JP2019211964A - Check rule creation device and check rule creation method - Google Patents

Abstract

To create a check rule capable of being uniquely interpreted for a check list checking that there is no problem in work related to information systems.SOLUTION: A check rule creation device comprises: a check list pre-processing unit 101 for extracting predetermined words from description included in a check item and creating a shaping check list listing the words for each check item of the existing check list; a natural language conversion information accumulating unit 112 for accumulating natural language conversion information associating a matter related to an information system and a word indicating it; an ambiguity evaluation unit 102 for extracting an ambiguous part as the check item of the information system in the shaping check list by comparing the word of the shaping check list and the word of the natural language conversion information; an ambiguous part complementing unit 103 for estimating the contents of the ambiguous part based on the natural language conversion information; and a check rule conversion unit 104 for creating a check rule by adding the estimated content to the extracted ambiguous part.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a check rule creation device for creating a check rule for checking that there is no problem in work related to an information system.

  In the development, operation, and maintenance work of information systems such as enterprise systems of enterprises or public offices, the person in charge of the work works while checking the check items in the checklist created in advance so that the work is not overlooked. I do. The checklist may be created by the person in charge of the work to be checked, or may be created by the person in charge of other work.

  The checklist creator is responsible for creating checklists such as guidelines on the development, operation, and maintenance of information systems to be checked (hereinafter also referred to as “check target systems”) and checklists of other existing information systems. While checking available information (hereinafter also referred to as “reference information”), check items suitable for the system to be checked are created manually.

  However, the reference information may be described in an ambiguous sentence. Also, the check items to be listed in the check list differ for each check target system. For this reason, it takes time to individually create a check list for the check target system. A technique for reducing the work of creating such a checklist is required.

  In Patent Document 1, check items are stored in advance in a database (hereinafter also referred to as “DB”), and when conditions for work to be performed are specified, check items necessary for checking work from the DB are determined based on the conditions. It is described that it is extracted and a new checklist is created.

  In Japanese Patent Laid-Open No. 2004-228867, a phrase combinator that does not receive a noun phrase is used for accurate syntactic analysis and semantic analysis of a sentence in which essential components such as a subject and an object are omitted. ”Is specified, and the part of speech of the omitted component is specified.

  In Patent Document 3, when a character string composed of a plurality of words separated by a specific character such as a space is input to a character input box such as a search box, if a word in the character input box is input, it is input in the past. A technique for predicting and complementing a phrase that will be input together with the phrase based on the history of the character string is disclosed.

JP 2012-53755 A JP 2007-133905 A JP 2007-58444 A

  However, since the technique of Patent Document 1 extracts and uses check items registered in the DB, the check target system is completely different from the system that is the target of the check by the check items stored in the DB. Appropriate check items may not be extracted. Further, if the check items registered in the DB are ambiguous, the extracted check items are also ambiguous.

  Although the technique of Patent Document 2 can identify the part of speech of the omitted component, it cannot identify the meaning of the omitted word itself, so the omitted part is complemented to improve the ambiguity of the Japanese sentence. It is not possible.

  The technique of Patent Document 3 supplements a phrase that will be input based on a combination of phrases in a character string that has been input in the past. Not suitable.

  An object of the present invention is to provide a technique for creating a check rule capable of uniquely interpreting a check list for checking that there is no problem in work related to an information system.

  A check rule creation device according to an aspect of the present invention is a check rule creation device for creating a check rule for checking an information system as a check target and checking that there is no defect in work related to the information system. For each check item in an existing check list that includes a description in natural language for checking that there is no defect in the work related to the information system, extract a predetermined word by morphological analysis of the description included in the check item A pre-processing unit that creates a shaping checklist that lists the words, a natural language conversion information storage unit that stores natural language conversion information that associates a matter related to an information system and a natural language word indicating the information, Compare the words in the shaping checklist with the words in the natural language conversion information And an ambiguity evaluating unit that extracts an ambiguous part as a check item of the information system in the shaping check list, an ambiguous part complementing unit that estimates the content of the ambiguous part based on the natural language conversion information, And a check rule conversion unit that creates the check rule by adding the estimated content to the extracted ambiguous part.

  According to one aspect of the present invention, it is possible to create a check rule that can be uniquely interpreted for a check list for checking that there is no problem in work related to an information system.

