JP2008209988A - Fmea sheet creation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technique enabling a person having no experience or knowledge to efficiently perform analysis a large quantity of case data and creation of a FMEA (failure mode and effects analysis)sheet. <P>SOLUTION: The FMEA sheet creation device comprises an item list storage means which stores an item list in which a plurality of items of FMEA sheet is defined together with the arrangement order thereof; a phrase extraction means which extracts phrases from a plurality of given case data; a semantic network generation means which generate a semantic network having each phrase as a node from a phrase group extracted by the phrase extraction means; a selection support means which displays the semantic network on a display device and makes a user select a plurality of phrases linked on the displayed semantic network; and a FMEA sheet generation means which generates a FMEA sheet having each phrase selected by the selection support means as the content of each item. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for analyzing and utilizing accumulated cases.

  When a defect occurs in a manufacturing process of a factory, it has been conventionally performed to investigate the cause of the occurrence based on the experience of the worker and knowledge about the product and to improve each time. A process FMEA (Failure Mode and Effects Analysis) is known as one of such improvement work methods.

  The process FMEA is a bottom-up improvement method for investigating the influence and cause corresponding to the generated defect (failure mode) when a defect is found in the manufacturing process. In the process FMEA, an FMEA sheet (also called analysis data, FMEA list, or simply FMEA) is prepared before the start of production. The FMEA sheet is a list in which processes, functions, phenomena (risks), effects, causes, countermeasures, and the like are associated with each expected failure occurrence pattern (failure mode).

  FIG. 15 shows an example of the FMEA sheet. This FMEA sheet includes processes, functions, risks, effects, and causes as items. In each row of the FMEA sheet, the value of each item corresponding to a certain failure mode is described. For example, in the first line, as the value of each item, the process: “PCB division”, function: “fix”, risk: “part crack (near V-cut)”, influence: “not function” , Cause: It is described as "stress such as sled".

  For example, when “PD contamination” occurs in the “PCB division” step, if it is desired to know the influence and the cause thereof, the FMEA sheet may be used in the following procedure. (1) First, a row corresponding to “PCB division” is searched from the column of “process”, (2) a row where “risk” corresponds to “dirty PD” is specified, and (3) specified Refer to the contents of "Impact" and "Cause" in the line. According to such a procedure, the fact that “PD contamination” causes “deterioration of detection function” and that “PD contamination” occurs due to “touching PD with bare hands” indicates that the user (process manager or work) Person) can easily know.

  FMEA sheets are generally created based on past defect countermeasures and complaint handling experiences and knowledge obtained from past cases (test reports, defect reports, complaint reports, past FMEA sheets, etc.). The Therefore, there is a problem that the effectiveness and usefulness of the FMEA sheet depends greatly on the experience, knowledge and ability of the creator. Also, there is a problem that it is extremely difficult for beginners with little experience and knowledge to create FMEA sheets.

  Therefore, Patent Document 1 proposes a system for supporting creation of an FMEA sheet. In this system, candidates for each item such as cause (occurrence condition), influence, countermeasure, and the like are displayed on the screen. Therefore, the creator can create an FMEA sheet by performing a selection operation according to the screen display.

If the FMEA sheet creation phase is considered as [1] collection of past cases, [2] case analysis (knowledge data), [3] FMEA sheet creation (documentation), the system Only the workload of the documenting phase [3] is reduced. However, it is not the documenting phase but the knowledge data conversion phase of [2] that the creator is most required for experience, knowledge, and ability (that is, in Patent Document 1, candidate data for each item. Failure mode / occurrence condition level file 22 by work action, influence evaluation table 24, countermeasure rank table 25 in FIG.
Etc.) Therefore, in the conventional system, the above-described problems related to the creation of the FMEA sheet are not essentially solved.

As a technique for performing semantic analysis of a large amount of data, a technique called a semantic network is known (see Patent Document 2).
JP 2005-182544 A JP-A-10-74210

  The present invention has been made in view of the above circumstances, and the purpose of the present invention is to perform analysis of a large amount of case data and creation of an FMEA sheet even for a person who does not have experience or knowledge. It is to provide the technology that makes it possible.

