JPH04169968A - Business processing specification generation method - Google Patents

Abstract

PURPOSE:To generate proper programs by obtaining the conformity of data structure while using a job knowledge base at the time of the nonconformity (difference) in the data structure. CONSTITUTION:A format 1 for plural kinds of slip data is defined in a slip data format definition processing 2 to be stored in a slip data format file 3. A job generation processing 4 extracts the reference/update relation between the data by using an job dictionary 5 and job knowledge 6 based on the slip data format stored in the slip data format file 3 and generates these processing specification 7 (transaction specification). At the time of nonconformity in data structure between the slip data in the reference(input)/update(output) relation, the conformity of both data structure is taken to generate transaction specification according to the structure. Thus, the transaction specification 7 for business processing job where plural slips with different data structure are related is automatically generated from the format of the slip.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention extracts the reference (input/update (output) relationship) of the data group from the form data group of the number of types of business processing specification generation rules, and The present invention relates to a method of recognizing the data structure of a group of form data, ensuring consistency between the structures when a mismatch (difference) occurs between the structures, and generating business processing specifications according to the structure.

[Prior Art] The following conventional examples can be cited as prior art related to the present invention.

Conventional Example (1) Japanese Unexamined Patent Publication No. 64-44541 (screen form design language method) The format is defined in a screen root data description language, and a screen form data design source program is generated.

The processing specifications for the part that processes screen form data are as follows:
Input from the keyboard. Generate an execution program from the screen form data design source program and processing specifications.

Conventional example (2) JP-A-1-147621 (program automatic generation method)
Definitions of files, form formats, processing specifications (expressed as conditional expressions (processing specifications for input data), generation expressions (processing specifications for output data) in the present invention), and tasks (relationship specifications for input and output data) are defined in the worksheet. Enter as an image. Furthermore, the long order of the data.

Enter the execution order and generate a batch processing job program.

Conventional Example (3) Japanese Patent Application No. 1-246290 (Form processing method) Input form data as an image (image processing), and generate the processing specifications using a form knowledge base.

[Problem to be solved by the invention]

(1) Prior Art (1) The program generated in this conventional example is a program for defining the format of screens and forms, and is not a program for processing screen and form data. In order to generate a screen/form data processing program, it is necessary to input the processing specifications.

(2) Prior Art (2) This conventional example is limited to batch processing operations, and cannot generate programs for interactive processing operations. Furthermore, it is assumed that the business (input/output data relationship specifications) and processing specifications are defined, and without these definitions, the program cannot be generated. Even when a business is defined, there is a structural mismatch between input and output data (if the input file is in ascending order and the output form is in descending order, there will be a mismatch in the order between the two. In this case, sort processing is usually required. ) occurs, so it is impossible to generate a correct program (or it is possible to generate it at all).

(3) Prior art (3) Business operations are made up of multiple types of data (form) processing intertwined (for example, inputting a slip, referencing and updating multiple ledgers based on this, and processing the ledger). duties such as outputting). However, in the embodiment of this conventional example, there is no description of operations involving a plurality of types of forms. The processing specifications generated by this conventional example do not correspond to the business processing described above.

An object of the present invention is to provide a business program generation method that should solve these conventional problems.

[Means for Solving the Problems] The method of the present invention is a method of recognizing the data structure of a plurality of types of form data groups from the formats thereof and generating business processing specifications according to the structure. a step of inputting formats for multiple types of form data groups; a step of extracting reference (input)/update (output) relationships of data groups from the formats using a business dictionary and a business knowledge base; There is a step to recognize the data structure of a group of form data in that relationship, and a step to ensure consistency between the data structures using a business knowledge base if there is a discrepancy (difference) between the data structures. and a step of generating business processing specifications according to the data structure based on the business knowledge base.

[Effect]

In the business processing specification generation method according to the present invention, processing specifications for a business group that processes a plurality of types of form data are generated.

