Past fault information search device

Abstract

To provide a past fault information retrieval device that can acquire information according to a purpose.SOLUTION: A past fault information retrieval device includes: a past fault information structuring unit that extracts a plurality of characteristic elements representing structure from a past fault information group and generates a hierarchically structured past fault information element group; a past fault information element sorting unit that compares the plurality of elements with a plurality of keyword groups, which are collections of keywords, and generates a plurality of related element groups sorted so as to assign semantically closest keyword groups; a past fault information database generation unit that adds reference information to each of the plurality of related element groups to create a plurality of related reference groups; and a past fault information retrieval unit that retrieves an input retrieval target for the plurality of related reference groups and outputs a retrieval result.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a past failure information search device for utilizing failure information that occurs in product development. [Background technology]

[0002] In order to reduce the cost of software development for products, there is a need to shorten the lead time for software development. In recent years, software development has become large-scale due to the accumulation of various information created in the past. In order to shorten the lead time, it is important to efficiently obtain reference information from the vast amount of past information when various new investigation tasks arise.

[0003] Among the information accumulated during software development, past failure information, which is a record of software failures that occurred in the past, is particularly noteworthy. Past failure information is considered to be highly effective because it contains details based on phenomena that actually occurred.

[0004] There are various situations in which past failure information can be utilized, such as when you want to check failure information from past occurrences of a similar phenomenon when investigating the cause of a new failure, when you are developing a system that reuses design elements whose details are difficult to grasp and when you want to check failure information that will allow you to understand the details and the scope of the impact when correcting them, and when you want to check failure information to understand what kind of failures the design elements you are trying to reuse have been involved in in the past.

[0005] However, outage information is usually written in natural language and does not necessarily follow a strict format. Also, because it is written in a situation where an outage is urgent, the text is not well thought out and the content that should be written is often written in a different place than where it should be. Therefore, even if you search for outage information in its original form, it is likely that it will be difficult to find the reference information you need.

[0006] Prior art for improving the searchability of large amounts of natural language information includes, for example, a technique disclosed in Patent Document 1, which uses techniques such as pattern matching to subdivide sentences and represent them as a hierarchical structure in order to make it easier to understand the structure of natural language sentences. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent No. 6232736 Summary of the Invention [Problem to be solved by the invention]

[0008] By using the technology disclosed in Patent Document 1 to perform a search on text that has been subdivided and hierarchically structured, it becomes easier to focus on the hit elements and the elements above and below them in the tree, making it easier to reach the information you need, which is thought to reduce the lead time for utilizing information.

[0009] However, as mentioned above, when it comes to fault information, the content that should be written is often written in a different place than where it should be written, and there was a problem that subdivision and hierarchical structuring alone could not address the issue.

[0010] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a past fault information search device that can acquire information according to a purpose. [Means for solving the problem]

[0011] A past trouble information search device according to the present disclosure includes: a past trouble information structuring unit that extracts a plurality of characteristic elements that represent a structure from a past trouble information group that accumulates trouble information that occurred in software in the past and is written in a natural language, and generates a hierarchically structured past trouble information element group; a past trouble information element organizing unit that compares the plurality of elements included in the past trouble information element group with a plurality of keyword groups that are a group of keywords prepared in advance, classifies and organizes the elements so as to assign a keyword group that is closest in meaning, and generates a plurality of related element groups that are groups of organized past trouble information; a past trouble information database generation unit that adds reference information to each of the plurality of related element groups to generate a plurality of related reference groups; and a past trouble information search unit that searches the plurality of related reference groups for an input search target and outputs search results, wherein the past trouble information search unit has a plurality of search units suited to a search purpose, and the plurality of search units are provided so as to correspond respectively to the plurality of related reference groups generated by the past trouble information database generation unit, and each of the plurality of search units searches the related reference group corresponding to the search target. [Effects of the Invention]