  In order to achieve the above object, the present invention adopts the following configuration.

  A first aspect of the present invention is an FMEA sheet creating apparatus for creating an FMEA sheet from case data in which cases are recorded in text, and a plurality of items of the FMEA sheet are defined together with their arrangement order. An item list storage means for storing an item list, a phrase extraction means for extracting a phrase from a plurality of given case data, and a semantic network having each phrase as a node from the phrase group extracted by the phrase extraction means Selected semantic network generation means, selection support means for displaying the semantic network on a display device, and allowing a user to select a plurality of words linked on the displayed semantic network, and the selection support means. FMEA sheet generation means for generating an FMEA sheet having each item as the content of each item. FMEA is a sheet creating device.

  According to this configuration, a large number of case data (the case data need only be raw data in which the case is recorded in text, and does not have to be in the state of knowledge data) are automatically analyzed, and words in the case data Is displayed in the form of a semantic network. This semantic network makes it easy to grasp the relationship between phrases and the frequency of occurrence (importance) of phrases. Then, an FMEA sheet can be created by an intuitive operation of selecting a linked phrase while looking at a semantic network. Therefore, if the present invention is used, the user can analyze the case and create the FMEA sheet easily and efficiently. The concept of “FMEA sheet” in the present invention includes a hard copy FMEA sheet, a screen-displayed FMEA sheet, and an FMEA sheet of electronic data (electronic document).

  It is preferable to provide a dictionary that stores a plurality of registered phrases together with corresponding FMEA sheet items. More preferably, the dictionary allows the user to add and delete registered words.

  As the number of case data increases, the number of extracted phrases increases, and the scale of the semantic network also increases. If so, there is a risk that the intuitive understanding and operation of the semantic network may be hindered. Therefore, the following ideas should be added when creating and displaying the semantic network.

It is preferable that the selection support unit changes the display mode of the node on the semantic network between a phrase that matches the registered phrase and a phrase that does not match the registered phrase. For example, the shape, color, size, etc. of the nodes may be varied. Moreover, you may hide only one side. Such display facilitates the distinction between words that match the registered words and words that do not. Typically, words that match the registered words should be conspicuous.

  The selection support means may change the display mode of the node of the phrase that matches the registered phrase according to the item of the corresponding FMEA sheet. For example, the shape, color, size, etc. of the node may be different for each item. Such a display makes it possible to know at a glance the correspondence between a word and an item in the FMEA sheet, which is useful when selecting a word or phrase to be adopted for the FMEA sheet.

  The selection support means is configured to select the two when the two linked phrases match the registered phrase, and the items of the FMEA sheet corresponding to each of the linked phrases are adjacent in the arrangement order of the item list. It is preferable that a link between two words is represented by a directed link indicating the arrangement order. Since this directed link serves as a guide that suggests the order in which words are selected, it is easier to select words to be used in the FMEA sheet.

  Preferably, the selection support means displays only words that match the registered words and that are linked in the arrangement order of the item list. This is because “a phrase that matches a registered phrase and is linked in the arrangement order of the item list” is likely to be an appropriate phrase as the contents of each item of the FMEA sheet. By displaying only such words and phrases, it becomes easier to understand the semantic network and select words.

  The selection support means may extract case data including all of one or more words in a selected state and display them on the display device as a word is selected or deselected on the semantic network. preferable. This function enables drill-up analysis and drill-down analysis of case data simultaneously with creation of FMEA sheets.

  In the case where the extracted case data includes information represented by numerical values, the selection support means preferably displays statistical information on the numerical values on the display device. Statistical information includes, for example, average, variance, and histogram. Such information serves as a reference for drill-up analysis and drill-down analysis of case data, and by extension, selection of words and phrases in the FMEA sheet.

  The present invention may be regarded as an FMEA sheet creating apparatus having at least a part of the above means, an FMEA sheet creating method including at least a part of the above processing, or an FMEA sheet creating program for realizing such a method. It can also be understood as a recording medium on which the program is recorded. Each of the above means and processes can be combined with each other as much as possible to constitute the present invention.