Input multiple types of form data and extract their reference (input) and update (output) relationships (generate a business framework).

Then, by recognizing the data structure of the related forms,
Generate a processing specification according to that structure. Reference (input)・
If a data structure mismatch occurs between form data that are in an update (output) relationship, the data structures of both are made consistent and processing specifications are generated in accordance with the structure.

[Example〕 Hereinafter, one embodiment of the present invention will be described based on the drawings.

FIG. 1 shows a business processing specification generation method in this embodiment. Format l of multiple types of form data,
It is defined in the form data format definition process 2 and stored in the form data format file 3.

The business generation process 4 generates business dictionary 5. from the form data format stored in the form data format file 3. Using business knowledge 6, reference/update relationships between data are extracted and processing specifications (business processing specifications) 7 are generated.

FIG. 2 is a diagram showing a hardware configuration when the present invention is implemented with a single processor.

FIG. 3 is a flowchart showing the processing steps of form data format definition processing 2. This will be explained below.

An initial screen is displayed (step 301). A form data format definition command is input (step 302). If the command type is a "ruled line" setting command (step 303). Display the form definition screen (step 304)
,, If the command type is a "species attribute" setting command (step 303)', column attribute setting processing is performed (step 3o5).

Other commands are processed in the same way depending on the command type.

Enter the command on the form definition screen (Step 3)
06). If the command type is "return", step 307) returns to the initial screen.

“Horizontal line”, “Vertical line n, , ll diagonal line”, “Rectangle’”
, ``Erase horizontal line'', ``Erase vertical line'', ``Erase diagonal line'', ``Erase rectangle'', ``Line attribute'' command (
Step 307), the corresponding horizontal line setting process (
Step 308) and vertical line setting processing (Step 309) are performed. If the "end" command is input (step 307), the format definition processing is ended and a data storage confirmation message is displayed (step 310). Then, a confirmation command is input (step 311). If the command type is a “storage” command (step 312)
.

Enter the file name (step 313) and store the data (step 314).

FIG. 4 is a diagram showing an example of a form format defined in the form data format definition process 2.

FIG. 5 is a flowchart showing the processing steps of the task generation process 4. This will be explained below.

The formats of the form data stored in the form data format file 3 are extracted one by one (-sheets) (step 501). Until retrieval of form data is completed (step 502)
, repeat steps 503 to 505. Analyzing the structure of the form data (step 503), processing specifications (arithmetic expression).

conditional expressions, etc.) (step 504). In the document structure analysis step 503, the header section, detail section, and generalization (trailer) section of the document data are recognized, and the repetitive structure of the detail section is analyzed. Then, form processing specification generation step 50
4. Business dictionary 5. The business knowledge 6 is used to generate a processing specification according to the structure. When the structural analysis and processing specification generation for each form data is completed (step 502),
Extracts reference/update relationships between individual form data and generates business data flows showing these input/output relationships (
Step 506). Then, process specifications for each business that constitutes the business data flow are generated (step 507).
.

Figure 6 shows the form processing specification generation process (Figure 5, Step 5).
04) is a flowchart showing the processing steps of step 04). This will be explained below.

Form data is input and an instance frame is generated (step 601). The structure of the form data is projected onto the fact type knowledge (frame) in the business knowledge 6. The attributes of the form and the attributes of the items that make up the form are displayed on the frame. Next, references and updates on the frame are performed using the experiential knowledge of IF-THEN (step 602). Then, a processing specification is generated based on the updated frame information and output to the business processing specification file 7 (step 603).

FIG. 7 shows the frame generation process (FIG. 6, step 601).
) is a flowchart showing processing steps. This will be explained below.

Generate an instance frame for form data (
Step 701). Set the form name on the frame. For all items making up the form (step 702
, 703, 705), and generates an instance frame (step 704).