[0012] According to the past fault information search device of the present disclosure, multiple elements of hierarchically structured past fault information are classified according to the specific meanings of the fault information, and searches are performed for each classification according to the search purpose, thereby making it possible to efficiently obtain information according to the purpose. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a block diagram showing a configuration of a past fault information search device according to a first embodiment. [Figure 2] FIG. 10 is a diagram illustrating an example of a past failure information group. [Figure 3] FIG. 10 is a diagram illustrating an example of a past failure information element tree group. [Figure 4] FIG. 10 is a diagram illustrating an example of fault-related keyword information. [Figure 5]10 is a flowchart illustrating an example of processing by a past failure information element organizing unit. [Figure 6] 10 is a conceptual diagram illustrating an example of processing by a past failure information element organizing unit. FIG. [Figure 7] FIG. 10 is a diagram illustrating an example of a group of organized past failure information. [Figure 8] 10 is a flowchart illustrating an example of processing by a past failure information database generating unit. [Figure 9] FIG. 10 is a conceptual diagram illustrating an example of processing by a past failure information database generating unit. [Figure 10] FIG. 10 is a diagram illustrating an example of a past failure information database. [Figure 11] FIG. 10 is a diagram illustrating an example of a search target. [Figure 12] FIG. 10 is a diagram illustrating an example of a similar failure information group. [Figure 13] FIG. 10 is a diagram illustrating an example of a detailed investigation method group. [Figure 14] FIG. 10 is a diagram illustrating an example of a group of quality influence factors. [Figure 15] FIG. 10 is a diagram illustrating an example of a design intention group. [Figure 16] FIG. 10 is a diagram showing an example of a search flow when a search is performed using a general search method. [Figure 17] FIG. 3 is a diagram showing an example of a search flow when a search is performed by the past fault information search device of the first embodiment. [Figure 18] FIG. 10 is a diagram showing an example of a search flow when a search is performed using a general search method. [Figure 19] FIG. 3 is a diagram showing an example of a search flow when a search is performed by the past fault information search device of the first embodiment. [Figure 20] 1 is a diagram illustrating a hardware configuration for realizing a past fault information search device according to a first embodiment. [Figure 21] 1 is a diagram illustrating a hardware configuration for realizing a past fault information search device according to a first embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] <First Embodiment> <Device configuration> 1 is a block diagram showing the configuration of a past problem information search device 1 according to a first embodiment of the present disclosure. As shown in FIG. 1, the past problem information search device 1 includes a past problem information structuring unit 3, a past problem information element organizing unit 10, a past problem information database generating unit 16, and a past problem information search unit 23, and searches for a search target 22 in a past problem information group 2 input from the outside based on problem-related keyword information 5 input from the outside, and generates and outputs a search result 28.

[0015] The past failure information group 2 is a collection of natural language documents that describe in some format the occurrence of software failures in a certain product, the investigation of the cause, the identified cause, and countermeasures such as temporary measures, permanent measures, measures to prevent recurrence, and horizontal deployment. An example 201 of the past failure information group 2 is shown in Figure 2.

[0016] In an example 201 shown in Figure 2, the information distinguished by the fault information ID: 1234, which is the fault identification symbol, includes natural language descriptions such as (1) how the fault occurred, (1-1) the situation in which the fault occurred, [State before the occurrence] "··· was XX", [Operation that caused the fault] "○○ was XX", (1-2) fault phenomenon "A phenomenon occurred in which △△ became ☆", and (2) investigation of the fault.

[0017] The past problem information structuring unit 3 generates a past problem information element tree group 4 by extracting characteristic elements that represent the structure of the past problem information group 2. For example, a line that starts with a symbol such as "(1), (2), (3)...", "(1-1), (1-2), (1-3)...", or "[ " is regarded as a title line and one element, and subsequent lines are regarded as child elements until the same type of symbol appears or the sentence ends, which can be realized by repeating this process. An example 301 of the past problem information element tree group 4 is shown in FIG. 3.

[0018] In the example 301 shown in Figure 3, the tree of information identified by the fault information ID: 1234 is hierarchically structured and listed in a tree-like manner, with the symbol (1) indicating "how the fault occurred" and the symbols (1-1) indicating "the circumstances under which the fault occurred," "the state before the fault occurred," and "the operation that caused the fault."

[0019] The fault-related keyword information 5 has information about each keyword that is expected to be included in the past fault information group 2 and the likelihood of meanings similar to that keyword, and is composed of a phenomenon-related keyword group 6, an investigation-related keyword group 7, a cause-related keyword group 8, and a countermeasure-related keyword group 9. An example 401 of the fault-related keyword information 5 is shown in FIG.

[0020] In the example 401 shown in Figure 4, it is assumed that the target natural language sentence is classified into one of four categories: "phenomenon," "investigation," "cause," and "countermeasure," and "phenomenon-related keyword group," "investigation-related keyword group," "cause-related keyword group," and "countermeasure-related keyword group" are listed.

[0021] The phenomenon-related keyword group 6 has one or more pairs of keywords related to the phenomenon that occurred as a failure and a value that indicates the likelihood of that relationship. In Figure 4, the "phenomenon-related keyword group" contains the keywords "phenomenon," "event," and [circumstances], and is assigned "10pt (points)," "10pt," and [8pt] as the increments of the likelihood value for classification, respectively.

