JPS63233428A - Software development supporting system - Google Patents

Abstract

PURPOSE:To easily execute the development of software by expressing a business concept area and an information concept area of the concept of a business computer to lead in a computer with mutual relation through abstracting, dividing and depending links. CONSTITUTION:When a costumer request specification 1 is inputted, the specification 1 is translated by using a requested specification translating dictionary 3 to form an incomplete model 4 for the customer request specification. Then, the model 4 is interpolated 5 by using a software specification concept data base 6 to form a complete model 7 for the customer request specification 1. Then, an information concept indicating an object and money is extracted 8 by using a software specification concept data base 9 to form a usable software specification name/program name 10. Then, expression type retrieval 11 is executed from a software specification program expression type library 12 by using the specification name/program name 10 to form a software specification program expression type 13 and conversation correction 14 is executed to obtain a complete software specification/program 15.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a software development support system, and in particular, the accumulation and use of a software specification concept database suitable for software designers creating software documents and programs. , storage and retrieval of software specification information, and support tools [Conventional technology] A system designer, software designer, or program designer can create a system designer, software designer, or program designer from a requirement specification written in the language of the business to be computerized. When creating software using existing software specifications and source programs, first of all, a technology for accumulating a software specification concept database and a technology for using this are required.

Regarding the conventional management of software specifications, see
NCM, NISIG-OWN (ACM, ST
ρP of GOA) (June 1984).

+32-142 Form Document
Manager+tSys? ,en+S
PECD0Q...IIl, 5Archil.

ec? , are And Implemen7a
Discussed in tinnJ.

Conventionally, when developing software, software designers manually create documents such as external specifications and program specifications, or designers refer to similar system specifications that have already been developed. was created. However, manually creating specifications is prone to errors, and furthermore, changing specifications requires a lot of effort.

[Problem that the invention seeks to solve]

In the past, if you were trying to search for specifications at a certain level for an award-winning system that had already been developed, such as a customer's system, by entering the name of the customer's system and its level of detail, you would be able to find the specifications of the requested software. You can search for books or source programs. However, if the user lacks knowledge about the business system for which he or she is searching, there is little chance that the user will be able to search for what he or she wants. Users with little business knowledge must go through a great deal of trial and error in order to find what they want. In other words, in the past, consideration was not given to making it possible for users with little knowledge of business systems to search for what they wanted. Because it is a unit, no consideration is given to searching whether functions that do not exist in existing fields exist in specifications in other fields. Therefore, in order to find the desired specifications, search multiple fields and find the necessary specifications;! I had to go through a process of trial and error.

Traditionally, in addition to systems, projects have also been
Although it was used as a unit for 1F searches, it did not take into account which specifications should be searched preferentially, so users had to repeat trial and error.

Furthermore, in the past, software specifications were created manually,
It is necessary to make corrections, etc. In this case, even when developing a system based on a system developed in the past, the task is to read the specifications many times, extract similar parts from the past, and then create new specifications. The purpose of the present invention is to improve these conventional problems, and to create suitable existing software specifications that can be used from a database. and programs can be selected efficiently, and specifications that are similar to the desired system can be searched from existing specification information without being aware of the field or project, improving the efficiency of software development work. Our goal is to provide a software development support system that enables

Means for Solving Six Problems Between C] In order to achieve the above object, the software support development system of the present invention (1) transforms the concept of business in which a computer is to be introduced and the concept of a computer into a The business concept area related to money and the information concept area showing aspects of systems related to people, things, and money are expressed in relation to each other by concept abstraction links, concept division links, concept dependence links, and inter-concept message exchange links. The feature is that it is registered in the software specification concept database. (2) When registering specification information in the database, extract the data objects that appear in the specifications and use the data object abstraction information, data object segmentation information, and data object agreement information between the specifications for search purposes. After accumulating the information, it is first matched with data object abstraction information, then matched with data object segmentation information, and if matching fails, approximation is performed from data object agreement information. (3) When searching for specifications in a database, data objects are extracted from input search request information and search priority information is added. At the same time, this information is used in the consent data object search/replacement process when creating an abstract data object link to search for specifications. Furthermore, (4) in order to create a data flow diagram, input system information from the terminal according to the Japanese syntax displayed by the system, and from the input Japanese keywords, past created software specifications The feature is that it searches for and automatically creates specifications for a new system.

[For production]

In the first embodiment of the present invention, first, required specification information (customer required specifications) input by a user is analyzed, and a model of the customer required specifications is created. The model is incomplete because the customer's required specifications are insufficient.

Next, the software specification concept database is used to complete the requirements specification model. Next, use the software specification concept database to extract concepts (data types) in the information concept area that correspond to the requirements specification model, and find existing software specification names and program names that correspond to those data types. Next, the name of the specification you found,
Search for the phenotype corresponding to the program name. Next, the searched specifications and programs are modified to meet the user's requirements and used.