A keyword corresponding to the item name is searched from the keyword dictionary in the business dictionary 5, and attribute names are set for each of the six frames to be generated as child frames of the class frame corresponding to the searched keyword.

The processing procedure will be described later according to the flowchart of FIG. After setting the attribute name for each frame, a relationship is created between the instance frame for the form and the instance frames of the items that are its constituent elements (step 706).

An example of the instance frame generation procedure is shown in FIG.

Generate an instance frame of "sales amount" of "sales slip". Search the keyword dictionary for the keyword for "sales amount". The keyword "amount" is obtained (801). Next, the class frame "amount" is searched on the frame (802). Then, as a child frame of the class frame, an instance frame with sales amount J is generated (803).

The procedure for setting attribute symmetry for each frame will be explained.

There are various ways to name the individual items that make up the form data. When defining only the keyword part (ex
, code), a case where only a qualified name is defined (ex, customer), a case where a full name is defined (ex, customer code), etc. To deal with this, a common name is held as frame attribute information. Setting of attribute names is performed in the following steps shown in FIG.

Using the item name as a key, a corresponding item is searched from the keyword dictionary in the business dictionary 5 (step 901). If a search item exists (step 902), a search is performed by position (step 903), and the search item name is set as an attribute name (step 904). Search processing by position extracts items adjacent to or in contact with the attribute name, cuts out only the qualified name, and if the item name searched from the keyword dictionary includes the qualified name, sets that item name as the attribute name. .

The search pattern by location is shown in Figure 1O.

If the keyword does not exist in the dictionary (step 902)
), search the abbreviation dictionary in the business dictionary 5 (step 90
5). Step 9o6)
. The search item name is set as the attribute name (step 907). If it does not exist in the abbreviation dictionary (step 9)
06), the form item name is set as the attribute name (step 908).

For example, the procedure for setting an attribute name for the item "Code" is shown. Search the keyword dictionary for the item corresponding to "code". "Product code" and "Customer code" are obtained. Adjacent to “code” (pattern in Figure 10)
Extract item name.

Assume that the retrieved item name is "Customer Name."

Extract the qualified name "customer" from the "customer name" and verify whether the previously extracted "product code" and "customer code" match this qualified name. As a result, "Customer Code" is obtained, and "Customer Code" is added to the attribute name of "Code".
Set.

FIG. 11 is a flowchart showing the processing steps of the frame reference/update process (FIG. 6, step 602) in the input form. This will be explained below.

Extract all items of form data (step 1001.
1002.1007). It is determined whether the retrieved item is an input item (step 1003) and what its class is (steps 1004 and 1005). Then, frames are referenced and updated for each class and input/output classification, and processing specifications are generated (step 1006). Each processing specification is composed of the processing specifications shown in FIG. Specifications for each process are generated based on the process skeleton shown in FIG. The part marked with ★★★, ◆◆◆
Set the instance name in the part indicated by and generate it. The processing steps are shown below.

FIG. 14 is a flowchart showing the processing steps of ledger reference processing.

Ledger reference processing refers to processing that refers to ledger information using input form items as keys.

First, search for the ledger name to refer to (step zol)
. Remove the keyword part from the attribute name of the ledger reference item (input form item) and use it as the ledger name. For example, if the attribute name is set as "Customer Code", "Code J" keyword part is removed and "Customer Code" is set.
(Customer Ledger) as the ledger name. Next, the ledger reference key is searched (step 1102). Search for the ledger side item that corresponds to the ledger reference item (input form item). and,
Set the searched ledger name 1 ledger reference key to the skeleton of the ledger reference processing specification and generate the processing specification (Step 1
103).

