file (R-Fil
e) obtained as 47 and then computer 4
The formatting program (Forma
tter) A time chart (T-Chart) can be obtained using the system resource 48.

Another result of the logic simulator 41 is sent by a response end (E'nd) 411 as a response object to a send (Send) 412 as a functional resource,
Furthermore, it is output to the own terminal (My Termina 413).

The description of requirements specification knowledge using the graph representation of GRKRL is performed by the user who uses the requirements specification knowledge acquisition support system in accordance with a stepwise refinement procedure. This will be explained next.

FIG. 1 shows an example of the configuration of a requirements specification knowledge acquisition support system using GRKRL. This system includes an interface module 11 that directly supports the description of requirements specification knowledge using GRKRL, an interview module 12 that guides the work of the user 10, an operation module 13 that provides functions for various operations on requirements specification knowledge, and requirements specification knowledge. It consists of a knowledge pace monocle 14 that stores and manages information.

Knowledge base 142 in this knowledge base module 14
The definition contents shown in the table of FIG. 3, knowledge for analyzing monitoring information, knowledge for generating questions and advice, acquired requirement specification knowledge, etc. are stored in a formalized manner.

The interface module 11 provides a multi-window function, whereby the user 10 operates directly from the provided window, and the behavior of the user is monitored by the monitoring sub-module 111, and the information is sent to the interview module 12. Ru.

FIG. 5 shows a typical window configuration provided by the interface module 11. As is clear from the figure, the user can improve his or her requirements specification knowledge by selecting and executing GRKRL expression elements 511 and their operation functions 512 from the menu 51 displayed on the window. can be expressed. First, as shown in reverse video in the figure, a rectangular icon with diagonal lines indicating a system resource is selected from the expression elements 511, and moved to a predetermined position on the screen using the movement key 513. Next, the function name Logic Simulator 41 is assigned to the system resource.

The screen also shows an example of a multi-popup menu 52 used in this layout work. That is, a process in which a system resource is selected from a resource class, an application software class is further selected from this system resource class, a general application software is selected from the selected application software class, and a logic simulator to be allocated is selected. It shows. Note that the generator 44. which is already visualized on the screen. The computer 46 and the like are similarly selected and assigned and displayed.

The operation module 13 shown in FIG.
Execute the command selected from the figure).

The results are displayed on the window, and at the same time, they are all sent to the knowledge base module 14 and updated together with the contents of the requirements specification knowledge stored in the knowledge base 142 therein. All such management of the knowledge base 142 is performed by the knowledge base management system 141.

On the other hand, the interview module 12
It analyzes the information in the processing submodule 111 and provides advice, warnings, and requests for information to the user as necessary.

FIG. 6 is an example of a state in which advice from the interview module 12 is presented. In the figure, advice from the system is displayed in a sub-window 61 at the upper right of the screen. In the illustrated example, a logic simulator (Logic Sai
The user is requested to input the attribute value of a system resource called mulator) 41. For this reason, the bottom right of the screen also shows a state in which the user has opened a spreadsheet window 62 for attribute description based on the request, and has written attribute values in the spreadsheet.

Such an interview function operates according to the control flow shown in FIG. That is, in step 71 of FIG. 7, monitoring information on the user's behavior is received from the monitoring submodule l11. Based on this, in each phase of operation extraction and primitive extraction shown in steps 72 and 73, knowledge for analyzing the monitoring information is extracted from the knowledge base 74 of the operation system and GRKRL system corresponding to each phase. The monitoring information is analyzed and inspected at the knob 75.

Based on the results, questions and advice are generated in step 76 based on an interview-based knowledge base 77,
The generated questions and advice are fed back to the user in response step 78.1. Window 61.
62 (FIG. 6).

The spreadsheet window 62 for attribute description is configured as shown in FIG. That is, a spreadsheet window is generated based on spreadsheet information such as that shown in FIG. 8(a), and this spreadsheet information is stored in the knowledge base 142 shown in FIG. When the information is written, it becomes as shown in FIG. 8(b).

FIG. 9 is an example of a window display when acquisition of requirement specification knowledge is completed through an interactive process performed according to such a step-by-step detailed procedure. It can be seen that the same description of the requirements specification knowledge shown in FIG. 4 is displayed in the window.

As mentioned above, by using the requirements specification knowledge acquisition support system provided as above, users can proceed with the description step by step while visually checking the contents of their requirements specification knowledge. . The requirements specification knowledge obtained in this way is stored and preserved in a knowledge base, allowing iterative modifications and changes. Furthermore, subsequent design work can be performed continuously based on the requirement specification knowledge stored in the knowledge base.

As described above, according to this embodiment, the requirements specification knowledge can be expressed as a graph structure, so it is possible for the user to describe the requirements specification knowledge by himself/herself, and the expression element conforming to the design methodology can be added to the description. A requirements specification knowledge expression language with an icon can be used, and requirements specification knowledge can be written on the screen with the support of intervening means. Additions and changes can be made easily.