It is a functional lineblock diagram of a check rule creation device by an embodiment. It is a functional block diagram of the check list creation apparatus shown in FIG. It is a flowchart for demonstrating operation | movement of the check list production apparatus shown in FIG. FIG. 3 is a diagram illustrating an example of a reference source check list illustrated in FIG. 2. It is a figure which shows an example of the morphological analysis result in the check list pre-processing part shown in FIG. It is a figure which shows an example of the analysis result of the geometric information in the check list pre-processing part shown in FIG. It is a figure which shows the specific example of the system configuration information which is one of the information accumulate | stored in the natural language conversion information storage part shown in FIG. It is a figure which shows the specific example of the file structure information which is one of the information accumulate | stored in the natural language conversion information storage part shown in FIG. It is a figure which shows the specific example of the metrics information which is one of the information accumulate | stored in the natural language conversion information storage part shown in FIG. It is a figure which shows the specific example of value and operator information which is one of the information accumulate | stored in the natural language conversion information storage part shown in FIG. It is a figure for demonstrating the process in the ambiguity part complementation part shown in FIG. It is a figure for demonstrating the process in the ambiguity part complementation part shown in FIG. It is a figure which shows the specific example of the check rule format accumulate | stored in the check rule storage part shown in FIG. It is a figure for demonstrating the process in the synonym learning part shown in FIG. It is a figure for demonstrating the process in the check rule reliability learning part shown in FIG.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a functional configuration diagram of the check rule creation device according to the present embodiment.

  As shown in FIG. 1, the check rule creation apparatus according to the present embodiment includes a computer 1 that includes a memory 10, an HDD 11, a CPU 12, an input device 13, and a display device 14. The memory 10 includes a check list preprocessing unit 101, an ambiguity evaluation unit 102, an ambiguity location complementation unit 103, a check rule conversion unit 104, a synonym learning unit 105, a check rule reliability learning unit 106, A natural language conversion information update unit 107 and a check rule output unit 108 are provided. In other words, the processing program stored in the HDD 11 is expanded on the memory 10, and the CPU 12 performs calculations to realize each function. The HDD 11 includes a shaping checklist storage unit 111, a natural language conversion information storage unit 112, a check work log storage unit 113, and a check rule storage unit 114. These indicate that various parts of the memory 10, that is, various data used for each function of the program are held in the HDD 11.

  Hereinafter, each function of the program developed on the memory 10 of the check list creation apparatus configured as described above and the operation of the check list creation apparatus based on the functions will be described.

  FIG. 2 is a functional block diagram of the check list creation apparatus shown in FIG. FIG. 3 is a flowchart for explaining the operation of the checklist creation apparatus shown in FIG.

  In the check list creation apparatus shown in FIG. 1, first, the check list preprocessing unit 101 executes the check preprocessing function with the reference source checklist 201 described in a natural language as an input.

  FIG. 4 is a diagram illustrating an example of the reference source check list 201 illustrated in FIG.

  The reference source checklist 201 shown in FIG. 2 is an existing checklist that includes a description in natural language for checking that the information system is a check target and that there is no problem in the work related to the information system. For example, as shown in FIG. 4, the reference source checklist 201 is described in an Excel file or the like in Japanese by an operator or the like, and includes a checklist title, a large item representing a check item attribute, a medium item, and a medium item. Upper attributes such as items may be included. Since the reference source checklist 201 performs linguistic and geometrical analysis by the checklist preprocessing unit 101, it does not specify a fixed format.

  First, the checklist preprocessing unit 101 performs morphological analysis on Japanese sentences existing in various places in the reference source checklist 201 as shown in FIG. 4, extracts predicates and body words as predetermined words, and separates them with commas. Convert to the enumerated form (step 1).

  FIG. 5 is a diagram illustrating an example of a morpheme analysis result in the checklist preprocessing unit 101 illustrated in FIG.

  As shown in FIG. 5, the checklist preprocessing unit 101 performs morphological analysis on the description included in the check item for each check item of the reference source checklist 201, and extracts a predicate and a body as a predetermined word, Convert to enumerated form separated by commas.

  Further, the check list preprocessing unit 101 analyzes the geometric information of the electronic file of the reference source check list 201 and relates to one check item in the reference source check list such as a title, a large item, and a medium item. Identify other parts and phrases.

  FIG. 6 is a diagram illustrating an example of an analysis result of geometric information in the checklist preprocessing unit 101 illustrated in FIG.