[0022] The investigation-related keyword group 7 has one or more pairs of keywords related to investigations to identify the cause of the problem and values ​​that represent the likelihood of that association. In Figure 4, the "investigation-related keyword group" contains the keywords "investigation," "narrowing down," and "isolation," and is assigned "10pt," "9pt," and [8pt] as the increments of the likelihood values ​​for the classifications, respectively.

[0023] The group of cause-related keywords 8 contains one or more pairs of keywords related to the cause of the failure that was discovered as a result of the investigation, and a value that indicates the likelihood of that relationship. In Figure 4, the "group of cause-related keywords" contains the keywords "cause," "factor," and "mechanism," and is assigned "10pt," "10pt," and [8pt] as the increments of the likelihood value for each classification.

[0024] Countermeasure-related keyword group 9 includes one or more pairs of keywords related to countermeasures, such as temporary measures for the cause of the failure, permanent measures, measures to prevent recurrence, and horizontal deployment, and values ​​that indicate the likelihood of the relationship. In Figure 4, the "Countermeasure-related keyword group" includes the keywords "countermeasure," "measure," and "measure to prevent recurrence," and is assigned "10pt," "9pt," and [6pt] as the increments of the likelihood value for the classification, respectively.

[0025] The "value representing the likelihood of the relevance" is defined in advance by the user of the past fault information search device 1. For example, if the keyword "event" is included in the target natural language sentence, the user assumes that the likelihood of classification as a "phenomenon" increases, and sets the value representing this increase to "10 points."

[0026] The user can determine keywords by checking a collection of past outage information and selecting keywords that are likely to be effective for classification based on the appearance of various keywords in the collection, or by excluding keywords that are likely to be noise.

[0027] The past fault information element organizing unit 10 generates an organized past fault information group 11 by classifying the meaning of each element of the past fault information element tree group 4 based on the fault-related keyword information 5. The organized past fault information group 11 is made up of a phenomenon-related element group 12, an investigation-related element group 13, a cause-related element group 14, and a countermeasure-related element group 15.

[0028] <Operation> 5 is a flowchart of the processing of the past fault information element organizing unit 10. Also, a conceptual diagram for explaining an example of processing in the past fault information element organizing unit 10 is shown in FIG.

[0029] 5 and 6, the processing of the past fault information element organizing unit 10 will be described below. When the processing starts, first, each element is acquired from each piece of fault information in the past fault information element tree group 4 (step S1).

[0030] Next, all "XX related keyword groups" are acquired from the fault-related keyword information 5 (step S2).

[0031] Next, all "keywords" are acquired from the acquired "XX related keyword group" (step S3).

[0032] Next, the elements of the past failure information element tree group 4 acquired in step S1 are compared with the "keywords" acquired in step S3 to confirm a match (step S4).

[0033] If there is a match (Yes), points are added (Step S5). If there is no match (No), the process proceeds to Step S6.

[0034] The above processing of steps S1 to S5 will be conceptually explained using Fig. 6. Fig. 6 shows a case where "(1) (1-2) A phenomenon occurs in which a fault event △△ becomes a star" is selected as an example of each element acquired from the fault information of the past fault information element tree group 4.

[0035] In response, a match with "Event 10pt" is checked as a "keyword" obtained from the "group of phenomenon-related keywords," and if a match is confirmed, 10pt is added. "Match" is defined as "closest in meaning."

[0036] Taking the example 401 of the fault-related keyword information 5 shown in Fig. 4 as an example, points are added to each related keyword group using a score calculation table 510 as shown in Fig. 6. In the score calculation table 510, initially, the "phenomenon-related keyword group," "investigation-related keyword group," "cause-related keyword group," and "countermeasure-related keyword group" are all "0 points," but when a score of "10 points" corresponding to the "phenomenon" is added in step S5, "10 points" is added to the "phenomenon-related keyword group" in the score calculation table 510.

[0037] Returning to the explanation of the flowchart in Figure 5, after adding the points in step S5, it is checked whether all "keywords" obtained in step S3 have been processed (step S6), and if all "keywords" have been processed (if Yes), the process proceeds to step S7, and if there are any unprocessed "keywords", the process from step S4 onwards is repeated.

[0038] In step S7, it is checked whether all of the "XX related keyword groups" obtained in step S2 have been processed. If all of the "XX related keyword groups" have been processed (if Yes), the process proceeds to step S8. If there are any unprocessed "XX related keyword groups," the process of step S4 is repeated.

[0039] As a result of processing steps S6 to S7 above, when all "XX-related keyword groups" have been processed, the score calculation table 510 will be as shown in FIG. 6, with "50 points" added to the "phenomenon-related keyword group," "20 points" added to the "investigation-related keyword group," "10 points" added to the "cause-related keyword group," and the "countermeasure-related keyword group" remaining at "0 points" without any additions.

