JP7011527B2 - Defect countermeasure support system - Google Patents

Description

The present invention relates to a technique for assisting in taking measures to reduce defects in software.

In order to reduce the defects of software, it is useful to take measures such as establishing an appropriate check procedure and education system. However, effective measures differ depending on the type of defect, so in order to take effective measures, it is necessary to identify the type of defect to be reduced.

The following Patent Document 1 describes a technique for searching a manual describing a coping method for a system failure. The document states, "Reduce the effort of manual search without presetting. The monitoring server 10 refers to the history information including the failure information regarding the failure of the monitored device and the classification of the failure to which the failure information corresponds, and uses the elements constituting the failure information included in the history information. Then, the procedure for converting each failure information into the data format of the vector representation is determined, and the failure information included in the history information is converted according to the procedure. The monitoring server 10 uses the failure information in the data format of the vector representation and the failure classification as learning data, and generates a judgment model to be applied to the judgment process for determining the failure classification. The monitoring server 10 converts the failure information generated when a failure occurs according to the procedure, and determines the classification of the failure from the failure information converted into the data format of the vector representation using the determination model. The monitoring server 10 refers to a manual including a fault classification of the monitored device and a countermeasure for the fault corresponding to the classification, and extracts a countermeasure from the manual corresponding to the fault classification. 』(See summary).

The following Patent Document 2 describes measures against system failures. The document states, "By promptly providing information on similar failure countermeasure cases by inquiries when a failure occurs, even clients with little experience in troubleshooting provide a method in which failure countermeasures can be taken by themselves in most cases from the failure countermeasure cases. The purpose is to do. With this as an issue, "Failure countermeasure information for classifying and countermeasures for faults (fault classification conditions, fault countermeasure cases for each classification condition, fault managers for each program, collection operation data for each program, etc.) is stored in the server. Then, for the failure inquired by the client, the failure countermeasure information such as the corresponding failure countermeasure example is extracted and displayed on the client. The client can take countermeasures by himself / herself by looking at the trouble countermeasure information. If you cannot take countermeasures by yourself, collect the fault countermeasure information and send the data to the person in charge of fault countermeasures for each program to request countermeasures. 』(See summary).

Japanese Unexamined Patent Publication No. 2013-254451 Japanese Unexamined Patent Publication No. 2002-297996

In the above-mentioned Patent Document 1, the correspondence relationship between the failure classification and the handling method manual is selected and stored in advance on the screen, and when a failure occurs, the handling method manual is specified according to the correspondence relationship. (See 0025, Fig. 4 of the same document). In such a procedure, the relationship between the failure classification and the coping method is considered to be determined almost uniquely. Therefore, when the coping method presented for one failure classification is not appropriate, another coping method is used. It is considered difficult to present. In addition, this document is a technique for dealing with individual failures, and it is considered that it is not supposed to specify measures to be used in improvement activities to reduce a certain defect classification.

In the above-mentioned Patent Document 2, a coping method is presented with reference to past failure countermeasure cases. For example, in 0038 and FIG. 3 of the same document, it is described that past response examples corresponding to the contents of the trouble inquiry are displayed. Similar to Patent Document 1, such a method is a technique for coping with individual failures, and it is considered that it is not supposed to specify measures to be used in improvement activities for reducing a certain defect classification. ..

The present invention has been made in view of the above problems, and an object of the present invention is to provide a technique for supporting taking useful measures in improvement activities for reducing defects in software.

The defect countermeasure support system according to the present invention has learned in advance the correspondence between the defect content of the software and the defect classification, and is a natural language character string that requires a document used to improve the defect classification. Upon receiving the request, the defect classification specified by the natural language character string is estimated by language analysis, and then the most accurate estimation result is returned.