Next, in the second embodiment, data objects of existing specifications are extracted, and a database consisting of data object abstraction information, 5 data object/segmentation information, and data object agreement information between specifications is created. is created in advance, and the data object extracted by analyzing the input search request information is first matched with the data object abstraction information to search for the data object with the lowest level of abstraction. Next, by matching the searched lowest level data object with the data object segmentation information, it searches for a data object that can be seen from a different perspective, and also searches for the accompanying specifications. do. Furthermore, if a failure occurs during matching, the data object agreement information is used to replace the existing data object with a data object that has a meaning similar to that of the existing data object. These operation numbers make it possible to search for specifications at a detailed level similar to the input search request information.

Next, in the third embodiment, data objects are extracted from search request information input to search for specifications, and abstract data object links and divided data object links are added to each data object. At the time of creation, it operates so that it can be determined whether a consent data object search is necessary or not.

The information obtained in this way becomes information that narrows down the requirements of the system, and it is possible to search for the desired specifications.

In addition, in the fourth embodiment, Japanese text indicating a system outline is input from a terminal, keywords are extracted from the text, and specifications developed and registered in the past based on the keywords are stored in a file. 7. The retrieved specification #l# (data flow diagram) is synthesized, a new one-level coarse data flow diagram is created and displayed on the design document, and corrections are made if necessary. This makes it possible to improve efficiency in program design.

Modify the developed specifications and programs to meet user requirements and utilize them.

〔Example〕

Hereinafter, one embodiment of the present invention will be explained in detail with reference to the drawings. Figs. 1 to 6 are explanatory diagrams of the first embodiment of the present invention.

FIG. 2 is an overall configuration diagram of a specification information search system showing a first embodiment of the present invention. In FIG. 2, 16 is an input/output terminal equipped with a keyboard for input and a display for output; 17 is a computer processing unit for searching;
18 is a database in which specifications and the like are registered. The database 18 stores three types of sub-databases, which are a requirements specification translation dictionary 25, a software specification concept database 26, and a software specification/source program expression library 27, which may be the last software specification/source. The program expression library 27 is a library that includes not only an index but also a specification document or source program text. The computer processing bag [17 includes an overall control section 19. Requirements specification translation department 20. Model interpolation unit 21. Data type/operation name search section 221 Expression selection/search section 23. and an expression dialogue modification unit 24 are provided. The overall control unit 19 controls the translation of requirement specifications, the search for software specifications, and the modification of search results. The requirements specification translation unit 20 also inputs the requirements specifications of the customer requirement text input by the user (business function name, data type name, data type name, etc. expressed by the specification/program to be searched for, etc.) of the customer requirement text input by the user. , the requirements specification translation dictionary 25 is used to analyze the Japanese language in the requirements specifications.

Furthermore, the model interpolation unit 21 converts the customer requirement specifications into the translation unit 20.
For the incomplete requirements specification model obtained as a result of translation, concepts and associations are interpolated by pattern matching of software specification concepts in the software specification concept database 2G to create a complete model. The data type/operation name search unit 22 compares the data type name and operator obtained from the complete model of the requirements specification with the information concept (data type) and operator in the software specification concept database 26, and - The expression selection/search unit 23 searches the expression library 27 for an expression corresponding to the data type/operation name found by the data type/operation name search unit 22. The expression dialog modification unit 24 is a part that modifies the specifications and source programs of the retrieved software.

The requirements specification translation dictionary 25 has the following data structure. Note that the data structure is expressed using Equation 8 of LTSP, which is a type of computer language.

In the last part of the notation, different times are inserted depending on the individual data.

(Requirements Specification Translation Dictionary (verb base) (object road (case particle base)...) (characteristic property...)) (noun base) (characteristic (Semantic feature>...) (Abstract noun>Abstract noun>...)) Next, the software specification 4# concept database 26 will be described in detail.

FIG. 3 is a configuration diagram of the software specification concept database in FIG. 2. 26 is a software specification 1fi concept database, which corresponds to 26 in FIG.
27 is a software specification source program representation library, which corresponds to 27 in FIG.

This database 26 contains one person' concept plane 29, '
“Object” concept plane 30, “Money” concept plane 31, “Information” concept plane 32, and auxiliary computer reality 11L
It is composed of concept planes 33. That is, by classifying requirements and demands into substantive concepts and conceptual concepts that play a major role in business activities, it is made easier for even laymen to understand. Among these, the "one person" concept plane 29, the "thing" concept plane 30, and the "money" concept plane 31 represent the actual concept of the business that appears in the requirement specifications. Furthermore, the [information] concept plane 32 is a conceptual concept viewed from a different angle. Information concepts are "one person", "thing",
The business concept of ``R-Kin'' is mapped from the hard point of ``Information,'' and is mapped to the concept of computer software.