The detailed steps of the ledger reference key search process (step 1102) are shown in FIG. A ledger reference item is searched (step 1201). An item corresponding to the ledger reference item (input form item) is searched from the ledger searched in step 1101. If a corresponding item exists in the ledger (step 1202), the corresponding item is set as a ledger reference key (step 1203). If it does not exist (step 1202), items containing the keyword of the ledger reference item (input form item) are searched from the ledger searched in step 1101 (step 1204). If it does not exist (step 1205), the ledger reference item (input form item) is set as a ledger reference key (step 1206). If it exists (step 1205), all attribute names of items including the keyword are extracted (step 1207), and attribute names of ledger reference items (input form items) are searched from among them (step 1208). If a corresponding item exists (step 1209), that item is set as a ledger reference key (step 1211).

If it does not exist (step 1209), the ledger reference item (
input form item) as a ledger reference key (step 1210).

FIG. 16 is a flowchart showing processing steps of other processing.

The processing specifications are generated by setting instance names in the parts indicated by ★★★ and the parts indicated by ◆◆◆ in the processing skeleton in FIG.

First, a processing skeleton corresponding to an input form item is extracted (step 1301). Steps 1302 to 1309) of extracting the operand of the processing skeleton, and extracting all instance frames (instance items) from the class frame having the same name as the operand (step 1304).

If it exists (step 1305), evaluate whether the extracted instance frame (instance item) is a component of the same form as the input form item (step 1306).
), if they are components of the same form (step 130
7) Set the instance frame (instance item) in the operand section of the processing skeleton. (Step 1
308).

As many as the number of operands (step 1303), step 1
Repeat steps 304 to 1309.

The processing specifications of the output form mainly consist of input form items and assignment statements (transfer statements) for reference ledger items.

An item corresponding to an item in the output form is searched from the input form and the reference ledger, and an assignment statement (transfer statement) for that item is generated.

Regarding the aggregation specifications (total amount, etc.), an item corresponding to the aggregation item is searched from within the output form, and an aggregation specification (ex, SUM (sales amount)) for that item is generated.

FIG. 17 is a flowchart showing the processing steps of frame reference/update processing (processing specification generation processing) (FIG. 6, step 602) in the output form. This will be explained below.

Extract the instance frame (instance item) corresponding to the item of the output form from the frame (Step 1
401.1409). For all items (Step 1
402), repeat the following.

Take out instance frame (instance item)
Evaluate the type of For trailer items (step 14)
03) Search the detailed class for the same class as the instance frame (instance item), and extract the instance frame (instance item). If the extracted instance frame (instance item) is a component of the same form as the instance frame (instance item) extracted in steps 1401 and 1409 (step 1404), generate aggregation specifications for the instance frame (instance item). (step 1405). For detailed items (step 1403)
, extracts another instance frame (instance item) of the same class as that instance frame (instance item), compares the attribute names of both, and extracts an instance frame (instance item) with the same attribute name from among them (step 1406 ).

Next, step 1407) extracts the two ledger names of the instance frame (instance item) extracted in step 1406, and generates a processing specification (step 1408). Extract the next instance frame (instance item) even above the frame (step 1409)
.

An example thereof is shown in FIG. Extract the ``commercial code'' from the ``A store sales report'' (■). Extract other instance frames (instance items) of the same class as the "quotient code". You will get a "product code" (■). "Commercial code"
The attribute names of "Product Code" and "Product Code" are the same. Next, the form name of the "product code" (instance frame (instance item) of the same class as the "commercial code") is extracted. rA store sales report” can be obtained (■)
. Then, a processing specification (posting (substitution) statement for "product code") is generated (■).

Next, the procedure for generating a business data flow will be explained.

FIG. 19 shows the flowchart.

The form data is taken out (step 1501).

The following is repeated for all forms (step 1502). Steps 1503 and 1504 of extracting the processing specifications generated for each item constituting the form data.
1506), the reference/update items in the processing specifications are analyzed to derive the relationship between input and output data (step 1505). Data with reference items is positioned as input, and data with update items is positioned as output (step 1507). The next form data is extracted (step 1508).