According to the defect countermeasure support system according to the present invention, improvement activities can be effectively supported by presenting a document used for improving a designated defect classification. In addition, in order to identify as quickly as possible which defect classification the user is trying to improve, the defect classification explicitly or implicitly contained in the request by the natural language string is inferred and the guess is made. Present the result as a response to the request. Therefore, the user can determine whether or not the learner understands which defect classification is specified. This makes it possible to quickly identify the defect classification desired by the user.

It is a block diagram of the defect countermeasure support system 10 which concerns on Embodiment 1. FIG. This is a configuration example of the defect list 210. This is a configuration example of the classification result display screen 220 provided by the client terminal 200. This is an example of the total screen 230 provided by the client terminal 200. This is a configuration example of the document inquiry screen 240 provided by the client terminal 200. This is an example of a screen when the document inquiry screen 240 presents a document. This is an example of a screen when presenting a summary of a document. It is a continuation of FIG. 7A. This is an example of a screen when the defect classification cannot be specified.

<Embodiment 1>
FIG. 1 is a configuration diagram of a defect countermeasure support system 10 according to the first embodiment of the present invention. The defect countermeasure support system 10 is a system in which a user such as a project manager advises countermeasures to be taken in order to reduce defects in the software. The defect countermeasure support system 10 has a defect countermeasure support server 100 and a client terminal 200.

The defect countermeasure support server 100 includes a defect classification unit 110, a language analysis unit 120, a document acquisition unit 130, and an artificial intelligence 140 (learning device). The defect classification unit 110 receives a request from the client terminal 200 to classify the defect contents in the defect list 210 described later, and returns the classification result to the client terminal 200. An example of the request will be described later. The language analysis unit 120 analyzes the natural language character string included in the document request described later. An analysis example will be described later. The document acquisition unit 130 receives a request from the client terminal 200 requesting a document to be used in an activity for reducing a specific defect classification, and presents the corresponding document to the client terminal 200. Specific examples will be described later. The details of the artificial intelligence 140 will be described later.

FIG. 2 is a configuration example of the defect list 210. The user of the client terminal 200 records, for example, the defects obtained by testing the software developed so far in the project to which he / she belongs as the defect list 210. The defect list 210 has No211, a module 212, a developer affiliation 213, a developer 214, an event 215, a cause 216, and a correction content 217 as data fields.

No. 211 is a serial number of the record in the defect list 210. The module 212 is data (for example, a module name) that identifies a module of software that has a defect. The developer affiliation 213 and the developer 214 refer to the module as a developer and its affiliation (for example, an in-house department, a development company, etc.). Event 215 describes the content of the defect detected in the test (ie, the defect itself). Cause 216 describes the cause of the defect. The modification content 217 describes the work content performed to correct the defect.

FIG. 3 is a configuration example of the classification result display screen 220 provided by the client terminal 200. The user issues a request via the client terminal 200 to request the defect countermeasure support server 100 to classify the defects described in the defect list 210. The client terminal 200 transmits the defect list 210 together with the request to the defect countermeasure support server 100. The defect classification unit 110 classifies defects according to the request and responds with the classification result. The client terminal 200 displays the defect classification received from the defect countermeasure support server 100 on the classification result display screen 220. The processing procedure performed by the defect countermeasure support server 100 will be described later.

The classification result display screen 220 has defect classification 221, accuracy 222, evaluation 223, correction classification 224, and output button 225 as display items. Further, No. 221 to the modified content 217 described with reference to FIG. 2 are also displayed.

The defect classification 221 is the result of the defect countermeasure support server 100 classifying the defects (that is, the event 215). For example, a classification such as "initialization error" corresponds to this. The accuracy 222 is the accuracy as a result of the defect countermeasure support server 100 classifying the defects. The evaluation 223 is a mark indicating whether or not the accuracy 222 is equal to or higher than a predetermined threshold value (for example, 75%). The correction classification 224 is an interface for the user of the client terminal 200 to manually correct the defect classification 221, and presents, for example, a list of defect classifications by a pull-down selection box. When the user presses the output button 225, the client terminal 200 outputs the data displayed on the classification result display screen 220 to the storage device on the client terminal 200 in an appropriate format. If the defect classification 221 is corrected by the correction classification 224, the corrected content is output.