Each plane 29, 30, 31, 32. The individual concepts on and 33 will be linked by relationships between concepts, such as concept abstraction links, concept division links, concept dependence links, and interconcept message exchange links. . Further, the concepts on each plane are linked by a concept-dependent link such that, for example, the sales department 37 on the "person" concept plane 29 receives and receives the sales report 45 on the "information" concept plane 32. [Information] The information concept on the concept plane 32 is linked with the software specification library 46 and the source program library 50 in the expression library 27 by an implementation form association 49.

FIG. 4 is a diagram showing a method for realizing the software specification concept database in FIG. 3 on a computer.
G is concept table 91.92, expression related table (1) 93,
(IT) 94, and concept link table 95.
This is realized using tabular data structures.

The tabular data structure 5111 has associations indicated by arrows in the figure, and as a whole functions as the software specification concept database 2G. The concept table 91 consists of columns of concept name, concept category, operation category, operation name, and pointer, and the concept name is the concept name of other concept tables 92, tables, information concept names of present form related table 94, and concept link table 95. Each concept name is associated with an arrow. Further, the pointer in the concept table 91 specifies the corresponding functional specification name and program name in the expression association table 93.

Next, the software specification/source program expression library 27 is a live library that stores software specifications as documents and source programs as texts.

FIG. 1 is a flowchart of the processing of the database search method showing the first example of the present invention. First, when customer requirement specification 1 is input, translation 2 is performed using requirement specification translation unit #3. , an incomplete model 4 of customer requirement specifications is created. Next, model interpolation 5 is performed using the software specification conceptual database 6, and then a complete model 7 of customer required specifications is created. Next, the "information" concept (data type) is extracted using the software specification concept database 9 (8), and the usable software specification name and program name 10 are created. Next, using this,
Software specification program expression library 12
A representation search 11 is performed from , and a software specification program representation 13 is created. Next, dialogue modification 14 is performed to obtain a completed software specification and program 15.

FIG. 5 is a detailed operation flowchart 1 of the database search system of FIG. 2.

First, a request sentence from the user is inputted using the input/output terminal 1G (step 51).The input is a Japanese sentence from which concept names and operation names can be extracted. The requirement sentence inputted by the specification translation unit 20 is analyzed using the requirement specification translation dictionary 25 (step 52
). Next, the requirements specification translation dictionary 25 and the complete book control unit 19
extracts concepts (nouns) and operation names (verbs) from the analyzed results (step 53). Next, relationships between concepts are expressed by a network within the range that can be captured (creation of an incomplete model) (step 54). As a result, the model interpolation unit 21 uses the software specification concept database to create an incomplete model based on the network.
The model interpolation unit 21 collates the network of "person", "thing", "money", and "information J" (step 55), and then, by collation, the model interpolation unit 21 creates an incomplete model of the requirement statement based on the concepts in the software specification concept database. (step 56). Next, the model interpolation unit 21 creates a complete model of the request statement (step 57). Next, the data type and
The operation name search unit 22 follows links to find concepts (data types) in the information plane that are related to the concepts in the model (step 58). Next, a specification or program according to the operation name grasped from the model of the requirement specification in the above data type is found (step 59). Next, a list of available specifications and programs is output ( Step Go), and select the specifications and program to be used by looking at the settings (step 61). Next, the dialog is modified using the specification editor or source program editor (step G2). Through the operations of steps 51 to 62, the software specifications required by the user,
Or you can get the source program.

FIG. 6 is a diagram showing a specific example of the software specification ":!f concept database 7 database used in the present invention.

An example will be described below with reference to FIG. 6 in which a request statement is translated using a software specification concept database and an incomplete model created is interpolated into a complete model. 70 is 8
This is an example of a request sentence, and 7] is an incomplete model that is a translation of the request sentence. For the incomplete model 71, concepts and relationships between concepts outside the model 71 are extracted using the software specification concept database, and are associated with concepts within the model 71. Specifically, the sales department 77 sells 79 the best-selling products 73, inputs product sales data 81 into a computer, and receives and receives sales reports 82 (85). Further, the product headquarters 78 receives the sales analysis table 83 (86) and purchases the best-selling products 73 (80). From the above relationship, from the computer system side, product sales data 81. which is a concept (data type) on the information plane. The data specifications of the sales report 82, the sales analysis table 83, or the functional specifications or programs of the software that creates these are searched for and modified.7 In this way, in this embodiment, simply the name of the software document, Alternatively, system designers who are not familiar with the business can more efficiently search for software based on the customer's business requirements, compared to searching for keywords entered by the user using a dictionary of synonyms and synonyms (compound words) of the document name. In other words, in this embodiment, the customer inputs a request statement based on the concept of the business, and by using a database that stores relationships between concepts such as existing software specification concepts, A complete model expressing the requirements is created from a requirements specification model based on customer requirements that are complete and are written centering on business concepts, and the complete model is used to create appropriate existing software specifications11 or Since source programs can be searched, modified, and used, the number of man-hours required by those involved in software development can be significantly reduced.