FIG. 20 shows an example of an input form. When we analyze the processing specifications for the "product code" of the "sales slip," we find the "product code" of the "product mask" in the reference items, so we position the "product master" in the reference (input) data.

Further, FIG. 21 shows an example of an output form. "Product code" of "Sales slip", [If you analyze the processing specifications for sales amount j, the reference item is "Product code" of "Sales slip", and the update item is "Sales amount" of "Sales mask", so refer to it. Position the "sales slip" as the (input) data and the "sales master" as the update (output) data.

FIG. 22 is a flowchart of the business process specification generation step shown in step 507 in FIG.

This will be explained below.

First, extract the business data flow generated in the previous step (
Step 1601). Repeat the following for all business data flows (steps 1602 and 1605).

The data structure of the input/output form of each business data flow is analyzed, and if there is a structural mismatch (described later) between them, it is resolved using the structure mismatch resolution rule in the business knowledge R6 (step 1603). .

Then, a business processing specification is generated from the input/output form structure (step 1604).

Rules for resolving discrepancies in the data structure of input and output forms will be explained. First, we will show the characteristics of office processing operations, and explain the principles of rules corresponding to them.

Software that processes office work needs to perform not only calculations but also data input/output control in order to organize and output program output results in a way that is easy for users to understand.

The following describes what kind of input/output control processing is required in office processing software.

(1) Data processing with different processing cycles Process input data at any time ("immediate" data processing) and output that data at a certain timing ("daily" data processing, "monthly" data processing, etc.) process. What is a processing cycle?
This is about the timing of data processing.

(2) Input/output data processing with different data structures (a)
A process of processing input data rR having a certain record arrangement (data structure), organizing the results into a different record arrangement OR different from the record arrangement of the input data, and outputting the result.

■I R' = (rl l LLl f! ”'+
rn)OR==(r4.nu, eh...,
rn)■I R: =(rnt rap rsr
``'r rn)OR: '' (two-yo r rnt)
rap rnt ”'+ rn) r, := r, :=
r,:=...:=rn:”Record r,:=r,,・
..., rn total record (b) A process of outputting a plurality of data OR++..., ORn with different order of output data records from input data IR having a certain record order.

I R: =(rnt rap rsr ”'+
r,,)OR1: =(rnt rso rap
”'* rn)ORj : = (rnt rn-
1, r, n-2, -, r, ) r, := r, := r, :
=...:”rnt:=Record (3) Aggregation process A process/operation that inputs input data IR with a certain record sequence at any time, calculates that data according to a specified aggregation formula, and outputs the result OR. Results (aggregated data)
Processing that processes subtotals, width totals, grand totals, etc., such that the data becomes the next input data.

Ding R: = (i r l, i r,
, i r, , −, i rn)1 r,:”I
r,:=l r,:””:=l rn2=record 1
r 1:”I r m:=l r m:;”・
y 1 rn'=(f lt f tp"・Hfn) f
, :=f, :=f, :=fn+=Item OR: ”= (
Or 1 g Q rat 0ray"'y 0rn)
Or, :”Or, :=Or, :=”'::O
rn:”Record ori:=Σir,j'fk (Vi
rj l 1nvariant (fk)) or r 1
:=Σori orm:=Σorlori :=o
rl:=orm:=Aggregation record The rules corresponding to the above processing will be explained.

(I) Data processing for "processing cycle mismatch""In data processing with different processing cycles J, there is a timing when input data and output data are processed, and until output data is created (until the specified processing cycle)
, it is necessary to buffer the input data. This processing can be defined as data processing in which the output record references multiple past input records in order to create output data, and such data processing can be defined as data processing with "processing cycle mismatch". I will call you. In data processing with [processing cycle mismatch], it is necessary to perform input processing, data accumulation processing, and output processing within the processing. In this case, data accumulation processing is added to data processing.