<Embodiment 1: Server processing procedure>
When the defect countermeasure support server 100 receives the classification request from the client terminal 200, the defect countermeasure support server 100 delivers each defect content described in the defect list 210 to the artificial intelligence 140. The artificial intelligence 140 learns in advance the correspondence between the defect content and the defect classification, and holds the result as the defect classification learning result 141. The artificial intelligence 140 classifies the received defect content into one of the defect classifications by using the defect classification learning result 141.

Event 215 does not necessarily describe the defect classification itself created in advance, and may be described by, for example, a natural language character string. The artificial intelligence 140 learns in advance the correspondence between the event 215 described by such an ambiguous expression and the defect classification by an appropriate learning method. Furthermore, even if the event 215 is the same, it may be regarded as a completely different defect content depending on the combination of the cause 216 and the correction content 217. Therefore, the artificial intelligence 140 can also learn the defect classification corresponding to the combination of the event 215 / cause 216 / correction content 217 in addition to the defect classification corresponding to the event 215.

As described above, the defect contents described in the defect list 210 are not uniquely classified for any defect classification, and there may be a plurality of candidates as to which defect classification they belong to. In the past, among such multiple classification candidates, the personnel of the defect countermeasure support team identified the content of the defect based on experience. Therefore, countermeasure support could only be provided within the resource range of the defect countermeasure support team. Therefore, in the first embodiment, the artificial intelligence 140 learns in advance the correspondence between the defect content and the defect classification, thereby automating the defect classification work.

However, the defect classification learning result 141 that the artificial intelligence 140 learns in advance does not describe the defect classification for all the defect contents. Therefore, the artificial intelligence 140 classifies the content of the defect and outputs the accuracy of the classification result. For example, the accuracy can be calculated based on the distance between the existing defect classification learning result 141 and the classification result. Other suitable methods may be used. The classification result display screen 220 displays the classification result output by the artificial intelligence 140 via the defect classification unit 110 and its accuracy as the defect classification 221 and the accuracy 222, respectively.

Even if the accuracy 222 is equal to or higher than the threshold value, it may be incorrect, so it is desirable that the user can correct the defect classification 221. For example, when the accuracy 222 is equal to or higher than the threshold value, the client terminal 200 displays a mark to that effect on the evaluation 223 and makes the correction classification 224 inoperable on the screen. When the user clicks the mark on the evaluation 223, the mark is turned off, and the defect classification after correction can be specified on the screen on the correction classification 224.

The correction candidate list presented by the correction classification 224 may be a specified one, and all the defect classifications described in the defect classification learning result 141 may be selected from the defect countermeasure support server 100 to the client terminal 200. It may be listed by notifying at that time.

FIG. 4 is an example of the aggregation screen 230 provided by the client terminal 200. The client terminal 200 can aggregate the classification results received from the defect countermeasure support server 100 and display the results as the aggregation screen 230. The tabulation screen 230 can tabulate the defect classification results in units such as, for example, for each defect classification / module 212 / developer affiliation 213 / developer 214. Other appropriate units may be classified.

The upper left of FIG. 4 is an aggregation result showing the ratio for each defect classification included in the defect list 210. The upper right of FIG. 4 is an aggregation result showing the ratio of the module 212 having the defect classification 1. The lower left of FIG. 4 is an aggregation result showing the ratio of developer affiliations 213 involved in defect classification 1. The lower right of FIG. 4 is an aggregation result showing the ratio of developers 214 involved in defect classification 1. By presenting such aggregated results, the user can grasp which defect classification the software tends to have, and can estimate the module, developer, etc. that are the cause of each defect classification. ..