In the present invention, by associating a business concept that represents an entity with an information concept used in computer software using an operation name that corresponds to a verb in a customer request sentence, existing software Since it is possible to search for, modify, and use appropriate items from specifications and source programs, even those who are not familiar with the business can produce software satisfactorily. Furthermore, since existing high-quality software can be used, the quality of the software being developed can be improved. In recent years, systems engineers and software engineers are increasingly required to improve their productivity, and one person may be in charge of multiple business software systems, or may have to tackle new tasks one after another in a short period of time. There are many cases where software is developed without being familiar with the business. In the future, the need for a method that allows the effective use of existing high-quality software by using incomplete requirement specifications of a detailed customer, even without business knowledge, as in this example, will become increasingly necessary. Now, a second embodiment of the present invention will be described with reference to FIGS. 7 to 11.

FIG. 8 is a diagram showing an example of the configuration of a computer system according to the second embodiment. In this embodiment, the interactive display terminal device FR118
, a computer processing unit 119, and a database 120 for storing software specification information. The computer processing unit 119 includes a search request information analysis unit 122 that analyzes search request information input from the interactive display terminal device @118 and extracts data objects, and performs matching of data objects and data object abstraction information. , a Yusu data object information creation unit 123 that creates abstract data object link information, and a divided data object link information creation unit that performs matching of data objects of abstract data objects and data object division information to create divided data object link information. unit 124, and a consent data object search unit 125 that replaces the existing data object with the consent data object if the matching process fails.
A specification information search unit 12G that searches for specifications based on divided data object link information, and an interactive display terminal device 1
A terminal control unit 127 that controls information input or output from the 18, and a specification information search control unit 21 that controls the execution of these are provided. Analysis dictionary 128
, data object abstraction information 129, data object division information 130, data object identity information 131, and specification information 132 are stored.

FIG. 7 is a search processing flowchart showing a second embodiment of the present invention.

In this embodiment, the dialog display terminal device shown in FIG.
18 uses the terminal control unit 127 to input the search request information shown in FIG. 7 (step 101), and the search request information analysis unit 122 refers to the request information analysis dictionary 28 for the input search request information By doing so, analysis processing of request information and data object extraction processing are performed (Step 1
02°103), obtain a data object. The extracted data object is processed by the abstract data object link information creation unit 123 shown in FIG. 104) and a matching check process (step 105).

Here, FIG. 9 is a diagram showing an example of information accumulation in a database according to the present invention, and FIGS. 10 and 11 are diagrams showing the relationship between data objects and specifications in the search processing process of the present invention.

FIG. 9(1) shows the storage format of the data object abstraction information 129 of the database 120 in FIG.
Figure (2) also shows the data object segmentation information 130.
9(3) is the storage format of the data object consent information 131, and FIG. 9(4) is the storage format of the specification information 132 in FIGS. 10(1), (2), and 11. In.

133-140, 14f3-148, 152-165 are data objects, and links 141-145, 150, 151゜156-164
It shows how to combine.

In step 103, data objects 133-140 shown in FIG. 1O(1) are extracted. In steps 4 and 5, the data object abstraction information 129 multiplied in the format shown in FIG. 9 (1) is referenced.6 Next, if the matching fails, the consent data object search shown in FIG. By referring to the data object consent information 131 stored in the format shown in FIG. 9(3), the unit 125 performs consent data object selection processing, that is, selects a consent data object to be used in a field similar to the target field. Search (step 19G).

Next, by performing abstract data object link information creation processing (step 107), link 1 is created for data objects 133 to 140 shown in FIG. 10 (1).
41 to 144 are combined.

The link at this time is data object 133.13
4,137, first, the link 141 from the data object 133 is temporarily connected to the data object 139. Next, link 141 from data object 139 is coupled to data object 133, and link 143 is coupled to data object 137, respectively.

Then, from the data object 137 to Rin, Ri 143
The data object abstract fhi information in FIG. 9(1) is also combined to connect the data object 134 to the data object 134, recombine only the double links as normal links, and delete the less than double links. and create it. In addition, when performing the same data object search using data object L3t3 in Figure 10 (1) as an example, the data object consent information shown in Figure 9 (3) is referred to and the target field is searched. As an approximated data object, Figure 10 (2)
Search for data objects 14G to 148 as in
Replace these. Then, it is checked whether these processes have been performed for all the data objects shown in FIG. 10 (2) (step 108 in FIG. 7).