[0040] Returning to the explanation of the flowchart in Figure 5, in step S8, the group of related keywords that will give the highest score for each element is identified, and an element is assigned to be added as a related element to the group of related elements corresponding to the identified group of related keywords in the organized past fault information group 11.

[0041] The element allocation process in step S8 will be conceptually explained using Fig. 6. As shown in the final score calculation table 510 in Fig. 6, if the "phenomenon-related keyword group" has the highest score, the element "(1)(1-2) Fault event △△ becomes a star" is added to the phenomenon-related element group 12 in the organized past fault information group 11 by element allocation. In this way, by adopting the method of adding a numerical value according to the number of matches, element allocation can be performed relatively easily.

[0042] Returning to the explanation of the flowchart in FIG. 5, it is checked whether the processing of steps S2 to S8 has been executed for each element acquired from each piece of fault information in the past fault information element tree group 4 (step S9). If the processing has been executed for all elements (if Yes), the processing of the past fault information element organizing unit 10 is terminated, and if there are any unprocessed elements, the processing from step S4 onwards is repeated.

[0043] FIG. 7 shows an example 601 of the organized past fault information group 11 thus created by the past fault information element organizing unit 10.

[0044] In an example 601 shown in FIG. 7, the past fault information identified by fault information ID: 1234 includes a "group of elements related to the phenomenon," a "group of elements related to investigation," a "group of elements related to cause," and a "group of elements related to countermeasures." The "group of elements related to the phenomenon" includes the element "(1)(1-1)[] The state before the occurrence was XXX" and the element "(1)(1-1)[] The operation that caused the fault was XXX."

[0045] The "Survey-related elements group" contains the element "(2)(2-1)[] Narrowing down occurrence elements occurs only when □□ is present."

[0046] The "group of causal factors" contains the element "(4)(4-1)[] Occurrence mechanism: Because XX occurs at the timing of ●●..."

[0047] The "Countermeasure-related element group" contains the element "(6)(6-1)[] Correction response Add waiting process to process foo of module hoge."

[0048] Returning now to the explanation of Fig. 1, the past fault information database generator 16 generates the past fault information database 17 by adding reference information to each element group of the organized past fault information group 11.

[0049] A flowchart of the processing of the past fault information database generating unit 16 is shown in Fig. 8. A conceptual diagram for explaining an example of the processing of the past fault information database generating unit 16 is shown in Fig. 9.

[0050] The processing of the past fault information database generating unit 16 will be described below with reference to Figures 8 and 9. When the processing starts, first, each piece of fault information is acquired from the organized past fault information group 11 (step S11).

[0051] Next, each "XX-related element group" is obtained from the fault information obtained in step S11 (step S12). Examples of the "XX-related element group" include a "phenomenon-related element group," a "investigation-related element group," a "cause-related element group," and a "countermeasure-related element group."

[0052] Next, each element is obtained from the "XX related element group" obtained in step S12 (step S13).

[0053] Next, reference information for referencing the element acquired in step S12 is generated (step S14).

[0054] The processing of steps S11 to S14 above will be conceptually explained using Fig. 9. Fig. 9 shows a case where a "phenomenon-related element group" is acquired from the fault information, and an element "(1)(1-2) A phenomenon occurs in which a fault event △△ becomes a star" is acquired from the "phenomenon-related element group."

[0055] To reference this element, reference information for "ID: 1234" is generated and added to the information for the element "(1)(1-2) Failure event: A phenomenon occurs in which △△ becomes a star," thereby generating a "phenomenon-related reference group."

[0056] Returning to the explanation of the flowchart in Figure 8, after generating the reference information in step S14, it is checked whether all of the elements acquired in step S13 have been processed (step S15), and if all elements have been processed (Yes), the process proceeds to step S16, and if there are any unprocessed elements, the process from step S14 onwards is repeated.

[0057] In step S16, it is checked whether all of the "XX related element groups" obtained in step S12 have been processed. If all of the "XX related element groups" have been processed (if Yes), the process proceeds to step S7. If there are any unprocessed "XX related element groups," the process of step S14 is repeated.

[0058] In step S17, it is confirmed whether all of the fault information acquired in step S11 has been processed, and if all of the fault information has been processed (if Yes), the processing of the past fault information database generation unit 16 is terminated, and if there is unprocessed fault information, the processing of step S14 is repeated.

[0059] FIG. 10 shows an example 801 of the past fault information database 17 thus created by the past fault information database generating unit 16.