The user grasps the tendency of defects in the software, for example, through the tabulation screen 230. The user issues a request to the defect countermeasure support server 100 via the client terminal 200 to present a document to be used in the improvement activity for reducing the defect. The defect countermeasure support server 100 identifies a document corresponding to the defect classification specified by the client terminal 200 using the countermeasure document learning result 142, and presents the document to the client terminal 200. The detailed operation of the defect countermeasure support server 100 at this time will be described in the second embodiment.

<Embodiment 1: Summary>
In the defect countermeasure support system 10 according to the first embodiment, the artificial intelligence 140 has learned in advance the correspondence relationship between the defect content and the defect classification as the defect classification learning result 141, and the defect classification unit 110 is from the client terminal 200. In response to the request of, the defect classification corresponding to the specified defect content and its accuracy are responded. As a result, the user can obtain the defect classification only by presenting the defect list 210 to the defect countermeasure support server 100. Furthermore, by presenting the accuracy of the classification result together, it is possible to add the judgment by the user himself / herself.

In the defect countermeasure support system 10 according to the first embodiment, the classification result display screen 220 presents a state in which the correction classification 224 cannot be selected for the defect classification 221 whose accuracy 222 is equal to or higher than the threshold value as an initial state, and evaluates 223. Click the mark above to change the state so that the correction classification 224 can be selected. As a result, it is possible to add the judgment by the user to the automatic classification result while reducing the burden on the user to manually classify the defects. Further, since the correction classification 224 cannot be selected in the initial state, it is possible to suppress an erroneous operation by the user.

In the defect countermeasure support system 10 according to the first embodiment, the aggregation screen 230 displays the aggregation result of the defect classification by the defect countermeasure support server 100. As a result, in addition to being able to grasp the tendency of defects in the software, it is possible to infer the factors that cause the defects in the software development process based on the aggregated results.

<Embodiment 2>
In the first embodiment, the function that the defect classification unit 110 mainly returns the defect classification result by the artificial intelligence 140 has been described. In the second embodiment of the present invention, specific operation examples of the language analysis unit 120 and the document acquisition unit 130 will be further described.

FIG. 5 is a configuration example of the document inquiry screen 240 provided by the client terminal 200. The client terminal 200 provides an interface for inquiring a document to be used in an improvement activity for a user to specify a defect classification and reduce the defect classification as a document inquiry screen 240. The document inquiry screen 240 can be provided as an interactive interface as illustrated in FIG.

By designating the defect classification on the document inquiry screen 240, the user issues a request to the defect countermeasure support server 100 so as to present a document to be used in the activity for reducing the defect classification. However, the user may not always know the name of the defect classification of the software, and may not clearly understand the improvement target although there is a vague improvement request. Therefore, the user can describe the request by using the natural language character string as illustrated by the request character string 241. The client terminal 200 transmits the request character string 241 to the defect countermeasure support server 100.

The language analysis unit 120 receives the request character string 241 from the client terminal 200 and analyzes the contents thereof. For example, the request character string 241 is decomposed into words, and a character string similar to the defect classification or the defect classification is extracted. The language analysis unit 120 estimates the character string contained in the request character string 241 that is closest to the defect classification included in the defect classification learning result 141, and returns the result to the client terminal 200. That is, the defect classification obtained by estimation with the highest estimation accuracy is returned.

The document inquiry screen 240 displays a confirmation window 242 asking the user whether or not the defect classification received from the language analysis unit 120 is the defect classification intended by the request character string 241. For example, when the defect classification estimated by the language analysis unit 120 is "initialization error", the confirmation window 242 as shown in FIG. 5 is displayed. The user selects "Yes" if the estimation result is correct, and selects "No" if it is not correct.

When the user selects "No", the client terminal 200 further issues a request to the defect countermeasure support server 100 so as to present the estimation result of the accuracy of the second rank or higher. The language analysis unit 120 returns the defect classification of the accuracy after the second rank obtained in the estimation performed earlier. The document inquiry screen 240 presents a list 243 of defect classifications received from the language analysis unit 120. Here, an example of presenting the defect classifications of the second to fourth ranks is shown.