That is, the determination is made by performing an abstract link creation completion check process. The created abstract data object link information is abstracted by the divided data object link information creation unit 124 shown in FIG. 8 by referring to the data object divided information 130 accumulated in the format shown in FIG. After performing a matching process (step 109) between the data object in the data object information and the data object division information, a matching check process (step 110) is performed. Next, if matching fails, perform the process of selecting a consent data object that approximates the target system (step 111) in the same way as when creating the abstract data object link information (step 106), and then split the data object. By the data object link information creation process (step 112), the first
Detailed data objects 152 to 15 as shown in Figure 1
It is determined whether or not all the processes for joining links 156 to 159 have been performed for 5 by performing a split link creation completion check process (step 113). Place 1! ! After completion, the terminal control unit 127 shown in FIG. 8 performs the specification name search process (step 114) using the data object name and operation name by referring to the specification information 132 accumulated in the format shown in FIG. 9 (4). By doing this, link 1 in Figure 11
Follow 60.162.164 to specification name 161,1
63. ] 65 ;! Next, the specification name 161,
163° and 165 are displayed on the interactive display terminal @118.

Next, the user of this system judges the displayed specification name and specifies the specification name that is considered necessary from the interactive display terminal device 118. Then, the terminal control unit 127
Specification information search center I! ! (Step 116) and
After searching the specification information 132 stored in the format shown in FIG. 9 (4) based on the input specification name, the terminal control unit 127 performs specification output processing (step 117), thereby allowing dialogue. Display terminal device i! 118 and display it, the desired specifications can be obtained.

In this example, by inputting the search request information, it is possible to obtain specifications that approximate the target system, so indicators during system design can be set early, and software development can be speeded up. It can be carried out. In addition, even inexperienced 30% engineers can easily obtain existing highly reliable specifications, so they can design reliable systems and improve software productivity. .

12 to 16 are explanatory diagrams of a third embodiment of the present invention.

FIG. 12 is a block diagram of a computer system showing a third embodiment. As shown in FIG. 12, the system of this implementation (σ) has seven computer processing units 202 consisting of an interaction terminal r!1201, a computer processing unit rI1202, a database 203, etc. A terminal control unit 2 that inputs information and search priority information and outputs specifications obtained by the search to the interaction terminal device 5!201.
05, the search request information analysis unit 20G that extracts data objects from search request information, and the abstraction of data objects! ! an abstract data link creation unit 207 that creates information indicating relationships, a divided data link work unit 208 that creates relationship information of detailed data objects associated from abstract data link information, and a link creation unit 208 that creates information indicating relationships. If information cannot be created, a consent data object search unit 209 searches for a data object that agrees with the existing data object and replaces it, a search priority determination unit 210 controls data link creation processing according to search priority information, etc., and split data. A specification search unit 211 that searches for specifications from link information and a search control unit 204 that controls these processes are provided. The database 203 also includes a search request information analysis dictionary 212 for analyzing search request information, data object abstraction information 213 as information indicating the abstraction relationship of data objects when creating an abstract data link, and divided data At the time of data link creation, data object division information 214 as information indicating the detailed relationship of data objects, data object/consent information 215 as consent information of data objects, and specification information 216 are stored.

13 is a process flowchart of FIG. 12, FIG. 14 is a flowchart 1 of data object priority search processing in the data link creation process of FIG. 13, and FIG. FIG. 16 is a diagram showing a group of information given to the processing shown in the figure, and FIG. 16 is a diagram showing information obtained by the processing shown in FIGS. 13 and 14.

In this embodiment, the search control unit 204 shown in FIG. 12 inputs search request information to the interaction terminal device 201 under the control of the terminal control unit 2.5 in order to obtain search request information (step 200). . Next, in order to extract a data object from this information, the search request information analysis unit 20G
performs data object extraction processing according to the search request information analysis dictionary 212 of the database 203 (step 210), and in order to add search priority information to the data object obtained here, the terminal control unit 2
05 performs processing (step 220) for assigning search priorities to data objects on the dialog terminal device 2Q1, and the data object 21 as shown in FIG. 16(1) is
7 to 222 are output.

The process at this time is that the terminal control unit 205 outputs a table as shown in FIG. 15 to the dialogue terminal device 201. The boxes 217 to 255 shown in FIG. The data objects are data objects, and the lines 226 to 145 connecting these data objects are data links indicating mutual relationships between the data objects and the data objects, respectively.

Then, when the user inputs the priority, the end song control unit 205 adds this information to the data object. In the example shown in FIG. 15, the data object rAAA
”rAABJrAACJrBBBJrCCAJrDAD
J is extracted from the search request information, and "C", "A", "C", "B", and "n" are respectively extracted for these data objects.
Add a search priority of "C".