(2) Data processing for "mismatched order" (a) Addition to sort data processing Input/output data processing with different data structures such as the order of input data records and the order of output data records do not correspond to "mismatched order". ” data. Data processing that involves "mismatched order" requires sorting processing to match input and output data structures within the processing. In this case, sorting processing is added to the data processing.

(b) Addition to aggregation sort data processing In input/output data processing with different data structures, where the arrangement of input data records and the arrangement of records for aggregation items (arrangement of aggregation keys) do not correspond in chronological order,
An "inconsistency in order" occurs, and sorting processing is added in the same way as in (a).

(3) Data processing for "actual time discrepancy" (addition of aggregation processing) (a) Detailed record (rII rRe..., rn
) can be defined as data processing that outputs the aggregate record r, before outputting the aggregate record r, and can be defined as data processing that has a "real time discrepancy." ``cycle mismatch'' and ``order mismatch'' occur simultaneously, and after performing aggregation processing for r, (r-
It is necessary to output data in the order of v rat rat..., rn). In this case, aggregation processing is added to the data processing.

(b) Total record (rat rst ”・+ r
In data processing that outputs n) in the order of (r ny r n-1, ..., rl), as in (a), "real time mismatch" occurs, so it is necessary to add aggregation processing. There is.

In this case, the aggregate record (rll rII..., r
Aggregation processing is added in correspondence with n).

FIG. 23 shows an example of the step of generating business processing specifications from the input/output form structure (step 1604 in FIG. 21).

First, a processing specification regarding the input form is generated (■). Generate header/details input processing specifications (■-1) and header/details/trailer output processing specifications (■-1i
). In input processing, check specifications regarding input items are generated, and in output processing, calculation specifications regarding output items are generated. Next, if a structural mismatch occurs between the data structure of the input form and the data structure of the output form, a processing specification for resolving the mismatch is generated (■). If accumulation processing is required according to the rules shown above, create a processing specification that instructs to output the input form data to an external file or internal table, and if sort processing is required, create the data structure of the input form. Generate processing specifications that instruct conversion to the data structure of the output form. Then, processing specifications regarding the output form are generated (■). Generate calculation specifications for output items according to the data structure of the output form.

[Effects of the Invention] According to the present invention, it is possible to automatically generate processing specifications for office work in which a plurality of forms with different data structures are intertwined from the form of the forms. Regardless of the industry, a system template can be generated quickly from the form of the form. These verifications can prevent errors and omissions in the required specifications, and even designers and programmers with little business development experience can easily develop programs based on this template. Development time is shortened,
Productivity will be improved.

[Brief explanation of the drawing]

FIG. 1 is a flow chart showing an embodiment of the present invention, FIG. 2 is a hardware configuration diagram showing an embodiment of the present invention,
FIG. 3 is a flowchart showing the procedure of form data format definition processing, FIG. 4 is a diagram showing an example of a form format, and FIG.
FIG. 6 is a flowchart showing the procedure of business generation processing, and FIG. 7 is a flowchart showing the procedure of form processing specification generation processing.
The figure is a flowchart showing the procedure of frame generation processing,
FIG. 8 is a diagram showing an example of the frame generation procedure, FIG. 9 is a flowchart showing the procedure of attribute name setting processing, and FIG.
Figure 0 is a diagram showing search patterns based on position, Figure 11 is a flowchart showing the procedure of frame reference/update processing, Figure 12 is a diagram showing a list of each process that makes up the processing specification, and Figure 13 is , Diagram showing the skeleton of the processing specification, 1st
4 is a flowchart showing the procedure of ledger reference processing, FIG. 15 is a flowchart showing the procedure of ledger reference key search processing, FIG. 16 is a flowchart showing the procedure of processing other than ledger reference processing, and FIG. 17 is a flowchart showing the procedure of processing other than ledger reference processing. , a flowchart showing the procedure for frame reference/update processing in an output form, FIG. 18 is a diagram showing an example of the frame reference/update procedure, FIG. 19 is a flowchart showing the procedure for business data flow generation processing, and FIG. , FIG. 21 is a diagram showing an example of the business data flow generation procedure, FIG. 22 is a flowchart showing the procedure of the business process specification generation process, and FIG. 23 is a diagram showing an example of the business process specification generation procedure. . l... Form data format, 2... Form data format definition processing, 3... Form data format file, 4... Business generation processing, 5... Business dictionary, 6... Business knowledge, 7・
... Business Department No. 1 Figure 2 Opening 3 Figure 4 Figure 'f35 Figure Y q (2) u3rd /θ Figure 18 Opening'fI2θ Opening