As illustrated in FIG. 5, the result of estimating the defect classification specified by the request character string 241 is once presented for the following reasons. The defect classification specified by the user may not always match the defect classification held by the defect classification learning result 141. When the defect classification is specified by using a natural language character string such as the request character string 241, the tendency is further enhanced. Therefore, the defect countermeasure support system 10 first identifies the defect classification of the target for which the user intends to carry out the improvement activity. By carrying out the procedure as shown in FIG. 5, even when the defect classification is specified by an ambiguous expression, the user's wish can be determined at an early stage.

FIG. 6 is a screen example when the document inquiry screen 240 presents a document. When the user selects "Yes" in the confirmation window 242 or selects any of the defect classifications in Listing 243, the client terminal 200 is asked to present the documents to be used in the activity to reduce the selected defect classification. In addition, a request is issued again to the defect countermeasure support server 100.

The artificial intelligence 140 learns in advance the correspondence between the defect classification and the document that is effective to be used in the activity for reducing the defect classification, and holds the learning result as the countermeasure document learning result 142. ing. The document acquisition unit 130 identifies a document corresponding to the designated defect classification by using the countermeasure document learning result 142, and returns the specific result.

The document inquiry screen 240 presents the document received from the document acquisition unit 130 as a document list 244. It is not always necessary to present the content of the document itself. For example, the artificial intelligence 140 holds the title and the link destination of each document together in the countermeasure document learning result 142, and the document acquisition unit 130 can return the title and the link destination. The document list 244 may present its title and link. If multiple documents correspond, you can present the list. The same applies to FIG. 7B, which will be described later.

When there are a plurality of types of countermeasure documents, the document inquiry screen 240 may display the type confirmation window 245. The type confirmation window 245 is an interface for the user to select which document type is desired. As the document type, for example, the type of each document may be stored in the countermeasure document learning result 142 for each defect classification, and the document acquisition unit 130 may send this to the client terminal 200, or the specified document type. May be presented by the document inquiry screen 240 as an option. In FIG. 6, two document types, "educational material" and "check sheet", are illustrated.

FIG. 7A is an example of a screen for presenting a summary of a document. Depending on the document type, it may be useful for the user to present a summary before retrieving the document itself. For example, a standard document described by a spreadsheet is often described separately for each item, so it is relatively easy to present a summary for each item. Here, such an example will be described.

It is assumed that the user selects "Yes" on the confirmation window 242 described with reference to FIG. It is assumed that the user further selects the "check sheet" on the type confirmation window 245 in FIG. 7A. Therefore, the user is requesting to present a "check sheet" corresponding to the defect classification "initialization error".

When the document is described separately for each item, the artificial intelligence 140 learns the correspondence between the defect classification and the document, and stores the item described in the document in the countermeasure document learning result 142. can do. The document acquisition unit 130 returns the items corresponding to the specified defect classification and document type.

The document inquiry screen 240 displays the items received from the document acquisition unit 130 on the item list 246. It is not always necessary to display items obtained from a single document, and items contained in multiple documents corresponding to the same defect classification and document type may be displayed together. For example, check sheet 1 describes "deficiency in initialization" and "deficiency in initialization position" as check items related to defect classification "default initialization error", and check sheet 2 describes defect classification "initialization error". When "missing consideration of initial value" is described as a check item for, the item list 246 can be displayed as shown in FIG. 7A.

FIG. 7B is a continuation of FIG. 7A. When the user selects any item on the item list 246, the client terminal 200 requests a summary of the item from the defect countermeasure support server 100. Here, an example of requesting a summary of the check item "Insufficient initialization" is shown.