In addition, priority rAJ indicates that no search is required, that is, this data object is deleted when creating a data link,
Priority "rB" means an arbitrary search, that is, even if a link to this data object cannot be created at the time of data link creation, replacement with the consent data object is not performed. Furthermore, the priority rCJ indicates a mandatory search, that is, if a link to this data object cannot be created at the time of data link creation, a consent data object is searched and replaced.

The abstract data link creation unit 207 performs abstract data link creation processing (step 230) and termination determination processing (step 240) on the data object to which the search priority and ranking have been added, according to the data object abstraction information 213 of the database 203. By this, the first (3
Data objects t-217 to 222 in FIG. 16 (1) are data objects l-21 as shown in FIG.
Links 226 to 231 are connected to links 7 to 225, and information indicating abstract relationships between data objects appearing in the search request information is created. Figure 16 (2
), the data object 217 is composed of data objects 218, 219, and 223, which means that the data object 218 is composed of the data object 221, the data object 223 is composed of the data objects 224 and 225, and so on. In the process of link combination, the search priority order information is used by the search priority order determination unit 210 in the order shown in the processing flow of FIG. 14. First, the search priority is checked (step 261), and if the data object at this time has priority "A", the data object is deleted (step 262). Furthermore, when the data object has the priority rT3J shown in FIG. 15, the data object abstraction information search process (step 2
63) to perform link combination. At this time, even if the search combination is NG, the data object is not deleted and is left as -j:. Then, when the data object has the priority level "C" shown in FIG. 15, data object abstraction information search processing (step 265) is performed and link connection is performed. If the search result at this time is NG, the consent data object search unit 209 retrieves the consent data object information 215 from the database 203.
Accordingly, the same sound information search and replacement process (step 267) is performed to replace the data object with a data object that is approximate or has the same meaning, and the data object abstraction information search process (step 265) is performed again. Figure 16 (1
) to FIG. 16(2). In the example of data object link N, data objects 217, 218° 220 have priority [C", data object 21! ],
221 has priority level "B", and data object 222 has priority level rAJ. Data object 217 is data object 2 in data object abstraction information search.
Since 18 was searched, link 226 is combined. The data object 218 combines the link 228 by searching for the data object l-221 in the data object abstraction information search. data object 22
1 was NG in abstract information search, but priority r13J
Therefore, it is left without being deleted. The data object 219 is the data object 217 in the abstract information search.
, so link 227 is combined. Data object 222 has priority "A" and is therefore deleted. Data object), 220 was NG in the data object abstraction information search, but since it had a priority of "C", the consent data object was searched and replaced with data object 223. data object 2
23 has searched for data objects 217, 224, and 225 by data object abstraction search, so links 229, 230, and 231 have been combined.

The abstract data link information is generated by the divided data link creation unit 20
8 performs the divided data link creation process (step 2.50) and end determination process (step 260) in FIG. 13 according to the data object division information 214 of the database 203, and the data object 217 in FIG. data objects 232, 233 and links 239, 240, and data object 223 with data objects 234, 235 and links 241, 24.
2, and data object 224 and data object 238 and link 245, and data object 225 and data object 236 and link 2.
43 and 244, respectively, to create a system of data objects 232, 234, 236 and a system of data objects 233, 235, 237, 238 for the abstraction relationships of data objects 217, 223, 224, and 225.
Create information indicating a refinement relationship with a system of At this time, the data object E extracted from the search request information that remains when creating the abstracted data link is used to create the divided data link, and the search priority is also used in the same way.

Based on this divided data link information, the specification search unit 211 performs a specification search process (step 270) and selects a specification from the specification information 21G of the database 203, and then the terminal control unit 205 Specification output processing (step -280) is performed on the specification sheet.

In the example of FIG. 16(3), the data object 233
links 248 to specification 249, data object 235 links 250 to specification 11' 251
The data object 237 can obtain the specification tlHF 253 through the link 252, and the data object 238 can obtain the specification 255 through the link 254, respectively.

In this way, in this embodiment, the database 203
When searching the data object abstraction search 2'13 and the data object segmentation information 214 shown in Figure 2, it is specified whether or not to search the data object consent information 215 by adding a search priority to the search request information. Therefore, it is possible to search for specifications that narrow down the user's Φ requirements, and it is possible to obtain specifications that approximate the target system. Therefore, even an inexperienced designer can quickly design a system or program based on highly reliable specifications.

Next, a fourth embodiment of the present invention will be described with reference to FIGS. 17 to 22.