[0060] 10, a "phenomenon-related reference group," a "survey-related reference group," a "cause-related reference group," and a "countermeasure-related reference group" are listed. Note that the reference information for each reference group is added as "ID: 1234" to the beginning of each element shown in FIG. 7.

[0061] Returning now to the explanation of Fig. 1, the past fault information search unit 23 searches the past fault information database 17 for the search target 22 and outputs a search result 28. The past fault information database 17 is made up of a phenomenon-related reference group 18, an investigation-related reference group 19, a cause-related reference group 20, and a countermeasure-related reference group 21.

[0062] The past fault information search unit 23 is composed of a phenomenon-related search unit 24, an investigation-related search unit 25, a cause-related search unit 26, and a countermeasure-related search unit 27, which are respectively associated with the phenomenon-related reference group 18, the investigation-related reference group 19, the cause-related reference group 20, and the countermeasure-related reference group 21 of the past fault information database 17.

[0063] The past failure information search unit 23 has an algorithm for searching natural language sentences. As the algorithm, known algorithms such as morphological analysis and N-Gram, which divide natural language sentences into character strings, and Doc2Vec, which makes sentences evaluable as numerical values, can be used, and there is no need to develop an algorithm, which can reduce costs.

[0064] The search target 22 may be a new fault phenomenon that indicates a fault phenomenon that has newly occurred during development, or a reused design element that indicates information about a design that is being reused in a reused design.

[0065] A reused design is a design that reuses components of software created in the past, and the information about the original design is information about the design of the components being reused, such as specifications, design documents, and parts of the source code.

[0066] An example 901 of the search target 22 is shown in Fig. 11. In the example 901 shown in Fig. 11, "△△ has changed to ☆" is entered as a "new problem phenomenon," and "XX processing of ●● function" is entered as a "reused design element."

[0067] The phenomenon-related search unit 24 mainly inputs a new fault phenomenon as a search target 22, and searches the phenomenon-related reference group 18 to extract a similar fault information group 29, which is a collection of similar fault information having an occurrence phenomenon similar to the input new fault phenomenon. An example 1001 of the similar fault information group 29 is shown in FIG.

[0068] 12 shows an example 1001 in which, when the new fault phenomenon is "△△ changed to ☆", "a phenomenon occurred in which △△ became ☆" is searched from the phenomenon-related reference group 18, and "ID: 1234 (1) (1-2) fault event a phenomenon occurred in which △△ became ☆" is written as similar fault information. In this way, using the new fault phenomenon as the search target 22 and searching with the phenomenon-related search unit 24 is useful when analyzing the new fault phenomenon.

[0069] The investigation related search unit 25 mainly inputs reused design elements as search targets 22 and searches the investigation related reference group 19 to extract a detailed investigation method group 30, which is a collection of detailed investigation methods that are records of detailed investigations of design elements that are similar to the input reused design elements. The investigation related search unit 25 is useful, for example, when it is difficult to understand the content and structure of specifications, design documents, or source code, and you want to search for records of past investigations of those elements. An example 1101 of the detailed investigation method group 30 is shown in Figure 13.

[0070] In the example 1101 shown in Figure 13, when the reused design element is "XX processing of ●● function," "XX is used for XX processing in ●● function" is searched from the investigation-related reference group 19, and the detailed investigation method is described as "ID: 1234 (2) (2-1) Investigation content details XX is used for XX processing in ●● function."

[0071] The cause relation search unit 26 mainly inputs reused design elements as search targets 22 and searches the cause relation reference group 20 to extract a quality impact element group 31, which is a collection of quality impact elements that are records showing the impact that design elements similar to the input reused design elements have had on other design elements. The cause relation search unit 26 is useful, for example, when considering modifications to specifications, design documents, or source code, in order to investigate the degree of impact, and therefore to find out to what extent a particular element has affected other elements in the past. An example 1201 of the quality impact element group 31 is shown in Figure 14.

[0072] In the example 1201 shown in Figure 14, when the reused design element is "XX processing of ●● function," "XX stopped due to a delay in XX processing in ●● function" is searched from the cause-related reference group 20, and an example is shown in which "ID: 1234 (4) (4-2) Mechanism of failure occurrence XX stopped due to a delay in XX processing in ●● function" is written as a quality impact element.

[0073] The countermeasure related search unit 27 mainly inputs reused design elements as search targets 22 and searches the countermeasure related reference group 21 to extract a design intention group 32, which is a collection of design intentions that are records showing what kind of faults caused past modifications to design elements similar to the input reused design elements. The countermeasure related search unit 27 is useful, for example, when you know the contents and structure of the specifications, design documents, and source code, but cannot understand the design intention, and you want to search for fault information related to that design intention. An example 1301 of the design intention group 32 is shown in Figure 15.