The artificial intelligence 140 can store a summary of the items in the countermeasure document learning result 142 together with the items described in the document. For example, the section title described in the document can be stored as an item name, and the contents of the section can be stored as a summary. It is not always necessary to store a summary of the content, and the content of the document itself may be a summary. When storing the summary, the summary created by the manual may be stored, or the summary generated by an appropriate method may be stored. The document acquisition unit 130 returns a summary corresponding to the specified defect classification / document type / item.

The document inquiry screen 240 displays the summary received from the document acquisition unit 130 on the summary 247. Summary 247 also provides a link to the document. If the content of the summary 247 matches the wishes, the user can obtain the document via the link.

FIG. 8 is an example of a screen when the defect classification cannot be specified. When the user selects "None" on the list 243, the defect classification unit 110 cannot specify the defect classification from the request character string 241. In this case, since it is necessary to change the request character string 241, the document inquiry screen 240 displays a message 248 prompting the user to that effect.

<Embodiment 2: Summary>
In the defect countermeasure support system 10 according to the second embodiment, the language analysis unit 120 analyzes the request character string 241 input by the user in natural language to determine the defect classification desired by the user to present the document. Estimate and return the most accurate estimation result. The document inquiry screen 240 presents the estimation result on the confirmation window 242. If the estimation result is incorrect, the estimation result of the second or higher rank can be presented on the list 243. As a result, the user can freely describe the request in natural language, and the defect classification desired by the user can be determined at an early stage, so that the user can input the request from the viewpoint of the request input format and operability. The convenience of can be ensured.

In the defect countermeasure support system 10 according to the second embodiment, the artificial intelligence 140 stores the title of the document together with the countermeasure document learning result 142, and the document acquisition unit 130 is the document corresponding to the defect classification specified by the user. Reply the title. The document inquiry screen 240 displays its title and a link to the document on the document list 244. As a result, the user can grasp the contents of the document corresponding to the specified defect classification in a list.

In the defect countermeasure support system 10 according to the second embodiment, the artificial intelligence 140 stores the items in the document and the summary thereof together with the countermeasure document learning result 142, and the document acquisition unit 130 classifies the defects specified by the user. Reply the item and summary in the corresponding document. The document inquiry screen 240 displays the item on the item list 246 and displays a summary 247 of the item specified by the user. As a result, the user can grasp the contents to some extent even before browsing the document corresponding to the specified defect classification.

<Embodiment 3>
In the above embodiment, it has been described that the client terminal 200 displays the screens of FIGS. 4 to 8, but these screens may be generated by the client terminal 200 by themselves or the GUI generated by the defect countermeasure support server 100. (Graphical User Interface) may be drawn by the client terminal 200. For example, when a GUI is configured as a Web application, the latter configuration is used. However, by using a script that operates on the Web browser, it is possible to change a part of the GUI only on the client terminal 200. For example, the case where a technique such as Dynamic HTML is used corresponds to this.

<About a modification of the present invention>
In the above embodiment, the defect classification unit 110, the language analysis unit 120, and the document acquisition unit 130 can be implemented by, for example, executing a program in which these functions are implemented by an arithmetic unit such as a CPU (Central Processing Unit). .. The same applies to the artificial intelligence 140. The artificial intelligence 140 and each learning result do not necessarily have to be mounted on the defect countermeasure support server 100, and the same function can be obtained by implementing the artificial intelligence 140 and each learning result on another artificial intelligence server and then communicating between the defect countermeasure support server 100 and the artificial intelligence server. May be implemented.