FIG. 17 is a functional block diagram of a search system showing a fourth embodiment of the present invention. In FIG. 17, 301 is a display terminal equipped with a keyboard for inputting data;
302 is a functional unit that reads the data (Japanese text); 30
3 is a functional unit that extracts keywords from the loaded Japanese text,
304 is a functional unit that searches for a specification (data flow diagram) from the file 305 based on the keyword, and 306 generates one existing data flow diagram from the searched multiple specifications and displays it on the terminal. A functional unit 307 is a functional unit in which the design document makes modifications to the flow diagram. In addition,
Here, two display terminals 301 are provided.

FIG. 18 is an operation flowchart of FIG. 17, and FIG. 19 is a diagram showing an example of Japanese sentences.

First, a Japanese sentence expressing the existing system requirements is read from the terminal 301, and as an example of the format of this Japanese sentence, a sentence as shown in FIG. 19 (1) is read (step 308). The part shown here is where the designer fills in the text, and the other words in FIG. 19 (1) are presented by the system. Assuming that the sentence shown in FIG. 19 (2) is input from the terminal, the noun part is extracted from the given Japanese sentence in FIG. 18 (step 309). Here, the first
This means that the words surrounded by the boxes in Figure 9 (3) have been extracted. In other words, the keywords extracted are "rpos terminal", "operator", "sales management", "accounting management", and "single item management". Keywords that indicate "sales management", "accounting management", "single item management"
indicates the function to achieve it.

Next, existing data flow diagrams are searched from the file based on the three keywords representing these functions (step 310). Here, a plurality of items (three items) are searched and extracted. It is now assumed that three data flow diagrams shown in FIG. 20 have been retrieved. These three data flow diagrams express input data and output data to realize each function. These flow diagrams, as shown in the Japanese text shown in Figure 19 (2) and (3), are useful when building a store management system that has the functions of sales management, accounting management, and single item management. It is an attempt to In order to combine these three functions into one series of functions, common data is extracted between each function such as POS, operator, sales table by terminal, sales table by operator, and ink face information. Then, a summary data flow diagram as shown in FIG. 21 is generated (step 311).

This is displayed on the terminal (step 312) and checked by the designer, but of course there is information that the designer does not need or is missing, so next we will modify this diagram. (Step 313).

In other words, after deleting, adding, changing data, etc. and eliminating unnecessary or missing data, this data flow diagram is registered in a file (step 314). It is registered in association with the data flow diagrams of ``Management'', ``Accounting Management'', and ``Single Item Management''. Then, when necessary, it is output to a printer (step 315).

FIG. 22 is a diagram showing the link relationships of each data flow diagram. In FIG. 22, the dotted bidirectional arrow indicates the new link information. In this way, when trying to develop the next system by linking and registering newly generated items one after another with the previous ones,
If you enter the keyword ``store front management,'' this data flow diagram will be searched and can be reused.

In this example, by inputting Japanese sentences that follow a certain syntax, it is possible to search for previously created data flow diagrams and automatically generate a new data flow diagram. When developing a system for
As software specifications created earlier are accumulated one after another, and as the number of development cycles increases, the reuse rate improves and development efficiency increases.

〔Effect of the invention〕

As explained above, according to the present invention, even a user who is not familiar with M connection methods or operations can input request statements for software specifications, source programs, data flow diagrams, etc. stored in a database. Since similar specification documents with high reliability can be checked, the required number of man-hours can be significantly reduced, and the quality of the software being developed can be improved.

[Brief explanation of drawings]