[0074] In the example 1301 shown in Figure 15, when the reused design element is "XX processing of ●● function," "For this reason, add redundant processing to XX processing in ●● function" is searched from the countermeasure related reference group 21, and the design intent is described as "ID: 1234 (7) (7-1) Correction point For this reason, add redundant processing to XX processing in ●● function."

[0075] <Effects> In order to explain the accuracy of searching for fault information in the past fault information search device 1 of embodiment 1, examples of the search flow when searching using a general search method and when searching using the past fault information search device 1 are shown in Figures 16 and 17.

[0076] A search flow 1401 shown in FIG. 16 explains a search method of selecting the largest number of consecutive matching characters as a general search method, and shows a method of directly searching the past failure information group 2.

[0077] FIG. 16 shows an example in which, when searching for a new fault phenomenon 1402 with the search target "△△ changed to ☆," past fault information 1403 having a record of a similar fault phenomenon occurring and past fault information 1404 having a similar wording recorded as the cause of the fault occurrence are extracted as search results.

[0078] That is, the past fault information 1403 is recorded as past fault information identified by fault information ID: 1234, including "(1) How the fault occurred" and "(1-2) Fault phenomenon: A phenomenon occurred in which △△ became a ☆," and five characters match the new fault phenomenon 1402, "△△ changed to a ☆."

[0079] On the other hand, the past fault information 1404 is recorded as past fault information identified by fault information ID: 4321, and includes "(4) Cause of the fault" and "(4-2) Details of the cause of occurrence: Due to the change of △△ to ☆...", which matches nine characters with the new fault phenomenon 1402, "△△ changed to ☆".

[0080] As a result, a search result 1405 of "△△ changed to ☆" is obtained, and past fault information 1404 of fault information ID: 4321 in which similar wording is recorded as the cause of the fault occurrence is extracted.

[0081] In this way, if the similarity of sentences is determined only by the degree of match of wording, past fault information 1403 that has a record of a similar fault phenomenon occurring may not be extracted as a search result, and past fault information 1404 that should not be extracted may end up as a search result 1405.

[0082] A search flow 1402 shown in FIG. 17 shows a method in which the past trouble information search unit 23 searches the past trouble information database 17 for the search target 22 .

[0083] FIG. 17 shows an example in which, when searching for a new fault phenomenon 1502 with the search target "△△ changed to ☆", past fault information 1503 having a record of a similar fault phenomenon occurring is extracted as a search result, but past fault information 1504 having a similar wording recorded as the cause of the fault occurrence is not extracted as a search result.

[0084] That is, FIG. 17 shows that by performing a search on the phenomenon-related reference group 1505 of the past fault information database 17, “ID: 1234 (1) (1-2) Fault event A phenomenon occurred in which △△ becomes a ☆” of the past fault information 1503 included in the phenomenon-related reference group 1505 is extracted as a search result 1507.

[0085] This is because the phenomenon-related search unit 24 is intentionally selected and used in the search by the past fault information search unit 23, and the phenomenon-related search unit 24 searches only the phenomenon-related reference group 18, i.e., the phenomenon-related element group 12 referenced therefrom.

[0086] On the other hand, "ID: 4321 (4) (4-2) Details of the cause of occurrence Due to the change of △△ to ☆..." in the past fault information 1504 included in the cause related reference group 1506 is not searched for and is not extracted as a search result.

[0087] In this way, according to the past trouble information search device 1 of the first embodiment, the past trouble information 1503 that should be extracted as the search result 1507 is output, and it is clear that the search for trouble information is accurate.

[0088] <Example of using reused design elements> The search flow explained using FIGS. 16 and 17 shows the case where a new fault phenomenon is the search target, but similar results can be obtained when a reused design element is the search target.

[0089] FIG. 18 shows an example of a search flow when a search is performed using a general search method, and FIG. 19 shows an example of a search flow when a search is performed using the past fault information search device 1.

[0090] Figure 18 shows an example in which, when searching for a reused design element 1602 with "XX processing of ●● function" as the search target, past fault information 1603 containing records of modifications to the reused design element and past fault information 1604 containing records of details of the reused design element are extracted as search results.

[0091] That is, the past fault information 1603 is past fault information identified by fault information ID: 4567, and records "(7) How the fault occurred" and "(7-2) Correction points: Add redundant processing to XX of ●● function...", which matches seven characters with the reused design element 1602, "XX processing of ●● function".