  As illustrated in FIG. 6, the checklist preprocessing unit 101 performs all the check items of the reference source checklist 201 in accordance with the hierarchical structure represented in the geometric structure, including the related words and phrases of other parts. List them separated by commas (step 2). Furthermore, when a negative word such as “... not” is related to a certain word, the checklist preprocessing unit 101 also describes the presence / absence of the negative word and the word related to the negative word. For example, in the example shown in FIG. 5, “not present” included in the check items for “capacity, check” for the large item and “file” for the middle item is targeted, and “exist” is a negative word. Corresponds to the word involved. In addition, if the Japanese sentence is omitted or banned and the presence or absence of a negative word cannot be determined, the fact that it is unknown is described. For example, in the reference source check list 201 shown in FIG. 4, the item “Approval column is blank” is set to “undecidable”.

  The checklist preprocessing unit 101 creates a shaping checklist that enumerates the words extracted from the descriptions included in the check items of the reference source checklist 201 by performing the above-described morphological analysis and geometric information analysis. And stored in the shaping checklist storage unit 111. Note that it is possible to perform only morphological analysis in the checklist preprocessing unit 101, but the creation of check rules can be strengthened by analyzing geometric information.

  Here, the natural language conversion information storage unit 112 shown in FIG. 1 stores natural language information in which items related to the information system are associated with words in the natural language indicating the information system. The information will be described in detail.

  FIG. 7 is a diagram showing a specific example of system configuration information which is one of the information accumulated in the natural language conversion information accumulation unit 112 shown in FIG.

  As shown in FIG. 7, the natural language conversion information accumulating unit 112 shown in FIG. 1 includes information system components (left side in the figure) for which a check rule is to be created, and each component of the information system. System configuration information that associates natural language words predicted to be represented (right side in the figure) is accumulated by an input from an operator or the like. In the figure, data and objects that cannot be used are indicated by shading.

  FIG. 8 is a diagram showing a specific example of file configuration information which is one of the information accumulated in the natural language conversion information accumulation unit 112 shown in FIG.

  As shown in FIG. 8, the natural language conversion information storage unit 112 shown in FIG. 1 includes file configuration information that enumerates files including words in a file that are used in an information system that is a target for creating a check rule. , Accumulated by the input of the operator.

  FIG. 9 is a diagram illustrating a specific example of metrics information which is one of the information accumulated in the natural language conversion information accumulation unit 112 illustrated in FIG.

  As shown in FIG. 9, in the natural language conversion information storage unit 112 shown in FIG. 1, what kind of trouble can occur in which component of the system or file when checking the system state or file state, Should be checked to determine whether there is a problem of the cause related to the component and the cause of the problem related to the file that are assumed in the work related to the system for which the check rule is to be created. Metric configuration information that associates metrics indicating matters with Japanese words that are predicted to represent metrics for detecting each malfunction of each component of the system, that is, words in natural language related to the metrics, It is accumulated by the input of the worker.

  FIG. 10 is a diagram illustrating a specific example of value / operator information that is one of pieces of information stored in the natural language conversion information storage unit 112 illustrated in FIG.

  As shown in FIG. 10, the natural language conversion information storage unit 112 shown in FIG. 1 includes a natural language word indicating the status of a metric that distinguishes the presence or absence of a defect and its metrics when checking the system status and file status. Value operator configuration information in which an operator and a value indicating the state are associated with each other is stored.

  The ambiguity evaluation unit 102 compares the words accumulated in the natural language conversion information accumulation unit 112 as described above with the words in the shaping checklist accumulated in the shaping checklist accumulation unit 111, thereby forming the shaping checklist. The ambiguous part is extracted as a check item of the information system in (Step 3).

  First, the ambiguity evaluation unit 102 acquires system configuration information and file configuration information from the natural language conversion information storage unit 112, selects one row of the shaping checklist stored in the shaping checklist 111, and includes it in that one row. It is searched whether there is a match between the word described in the system configuration information and the file configuration information. If the word selected from the formatting check list matches the word in the system configuration information, it is determined that this check item is an item for checking the status of the system component and is in the file configuration information. If it matches the word, it is determined that this check item is an item whose file status is to be checked.

  When it is determined that the check item is an item whose status is a system component, the ambiguity evaluation unit 102 matches the word included in the selected system configuration information with the word included in the system configuration information. Search for. The system configuration corresponding to the matched word is defined as “check target factor”.