In addition, requirements specification knowledge can be expressed with the simple operation of selecting an icon from the menu displayed on the screen, writing a function name in the icon, and connecting it with a link. Since it is stored, it is possible to consistently manage and use requirements specification knowledge, and it is also possible to provide system designers with formalized design information.

Note that although the above-described embodiment deals with acquiring requirement specification knowledge in CC8, the present invention is not limited to this, but can be applied to a wide range of system design processes.

[Effects of the Invention] According to the present invention, vague and ambiguous requirements specification knowledge can be clarified and expressed in a graph structure while being interviewed on screen using a visual requirements specification knowledge expression language. Therefore, efficient and flexible requirements specification knowledge can be acquired.

[Brief explanation of drawings]

Figure 1 is a functional block diagram showing an embodiment of the requirements specification knowledge acquisition support system according to the present invention, Figure 2 is an explanatory diagram of conventional and future system design processes, and Figure 3 is a visual requirements specification knowledge representation. An explanatory diagram showing the definition of description elements of the language (GRKRL), Figure 4 is a block diagram showing an example of requirements specification knowledge expression based on GRKRL, and Figure 5 shows an example of requirements specification knowledge acquisition window such as menu selection and pop-up menu. The display screen shown in FIG. 6 is an advice window. A display screen diagram showing an example of the requirements specification knowledge acquisition window of the spreadsheet window, Fig. 7 is a control flow diagram of the interview module, Fig. 8 is an explanatory diagram showing an example of the spreadsheet configuration, and Fig. 9 shows the requirements specification in the final stage. FIG. 3 is a display screen diagram showing an example of a knowledge acquisition window. IO is a user, 11 is an interface module as an interface, 12 is an interview module as an interviewing means, 14 is a knowledge base module as a knowledge base means, 41 to 49, 411, 412,
413, 441, 451, and 481 are icons that are expression means for expressing requirement specification knowledge as a graph structure based on the visual knowledge expression language system, and 111 is a monitoring submodule as a monitoring means. Figure 1 of this example: Procedures included (self-proposal) December 22, 1989 Director General of the Japan Patent Office Yoshi 11 Moon Takeshi 1, Indication of the case 1999 Patent Application No. 065993 2, Invention Name of Acquisition Method of Requirement Specification Knowledge in System Design 3, Relationship with Amendment Person Case Patent Applicant Address (105) 1-7-12 Toranomon, Minato-ku, Tokyo Name (029) Representative of Oki Electric Industry Co., Ltd. Nobumitsu Kosugi 4, agent address (108) 4-10-3-6 Shibaura, Minato-ku, Tokyo.
Contents of supplementary IE (1) The claims are amended as shown in the attached sheet. (2) On page 6, line 18 of the specification, "for storing knowledge base" is corrected to "for storing knowledge." (3) Amend Figure 1 of the drawing as shown in the attached sheet. 5. Attachment to "Claims" and "Detailed Description of the Invention" of the specification subject to amendment In this method, requirements specification knowledge is constructed as a visual knowledge expression language based on the target system design methodology, which can be used by users themselves, and requirements specification knowledge is expressed as a graph structure based on this language system. a monitoring means for monitoring a user's description of requirements specification knowledge based on the language system performed on a display screen via an interface and generating monitoring information; It formalizes, stores and manages the requirements specification knowledge expressed in , absorbs changes in the requirements specification knowledge, analyzes the above monitoring information, and supports the user's writing work or complements the requirements specification knowledge. A knowledge base means for storing W money to generate advice and questions for the purpose, and a knowledge base means to analyze the above monitoring information based on the above knowledge base and support the user's writing work or improve requirements specification knowledge according to the analysis result. A method for acquiring requirements specification knowledge in system design, comprising: an interview means for generating supplementary advice and questions based on the above knowledge base and displaying the generated advice and questions on a screen.

Claims (1)

[Claims] A method for acquiring requirements specification knowledge used in the system requirements specification performed at the initial stage of system design, which can be used by the user himself and is based on the target system design methodology. Requirements specification knowledge is configured as a visual knowledge expression language, and an expression means for expressing requirements specification knowledge as a graph structure based on this language system, and an expression method based on the above language system that is executed on a display screen via an interface. A monitoring means that monitors the description work by users of requirements specification knowledge and generates monitoring information, and a system that formalizes, stores and manages requirements specification knowledge expressed in the above language system, and absorbs changes in requirements specification knowledge. and knowledge base means for storing a knowledge base for analyzing the monitoring information and generating advice and questions for supporting the user's writing work or supplementing requirements specification knowledge; The monitoring information is analyzed based on the above knowledge base, and based on the analysis results, advice and questions are generated based on the above knowledge base to support the user's writing work or to supplement requirements specification knowledge. A method for acquiring requirements specification knowledge in system design, comprising: an interview means for displaying questions on a screen.