Claims (1)

[Claims] 1. Reference to a data group from a plurality of types of form data groups (
Input) and update (output) relationships are extracted, the data structure of the form data group in this relationship is recognized, and the mismatch between the structures (
If a discrepancy) occurs, ensure consistency between the structures and create processing specifications (differences) for multiple form data in the relationship according to the structure
A method for generating business processing specifications), which includes a business dictionary that manages a group of items constituting the form data group, and a reference (input)/update (output) relationship of the form data group. A step of inputting the format of the data group using a business knowledge base that manages the rules used to generate specifications, and a step of referencing (inputting) the data group using the business dictionary and business knowledge. Extract the update (output) relationship, recognize the data structure of the form data group in the relationship, and if there is a discrepancy (difference) between the structures, ensure consistency between the structures, and perform operations according to the structure. 1. A business processing specification generation method, comprising the step of generating a processing specification. 2. In the business processing specification generation step according to claim 1, the step of recognizing the structure from the format of the data group, and the generation of processing specifications for each data unit using a business dictionary and a business knowledge base. a step of extracting the reference (input)/(output) update relationship of the data group in the processing specification to generate a business data flow group; and a step of generating the processing specification for each business data flow. 2. The business processing specification generation method according to claim 1. 3. The business dictionary according to claim 1 manages a group of items constituting the group of form data, and includes a keyword dictionary that manages keyword items, and a keyword dictionary that manages the group of items using abbreviations of the group of items as keys. 2. The business process specification generation method according to claim 1, further comprising an abbreviation dictionary. 4. The business knowledge base according to claim 1 includes a frame that manages the attributes of the form data group, a rule used to recognize the structure of the form data unit and generate its processing specification, and a reference for the form data group ( 2. Business processing specification generation according to claim 1, characterized in that the business processing specification generation comprises a rule used to extract an input)/(output) update relationship, and a rule used to generate a processing specification of a business in the relationship. Method. 5. In the step of generating processing specifications for each form data according to claim 2, the step of projecting the structure of the form data onto the frame of the business knowledge base (giving the attribute of the form data on the frame); Reference of the frame
2. The business processing specification generation method according to claim 1, further comprising the step of generating a processing specification by updating the processing specification according to a processing specification generation rule. 6. In the step of generating a business data flow group according to claim 2, the step of extracting the processing specifications of each data unit of the form data group, and the step of referring to the form data group (
2. The business process specification generation method according to claim 1, further comprising the step of analyzing the process specification and generating a business data flow group using input) and (output) update relationship extraction rules. 7. In the processing specification generation step for each business data flow according to claim 2, the step of extracting the business data flow group generated in the previous step, and the step of extracting the business data flow group generated in the previous step, and using the business process specification generation rule, )・(Output) A step of recognizing the structure of a group of form data in an update relationship, a step of making consistency between the structures when a mismatch (difference) occurs between the data structures of a group of form data in the relationship, and a step of 2. The business process specification generation method according to claim 1, further comprising the step of generating a business process specification according to the structure. 8. The business processing specification generation rule described in the claims includes a rule used to recognize the structure of the form data group shown in the business data flow, a rule used to ensure consistency between the data structures, 2. The business process specification generation method according to claim 1, further comprising rules used to generate a process specification according to the structure.