  In addition, when it is determined that the check item is an item whose file status is to be checked, the ambiguity evaluation unit 102 searches the file to be checked and there is a word included in the selected line. For example, a search is performed from the left word of the shaping check list 111 shown in FIG. If the leftmost word is found, then the next word on the right is searched. This is repeated until no words are searched in the file to be checked. It is hierarchized in the search order, and this is defined as “check target factor” in this check item.

  Next, the ambiguity evaluation unit 102 acquires metric information from the natural language conversion information storage unit 112 and searches for a word in the selected line that matches the word described in the metric information. Metrics corresponding to the matched words are defined as “metric factors”.

  Further, the ambiguity evaluation unit 102 acquires the value / operator information from the natural language conversion information storage unit 112, and searches for a word in the selected line that matches the word described in the value / operator information. To do. The value / operator factor corresponding to the matched word is defined as “value / operator factor”.

  The ambiguity evaluation unit 102 stores a series of processes using the system configuration information, file configuration information, metrics information, and value / operator information acquired from the natural language conversion information storage unit 112 described above in the shaping check list 111. If all words in the formatted checklist are matched and all words match, “Reliability” is set to 100%, and “Checked Factor”, “Metric Factor”, “Value / Operator” The process of the ambiguity evaluation function related to the selected one line is finished with each value of “factor”.

  On the other hand, if there is no word associated with the natural language conversion information stored in the natural language conversion information storage unit 112 in the shaping checklist 111, the ambiguity evaluation unit 102 sets the part corresponding to the word as an “ambiguous part”. Judgment, and in addition to “check target factor”, “metric factor” and “value / operator factor”, the word has the same value and the value is determined, and processing in the ambiguity evaluation function for the selected one line Finish.

  Thus, an ambiguous part can be easily extracted by a simple process of associating words.

  Next, the ambiguous part complementation unit 103 supplements the ambiguous part of the shaping checklist line whose “reliability” determined by the ambiguity evaluation part 102 is not 100% by estimating the specific value of the factor ( Step 4).

  FIG. 11 is a diagram for explaining processing in the ambiguity portion complementation unit 103 shown in FIG.

  In the example shown in FIG. 11, the “check target factor” is determined only up to “CPU” of the “operation monitoring server” for the components (left side in the figure) of the information system for which the check rule is to be created. Of the five servers N011, N012, N013, S011, and S012 of the operation monitoring server, it is not possible to determine which server's CPU should be checked, and that part is extracted as an “ambiguous place” (in the figure). Top right).

In this case, "the operation monitoring server" N011, N012, N013, S011, 5 C 1 + 5 C 2 + 5 C 3 + 5 C 4 + 5 C for ₅ kinds-taking of S012 can be considered. Therefore, the ambiguous part complementation unit 103 sets all the “check target factors” to N011, N012, N013, S011, and S012, creates five check items, and creates the five check items “metric factor”, “value / All of the “operator factors” are the same as the initial values, and the processing ends with a reliability of 23% that each server can take (lower right in the figure).

  FIG. 12 is a diagram for explaining the processing in the ambiguity part complementation unit 103 shown in FIG. 1. In the ambiguity part complementation unit 103, it is determined whether or not there is a negative word in the word related to “operator factor”. An example of processing for which it was not possible to determine whether “=” or “≠” because it could not be determined.

  In the ambiguous part complementation unit 103, the word related to the “operator factor” is determined by the fact that the Japanese sentence is abbreviated or stopped, and the presence or absence of a negative word cannot be determined and the negative word cannot be determined. When it is not possible to determine whether “=” or “≠” for (in the figure, the left side), the ambiguous part complementation unit 103 determines that the factor is an ambiguous part (upper right in the figure).

  The ambiguous part complementation unit 103 searches the shaping checklist in which the ambiguous part is determined to determine whether or not there is a negative word in other items, and determines “=” and “≠” at a certain ratio of the negative word. . In the example shown in FIG. 12, since all the other nine items of the shaping check list have no negated word, the probability of “=” is high, and 9/10 is not a negated word. Then, the process is terminated with 90% reliability (lower right in the figure).