[0092] On the other hand, the past fault information 1604 is recorded as past fault information identified by fault information ID: 2345, and includes "(2) Fault investigation" and "(2-1) Investigation content details: XX is used for XX processing of ●● function...", which matches nine characters with the reused design element 1602, "XX processing of ●● function".

[0093] As a result, a search result 1605 of "XX processing of XX function" is obtained, and past fault information 1604 of fault information ID: 2345 having a record of details of the reused design element is extracted.

[0094] In this way, if the similarity of sentences is judged solely by the degree of wording agreement, past problem information 1603, which contains records of modifications to the reused design element, may not be extracted as a search result, and past problem information 1604, which merely records details of the reused design element, may end up as a search result 1605.

[0095] A search flow 1701 shown in FIG. 19 shows a method in which the past trouble information search unit 23 searches the past trouble information database 17 for the search target 22 .

[0096] Figure 19 shows an example in which, when searching for a reused design element 1702 with "XX processing of ●● function" as the search target, past fault information 1703, which has records of modifications to the reused design element, is extracted as a search result, but past fault information 1704, which records details of the reused design element, is not extracted as a search result.

[0097] That is, FIG. 19 shows that by performing a search on the countermeasure-related reference group 1705 of the past fault information database 17, “ID: 4567 (7) (7-1) Correction location: Added redundant processing to XX of ●● function” of past fault information 1703 included in the countermeasure-related reference group 1705 is extracted as a search result 1707.

[0098] This is because the countermeasure related search unit 27 is intentionally selected and used in the search by the past fault information search unit 23, and the countermeasure related search unit 27 searches only the countermeasure related reference group 21, i.e., the countermeasure related element group 15 referenced therefrom.

[0099] On the other hand, "ID: 2345 (2) (2-1) Investigation content details: XX is used for XX processing of ●● function" in the past failure information 1704 included in the investigation-related reference group 1706 is not searched for and is not extracted as a search result.

[0100] In this way, according to the past trouble information search device 1 of the first embodiment, the past trouble information 1703 that should be extracted as the search result 1707 is output, and it is clear that the search for trouble information is accurate.

[0101] As described above, according to the past fault information search device 1 of the first embodiment, each element obtained by subdividing and hierarchically structuring past fault information is classified according to the specific meaning of the fault information, and a search according to the purpose is performed for each classification, thereby making it possible to efficiently acquire information according to the purpose.

[0102] For example, by searching for elements related to a phenomenon, it is possible to examine past cases of similar failures to help investigate the cause and consider countermeasures for a recent failure, thereby improving the efficiency of failure response work.

[0103] In addition, by searching for elements related to the investigation, it is possible to look up past cases where detailed investigations have been conducted on elements that are difficult to grasp in detail, thereby making the investigation process more efficient.

[0104] In addition, by searching for elements related to the cause, it is possible to investigate the quality of a certain element and its impact on other elements, thereby improving the efficiency of risk analysis of design elements, reuse decisions, and correction estimates.

[0105] In addition, by searching for elements related to countermeasures, it is possible to investigate the design intent of a certain element and the impact on other elements when it is modified, making it possible to prevent degradation, etc.

[0106] <Hardware configuration of the past fault information search device> Each component of the past fault information search device 1 can be configured using a computer and is realized by the computer executing a program. That is, the past fault information search device 1 is realized by, for example, a processing circuit 1000 shown in Fig. 20. A processor such as a CPU (Central Processing Unit) or a DSP (Digital Signal Processor) is applied to the processing circuit 1000, and the function of each part is realized by executing a program stored in a storage device.

[0107] Dedicated hardware may be applied to the processing circuit 1000. When the processing circuit 1000 is dedicated hardware, the processing circuit 1000 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof.

[0108] In the past fault information search device 1, the functions of the components may be realized by individual processing circuits, or these functions may be realized together by one processing circuit.

[0109] 21 shows a hardware configuration in the case where the processing circuit 1000 is configured using a processor. In this case, the functions of each unit of the past fault information search device 1 are realized by a combination of software, etc. (software, firmware, or software and firmware). The software, etc. is written as a program and stored in the memory 1200. The processor 1100 functioning as the processing circuit 1000 realizes the functions of each unit by reading and executing the program stored in the memory 1200 (storage device). In other words, it can be said that this program causes a computer to execute the procedure and method of operation of the components of the past fault information search device 1.

[0110] Here, memory 1200 may be, for example, a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), HDD (Hard Disk Drive), magnetic disk, flexible disk, optical disk, compact disk, mini disk, DVD (Digital Versatile Disc) and its drive device, or any storage medium to be used in the future.