10: Defect countermeasure support system 100: Defect countermeasure support server 110: Defect classification unit 120: Language analysis unit 130: Document acquisition unit 140: Artificial intelligence 141: Defect classification learning result 142: Countermeasure document learning result 200: Client terminal 210: Defect list

Claims (12)

It is a defect countermeasure support system that supports taking measures to reduce the defects of software.
A learner that has learned in advance the first correspondence between the defect content and defect classification of the software.
A client terminal that issues a first request requesting the learner to classify the defects described in the defect list by using the defect list that lists the defect contents of the software.
Have,
The learner further learns in advance the second correspondence between the document used in the activity of improving the defect belonging to the defect classification and the defect classification.
The client terminal issues a second request to the learner, including a natural language string requesting that the document corresponding to the defect classification be presented.
The learner estimates the defect classification specified by the natural language character string by language analysis, and then outputs the most accurate estimation result as a response to the second request. Support system. The client terminal issues a third request to the learner requesting that the document corresponding to the estimation result of the defect classification received as a response to the second request be presented.
The learner is characterized in that one or more of the documents corresponding to the defect classification specified by the third request are presented in order from the one having the highest accuracy of the correspondence relationship as a response to the third request. The defect countermeasure support system according to claim 1. The client terminal issues a fourth request to the learner to notify that the estimation result of the defect classification received as a response to the second request is incorrect.
The learner is characterized in that one or more of the estimation results of the defect classification specified by the natural language character string are presented in order from the one with the second highest accuracy as a response to the fourth request. The defect countermeasure support system according to claim 1. When the client terminal receives an operation input specifying that the estimation result of the defect classification received as a response to the fourth request is incorrect, the client terminal outputs a message prompting to correct the expression of the natural language character string. 3. The defect countermeasure support system according to claim 3. The learner stores the title of the document in advance together with the second correspondence.
The defect countermeasure support system according to claim 2, wherein the learner presents the title of the document as a response to the third request. The learner stores in advance one or more summaries of the document including the description related to the defect classification in addition to the second correspondence.
The defect countermeasure support system according to claim 2, wherein the learner presents a list of summaries of the document in response to the third request. The client terminal issues a fifth request to the learner requesting that one of the list of summaries be specified and its contents be presented.
The defect countermeasure support system according to claim 6, wherein the learning device presents the contents of the summary designated by the fifth request as a response to the fifth request. As a response to the first request, the learner responds with the defect classification to which the defect described in the defect list belongs and its accuracy.
When the client terminal receives the response to the first request, the client terminal presents an accuracy mark suggesting on the screen for the defect classification whose accuracy is equal to or higher than the predetermined threshold value, and the accuracy is less than the predetermined threshold value. The defect countermeasure support system according to claim 1, wherein correction candidates are displayed on the screen for the defect classification. The client terminal presents a GUI in a state where a correction candidate cannot be selected on the screen for the defect classification whose accuracy is equal to or higher than a predetermined threshold value, and a switch for changing the GUI so that the correction candidate can be selected on the screen. Presented,
8. The client terminal is characterized in that it switches whether or not a correction candidate can be selected on the screen for the defect classification whose accuracy is equal to or higher than a predetermined threshold value according to an operation input input to the switch. Defect countermeasure support system described. The client terminal presents the switch as the accuracy mark on the screen.
When the accuracy mark indicates that the accuracy is equal to or higher than the predetermined threshold value, the GUI cannot select correction candidates on the screen for the defect classification having the accuracy equal to or higher than the predetermined threshold value.
When the accuracy mark indicates that the accuracy is less than the predetermined threshold value, the GUI is in a state where correction candidates can be selected on the screen for the defect classification having the accuracy equal to or higher than the predetermined threshold value. The defect countermeasure support system according to claim 9, wherein the defect countermeasure support system is characterized in that. The defect countermeasure according to claim 1, wherein when the client terminal receives a response to the first request, the defect classification to which the defect described in the defect list belongs is aggregated and the aggregated result is presented. Support system. The learner establishes the first correspondence in at least one unit for each defect classification, each module constituting the software, each developer who developed the software, or each organization to which the developer belongs. I have learned in advance,
When the client terminal receives the response to the first request, the defect classification to which the defect described in the defect list belongs is at least for each defect classification, each module, each developer, or each organization. The defect countermeasure support system according to claim 11, wherein the aggregation result is presented in any unit.

Images