  The FMEA sheet creation method according to the second aspect of the present invention is an FMEA sheet creation method for creating an FMEA sheet from case data in which cases are recorded in text, wherein the computer includes a plurality of items of the FMEA sheet. A process for reading an item list defined along with the order of arrangement, a phrase extraction process for extracting a phrase from a plurality of given case data, and a phrase group extracted by the phrase extraction process, each phrase as a node Semantic network generation processing for generating a semantic network, selection support processing for displaying the semantic network on a display device, and for allowing a user to select a plurality of phrases linked on the displayed semantic network, and the selection support FM that generates an FMEA sheet containing each word selected in the process as the content of each item A FMEA sheet creation method of performing a A sheet generation process.

The FMEA sheet creation program according to the third aspect of the present invention is an FMEA sheet creation program for creating an FMEA sheet from case data in which cases are recorded in text, and the computer includes a plurality of items of the FMEA sheet. A process for reading an item list defined along with the order of arrangement, a phrase extraction process for extracting a phrase from a plurality of given case data, and a phrase group extracted by the phrase extraction process, each phrase as a node Semantic network generation processing for generating a semantic network, selection support processing for displaying the semantic network on a display device, and for allowing a user to select a plurality of phrases linked on the displayed semantic network, and the selection support An FMEA sheet containing each word selected in the process as the content of each item And FMEA sheet generating process for forming a FMEA sheet creation program for execution.

  According to the present invention, even a person who does not have experience or knowledge can efficiently analyze a large amount of case data and create an FMEA sheet.

  Exemplary embodiments of the present invention will be described in detail below with reference to the drawings.

<System configuration>
FIG. 1 is a block diagram illustrating a functional configuration of the FMEA sheet creation apparatus. This FMEA sheet creation apparatus is a system for creating an FMEA sheet from case data. More specifically, the FMEA sheet creation apparatus is a system for supporting analysis (conversion into knowledge data) of a large number of case data (raw data) and creation of an FMEA sheet.

  As shown in FIG. 1, the FMEA sheet creation apparatus includes a phrase registration function 1, an item list selection function 2, a case data input function 3, a morpheme analysis function 4, a co-occurrence frequency calculation function 5, a semantic network generation function 6, and FMEA analysis. A support function 7, an FMEA sheet generation function 8, a phrase dictionary 10, and an item list DB (database) 11 are provided. This FMEA sheet device is a general-purpose device including a CPU (Central Processing Unit), a main storage device (RAM, etc.), an auxiliary storage device (hard disk, etc.), an input device (keyboard, pointing device, etc.), a display device, a communication device, etc. Consists of computers. Each function described above is realized by a program stored in the auxiliary storage device being read into the main storage device and executed by the CPU. The phrase dictionary 10 and the item list DB 11 are stored in the auxiliary storage device.

<Phrase dictionary>
The phrase dictionary 10 is a dictionary (dictionary data) that stores a plurality of registered phrases together with corresponding FMEA sheet items. In the phrase dictionary 10, phrases frequently used as contents (values) of items in the FMEA sheet, phrases to be used, assumed phrases, and the like are registered. The phrase registered in the phrase dictionary 10 may be a word such as “transistor”, a group of words (phrase or clause) such as “no reaction”, or “an abnormal lamp is lit”. It may be a sentence.

  FIG. 2 shows an example of a phrase dictionary applied to the process FMEA in the semiconductor substrate manufacturing process. In the example of FIG. 2, “component”, “failure mode”, “influence”, “process”, “cause”, “measure”, and the like are used as items of the FMEA sheet. Then, “CPU”, “capacitor”, “transistor”, etc. are registered as terms corresponding to the item “parts”, and “chip standing”, “wetting failure” as terms corresponding to the process “defective mode”. , “Short” etc. are registered.