[0111] The above has described a configuration in which the functions of each component of the past fault information search device 1 are realized by either hardware or software, etc. However, the present invention is not limited to this, and the configuration may be such that some of the components of the past fault information search device 1 are realized by dedicated hardware and other components are realized by software, etc. For example, it is possible to realize the functions of some of the components by the processing circuit 1000 as dedicated hardware, and to realize the functions of other components by the processing circuit 1000 as the processor 1100 reading and executing a program stored in the memory 1200.

[0112] It should be noted that the present disclosure allows modifications and omissions to be made to the embodiments as appropriate within the scope of the disclosure. [Explanation of symbols]

[0113] 1 Past fault information search device, 2 Past fault information group, 3 Past fault information structuring unit, 4 Past fault information element tree group, 5 Fault-related keyword group, 10 Organized past fault information group, 11 Organized past fault information group, 16 Past fault information database generation unit, 17 Past fault information database, 22 Search target, 23 Past fault information search unit.

Claims

1. a past fault information structuring unit that extracts a plurality of characteristic elements that represent a structure from a past fault information group that accumulates fault information written in a natural language that has occurred in the software in the past, and generates a hierarchically structured past fault information element group; a past trouble information element organizing unit that compares the plurality of elements included in the past trouble information element group with a plurality of keyword groups that are a group of keywords prepared in advance, classifies and organizes the elements so as to assign a keyword group that is closest in meaning to the elements, and generates a plurality of related element groups that are a group of organized past trouble information; a past failure information database generating unit that generates a plurality of related reference groups by adding reference information to each of the plurality of related element groups; a past failure information search unit that searches the plurality of related reference groups for an input search target and outputs a search result, The past failure information search unit It has multiple search sections tailored to the search purpose, The plurality of search units include: a past fault information search device provided so as to correspond to each of the plurality of related reference groups generated by the past fault information database generation unit, and each of the plurality of search units searches the corresponding related reference group for the search target.

2. The past failure information search unit A search algorithm for the natural language text is provided, The algorithm is:

2. The past fault information search device according to claim 1, further comprising an algorithm for dividing the sentence in the natural language into an array of character strings, and an algorithm for enabling the sentence to be evaluated as a numerical value.

3. The plurality of keyword groups are a phenomenon-related keyword group having one or more pairs of keywords related to a phenomenon occurring as a failure and values ​​indicating the likelihood of the relevance; a group of investigation-related keywords including one or more pairs of keywords related to an investigation to identify the cause of the failure and values ​​representing the likelihood of the relevance; a cause-related keyword group having one or more pairs of keywords related to the cause of the fault identified as a result of the investigation and values ​​indicating the likelihood of the relevance; a group of keywords related to measures including at least a temporary measure, a permanent measure, a measure to prevent recurrence, and a lateral deployment for the cause of the failure, and a group of keywords related to measures having one or more pairs of values ​​indicating the likelihood of the relevance; The past failure information element organizing unit 2. The past fault information search device according to claim 1, wherein, if a match is found in the comparison between the plurality of elements and the plurality of keyword groups, the semantically closest keyword group is assigned by adding values ​​representing the likelihood of the relevance of each of the plurality of keyword groups.

4. The search target is a new fault phenomenon that represents a fault occurrence phenomenon that has newly occurred in the development of the software; The past failure information search unit is one of the plurality of search units, 2. The past fault information search device according to claim 1, further comprising a search unit for extracting a group of similar fault information, which is a collection of similar fault information having a phenomenon similar to the new fault phenomenon, as the search result for the new fault phenomenon.

5. The search target is a reused design element representing information about a design to be reused in the reused design of the software development; The past failure information search unit is one of the plurality of search units, 2. The past failure information search device according to claim 1, further comprising a search unit that extracts, as a search result for the reused design element, a group of detailed investigation methods that are a collection of detailed investigation methods that are records of detailed investigations performed on design elements that are close to the reused design element.

6. The search target is a reused design element representing information about a design to be reused in the reused design of the software development; The past failure information search unit is one of the plurality of search units, 2. The past failure information search device according to claim 1, further comprising a search unit that extracts, as a search result for the reused design element, a group of quality affecting elements that is a collection of quality affecting elements that are records that represent the influence that design elements close to the reused design element have had on other design elements.

7. The search target is a reused design element representing information about a design to be reused in the reused design of the software development; The past failure information search unit is one of the plurality of search units, 2. The past fault information retrieval device according to claim 1, further comprising a retrieval unit that extracts, as a search result for the reused design element, a group of design intentions that is a collection of design intentions that are records indicating what faults have caused design elements similar to the reused design element to be corrected in the past.

Images