FIG. 1 is a search system showing a first embodiment of the present invention. FIG. 2 is a configuration diagram of a computer system showing the first embodiment; FIG. 3 is a diagram showing the content structure of the software specification concept database in FIG. 2;
The figure shows a method for realizing the software specification conceptual database in Fig. 2 on a computer, Fig. 5 is a processing operation flowchart of this embodiment, and Fig. 6 shows an example of implementing the software specification conceptual database of this embodiment. 7 is a processing operation flowchart showing a second embodiment of the present invention,
゛Figure 8 is a block diagram of the computer system of the second embodiment,
FIG. 9 is a diagram showing an example of information storage contents of the database in FIG. 8, FIGS. 10 and 11 are diagrams showing the relationship between data objects and specifications when operated according to FIG. 7, and FIG. 13 is a block diagram of a computer system showing a third embodiment of the present invention, FIG. 13 is a processing flowchart of the system of the third embodiment, and FIG. 14 is a diagram showing link information creation using the search priority information in FIG. 13. 15 is a diagram showing an example of inputting search priority information,
FIG. 16 is a diagram of the relationship between data objects and specifications as the process progresses in the third embodiment, and FIG.
A block diagram of the search system showing the embodiment, FIG. 18 is a processing flowchart of the fourth embodiment, FIG. 19 is a diagram of the input Japanese sentence format in the fourth embodiment, and FIG. 20 is a diagram of the fourth embodiment.
FIG. 21 is a newly generated data flow diagram, and FIG. 22 is a diagram showing the link relationship of each data flow diagram. 1: Customer requirement specifications, 2: Translation function block, 3゜25:
Requirements Specification Translation Dictionary, 4: Incomplete Model of Customer Requirements Specifications,
5: Model interpolation function block, 6゜26: Software specification concept database, 7: Complete model of customer requirement specifications, 8: "Information" concept (data type) extraction function block, 9: Software specification concept database, 10: Available software specification name, program name, 11; Expression search function block, 12, 27: Software specification, program expression library, 13: Software specification, program expression type, I4: Dialogue modification function block, 15: Completed software specifications, programs,
113, 118, 201, 301: input/output terminal, 17,
119,202: Computer processing device, 18,120,20
1 database, 19; overall control unit, 20: requirements specification translation unit, 21: model interpolation unit, 22: data type/operation name search unit, 23: expression layer selection/search unit, 24: expression versus word correction unit, 29 : Plane of human thoughts, 30; Plane of object concepts,
31: Total concept plane, 32; Information concept plane, 33
: Calculator': A current concept plane, 91, 92: Concept table,
93, 94: Expression related table, 95: Concept link table, I2
1: Specification information search control unit, 122: Search request analysis unit, 1
23: Abstract object creation unit, 124: Data object link information creation unit, 125: Agreement object search unit, 126: Specification information search unit, 127, 205: Terminal control unit, 20G: Request information analysis unit, 207: Data link creation unit, 208: Divided data link creation unit, 209
: Homophone data object search unit, 21o; Search priority determination unit, 211: Specification search unit, 212: Search request information analysis dictionary, 213: Data object abstraction information,
214: Data object segmentation information, 215: Data object consent information, 216: Specification information, 302
: Japanese sentence reading function block, 303: 1i4 sentence analysis keyword extraction function block, 304: Data flow diagram search function block, 305: File, 306: lj
E-simplified data flow diagram generation, display function block, 3゜7:
Flow diagram confirmation correction, storage function block. Figure 7 Figure l O Figure 11 Figure 13 Figure 14 Figure 15 Figure 16 Figure 16 M l a Figure 19 (1) The system is
This is to realize the following. ■ The store management system realizes rMS0guchi by using POS terminals and operators who operate them. Figure 20 Figure 20

Claims (1)

[Claims] 1. A file for storing software specifications and source programs, an input/output terminal means, means for analyzing search request information input by a user from the input/output terminal means, and a means for analyzing search request information inputted by the user from the input/output terminal means. An abstraction concept that appears in a software specification and a source program, and a division of the abstraction concept, in a software development support system using existing software, which has means for searching the above-mentioned files for a matching software specification and source program. extract concepts, combine the extracted concepts into a link structure, and store it in the above file.
The relationships between the concepts appearing in the user's request information are checked, and the concepts with a high degree of agreement and the software specifications and source programs related to the concepts are selected by following links, and the selected specifications are A software development support system characterized by modifying documents, etc. and creating new specifications, etc. 2. As the above-mentioned abstracted concepts and divided concepts, the business concepts and computer concepts that appear in the business in which computers are to be introduced are divided into substantive business concepts representing "people,""things," and "money," and the substantive It is characterized by providing a software specification concept database that separates business concepts from lateral "information" concepts and relates them through concept abstraction links, concept division links, concept dependency links, and interconcept message exchange links. A software development support system according to claim 1. 3. For data objects that appear in software specifications as the above abstraction concepts and division concepts,
Abstraction and division information between specifications, respectively,
A database is provided that stores data objects with consent information, and data objects extracted from request information are first abstracted by combining them into a link structure in units of one level higher and one level lower based on the degree of matching with abstracted information. After generating data object link information, link the data objects from the highest level to the lowest level of the abstract data object link information in units of one level higher and one level lower in the direction of division based on the degree of matching with the segmentation information. 2. The software development support system according to claim 1, wherein divided data object link information is generated by linking to a structure and an existing specification document is searched. 4. In the process of searching the existing specifications mentioned above, if the matching degree test fails, the data object is replaced with a data object having a similar meaning even if it is in a different field based on the data object consent information. A software development support system according to claim 1 or 3. 5. When comparing the accumulated contents of the above file with concepts related to the user's request information, extract the data object from the request information, add search priority information to the data object, and compare it with the data object. According to claim 1, the search priority information is used for search and replacement of agreed data objects when creating abstract data object links and divided data object links to search for a desired specification. software development support system. 6. When inputting the above-mentioned user request information, if you input a specification expressing the system outline in Japanese according to a certain rule, the search request information analysis means reads the Japanese sentence,
After parsing the syntax and extracting keywords and searching for the data flow diagram from the above file using the keywords,
Generate and display several searched data flow diagrams,
2. The software development support system according to claim 1, wherein the data flow diagram is modified.