  Next, in the check rule conversion unit 104, the check item that is 100% reliable in the ambiguity evaluation unit 102 and the ambiguous part determined by the ambiguity evaluation unit 102 are estimated by the ambiguity part complementation unit 103. Check items that have been improved by adding the check items are combined into a single check rule format, and a check rule is created and stored in the check rule storage unit 114 (step 5).

  FIG. 13 is a diagram showing a specific example of the check rule format stored in the check rule storage unit 114 shown in FIG.

  As shown in FIG. 13, by the series of processes described above, the description contents are formulated for each of the four factors of the check target, metrics, value, and operator, and the check contents are uniquely identified by the machine or the operator. Can be interpreted.

  In the check rule output unit 108, the check rules stored in the check rule storage unit 104 are displayed and output on the display device 14 of FIG. 1, so that an operator or the like can view the check rules (step 6). .

  As described above, a check to check that there are no problems in the work related to the information system by using the existing check list and creating an information system check rule while suppressing ambiguity in the existing check list A check rule that can be uniquely interpreted for a list can be created.

  The functions and processes of the synonym learning unit 105, the check rule reliability learning unit 106, and the natural language conversion information update unit 107 illustrated in FIG. 1 will be described below.

  In the synonym learning unit 105 shown in FIG. 1, a check work log in which the check work actually performed and details of the results are written is acquired from the check work log storage unit 113, and from the shaping check list storage unit 111. When all the created checklists created so far are acquired and there are a plurality of words in the check check log that indicate the event for one event in the check work log, the word is defined as a synonym.

  FIG. 14 is a diagram for explaining processing in the synonym learning unit 105 shown in FIG.

  For example, as shown in the upper diagram of FIG. 14, in the shaping checklist acquired from the shaping checklist storage unit 111, words different from “usage rate”, “used percentage”, and “usage percentage” are arranged. As shown in the lower diagram of FIG. 14, in the check work log acquired from the check work log storage unit 113, when a program for measuring the resource usage rate is activated for all of the check items, the synonym learning unit 105 defines "usage rate", "used percentage", and "usage rate" as synonyms.

  In this way, the synonym learning unit 105 compares the shaping check list stored in the shaping check list storage unit 111 with the check work log of the information system stored in the check work log storage unit 113, and the shaping check list. For words that are different from each other, words that have the same check operation are defined as synonyms.

  Further, the check rule reliability learning unit 106 shown in FIG. 1 acquires the shaping check list from the shaping check list storage unit 111 and details of the check work and results actually performed from the check work log storage unit 113. The check work log in which is written is acquired, and the rule of the check work with respect to the check rule is learned.

  FIG. 15 is a diagram for explaining processing in the check rule reliability learning unit 106 shown in FIG.

  For example, as shown in the left diagram of FIG. 15, in the check rules stored in the shaping checklist storage unit 111, it is unclear which server should be checked when checking the usage rate of the memory and CPU. On the other hand, as shown in the right diagram of FIG. 15, if the check work logs accumulated in the check work log accumulation unit 113 all indicate that only the active server is checked, a check rule In the reliability learning unit 106, when checking the resource usage rate, the law learning that the active server should be performed is learned.

  In this way, the check rule reliability learning unit 106 compares the check rule stored in the shaping check list storage unit 111 with the check work log of the information system stored in the check work log storage unit 113 to check the check rule. On the basis of the correspondence status between the information system and the check work log of the information system, the rule of the check work for the check rule is learned.

  Further, the natural language conversion information updating unit 107 shown in FIG. 1 receives the synonym defined by the synonym learning unit 105 and the event learned by the check rule reliability learning unit 106, and uses the contents thereof. The information stored in the natural language conversion information storage unit 112 is updated. As a result, the synonym defined by the synonym learning unit 105 is reflected in the process in which the ambiguity evaluation unit 102 compares the words, and the law learned by the check rule reliability learning unit 106 is ambiguous. The sex part complementation unit 103 is reflected in the process of estimating the contents of the ambiguous part.

  As a result, as the number of check work logs stored in the check work log storage unit 113 increases, the accuracy of the information stored in the natural language conversion information storage unit 112 increases, and the ambiguity evaluation unit 102 and the ambiguity part complementation unit 103 increase. Can be performed with higher accuracy, and the reliability of the check rule is improved.

DESCRIPTION OF SYMBOLS 10 ... Memory, 11 ... HDD, 12 ... CPU, 13 ... Input device, 14 ... Display apparatus, 101 ... Check list pre-processing part, 102 ... Ambiguity evaluation part, 103 ... Ambiguous part complementation part, 104 ... Check rule conversion , 105 ... synonym learning part, 106 ... check rule reliability learning part, 107 ... natural language conversion information update part, 108 ... check rule output part, 111 ... shaping checklist storage part, 112 ... natural language conversion information storage part 113: Check work log storage unit 114: Check rule storage unit

Claims (7)

A check rule creation device for creating a check rule for checking an information system and checking that there is no defect in work related to the information system,
For each check item of an existing check list that includes a description in natural language for checking that there is no defect in the work related to the information system, the description included in the check item is subjected to a morphological analysis. A pre-processing unit that extracts words and creates a formatting checklist that lists the words;
A natural language conversion information storage unit that stores natural language conversion information that associates a matter related to an information system with a natural language word indicating the information system;
An ambiguity evaluation unit that extracts an ambiguous part as a check item of the information system in the shaping checklist by comparing the word of the shaping checklist and the word of the natural language conversion information;
An ambiguous part complementing unit that estimates the contents of the ambiguous part based on the natural language conversion information,
A check rule conversion unit that creates the check rule by adding the estimated content to the extracted ambiguous part;
A check rule creation device having Comparing the shaping checklist with a check work log of the information system, for different words included in the shaping checklist, defining a word having the same check work as a synonym and defining the synonym as the ambiguity The evaluation unit further includes a synonym learning unit that is reflected in the process of comparing words.
The check rule creation device according to claim 1. The check rule is compared with the check work log of the information system, and based on the correspondence status between the check rule and the check work log of the information system, the check rule of the check rule is learned, and the check The ambiguous part complementing unit further includes a check rule reliability learning unit that reflects the process of estimating the content of the ambiguous part.
The check rule creation device according to claim 1. The preprocessing unit enumerates words for each check item in the shaping checklist according to a hierarchical structure represented in a geometric structure of the existing checklist.
The check rule creation device according to claim 1. The ambiguity evaluation unit associates a word in the shaping checklist with a word in the natural language conversion information, and if there is no word to be associated with a word in the natural language conversion information in the shaping checklist, the word The part corresponding to
The check rule creation device according to claim 1. Problems assumed in the work related to the information system include problems caused by causes related to the components of the information system, and problems caused by causes related to files used in the information system,
The natural language conversion information is a file that lists system configuration information that associates system configuration elements of the information system with words representing the system configuration elements, and files that are used in the information system and include words in the files. Metric structure in which configuration information is associated with a metric for determining whether there is a cause defect related to the component and a cause defect related to the file, which are assumed in the work related to the information system, and a word related to the metric Information, value operator configuration information in which a word indicating the state of the metric that distinguishes the presence or absence of the defect, and an operator and a value indicating the state of the metric are associated with each other, and
The ambiguity evaluation unit includes the words listed in the check items of the shaping checklist, words included in the system configuration information, words included in a file whose file name is indicated in the file configuration information, Based on the above, it is determined whether the check item is an item for checking the status of the system component or an item for checking the status of the file, and the check item is an item for checking the status of the system component A metric for determining the check item is specified based on whether the check item is an item to be checked, the word listed in the check item, and the word included in the metric configuration information And based on the words listed in the check item and the words included in the value operator configuration information, Identify combinations of operators and a value indicating the state of the metrics to distinguish whether the defect as determined by serial check items,
The check item word, the word included in the system configuration information or the file configuration information, the word related to the specified metric in the metric configuration information, and the specified operation in the value operator configuration information Identifying an ambiguous part of the shaping checklist based on a word associated with a combination of a child and a value;
The check rule creation device according to claim 1. A check rule creation method for creating a check rule for checking an information system and checking that there is no defect in work related to the information system,
For each check item of an existing check list that includes a description in natural language for checking that there is no defect in the work related to the information system, the description included in the check item is subjected to a morphological analysis. Extract words, create a formatting checklist that lists the words,
Accumulate natural language conversion information that associates information system related items with natural language words that represent them,
By comparing the word of the shaping checklist and the word of the natural language conversion information, to extract the ambiguous part as a check item of the information system in the shaping checklist,
Estimating the content of the ambiguous part based on the natural language conversion information,
Adding the estimated content to the extracted ambiguous part to create the check rule;
Check rule creation method that the computer executes.

Images