It is preferable that a large number of words / phrases are preset in the word / phrase dictionary 10 (in an initial state of the apparatus). However, it is difficult to cover all cases that occur in an actual manufacturing process with only a preset dictionary. Therefore, the FMEA sheet creation apparatus of this embodiment has a phrase registration function 1. This phrase registration function 1 adds a new phrase, edits a registered phrase,
It is a function that provides a user interface such as deleting registered words. For example, on the input screen as shown in FIG. 2, when the user inputs an item corresponding to a new phrase, the phrase and the item are associated with each other and registered in the phrase dictionary 10.

<Item list>
The item list DB 11 is a database (item list storage means) that stores an item list. The item list is data in which a plurality of items on the FMEA sheet are defined together with their arrangement order. The item list DB 11 can store one or more item lists.

  FIG. 3 shows “process → failure mode → cause → countermeasure” and “part → failure mode → cause → effect → detection means → countermeasure” as examples of the item list. The item list should also be preset in advance. The user can also add, edit, and delete item lists.

<Analysis of case data>
The case data analysis process in the FMEA sheet creation apparatus will be described with reference to the flowchart of FIG. This device uses a semantic network to analyze enormous case data.

  First, in step S10, the item list selection function 2 reads all item lists from the item list DB 11, displays them as a list, and prompts the user to select an item list. When the user selects the item list, the process proceeds to the next process. If there is no desired item list in the list, the user can create a new item list in this step.

  In step S11, the case data input function 3 reads a case data group to be analyzed. It is assumed that the storage location (folder, directory, or recording medium) of the case data to be analyzed is designated by the user or predetermined.

  Any data can be handled as case data as long as it is data in which the case is recorded in text. For example, FIG. 5 shows defect reports (documents 1 to 3) as an example of case data. In this way, the case data handled by this device does not need to be specially structured or converted to knowledge data (of course, it may be structured or written in accordance with predetermined rules). However, it may be raw data written in any format and free text.

  In step S12, the morphological analysis function 4 performs a morphological analysis on the text included in the case data and extracts a phrase. That is, in the present embodiment, the morphological analysis function 4 corresponds to the phrase extraction unit of the present invention. As conceptually shown in FIG. 6, the morphological analysis function 4 extracts a plurality of phrases from each case data (a set of extracted phrases is referred to as a “phrase set”). If there is a possibility that the number of phrases may become enormous if all the results of morphological analysis are adopted, it is preferable to select phrases to be extracted according to a certain standard. For example, it is preferable to preferentially extract words that are registered in the word dictionary 10 rather than words that are not registered. Once the word / phrase set has been created for all case data, proceed to the next process.

  In step S13, the co-occurrence frequency calculation function 5 obtains a co-occurrence frequency table based on the phrase set of each case data. “Co-occurrence” refers to appearing in the same document (example data), and “co-occurrence frequency” refers to the frequency of occurrence of co-occurrence in two words. It can be said that the combination of words with higher frequency of co-occurrence has a stronger relationship between words.

  FIG. 7 conceptually shows how to obtain the co-occurrence frequency table. That is, the co-occurrence frequency calculation function 5 (1) first counts the number of appearances (number of case data that appears) for each word, and (2) the number of co-occurrence ( The number of case data that appears at the same time) is counted. (3) Then, the co-occurrence frequency calculation function 5 calculates the co-occurrence frequency for each of the combinations of the two phrases by dividing the co-occurrence number by the appearance number. A co-occurrence frequency table is shown in (3) of FIG. In this embodiment, the processing procedure of obtaining the co-occurrence frequency table after extracting words from all case data is adopted, but the number of occurrences and the number of co-occurrence are counted in parallel with the word extraction. A processing procedure may be adopted.

  In step S14, the semantic network generation function 6 generates a semantic network (also referred to as a co-occurrence frequency network) based on the co-occurrence frequency table. The semantic network generated here has each word / phrase (each word / phrase extracted in step S12) included in the co-occurrence frequency table as a node, and links between words / phrases having a co-occurrence frequency equal to or higher than a predetermined threshold. It is a thing. That is, in this embodiment, the co-occurrence frequency calculation function 5 and the semantic network generation function 6 correspond to the semantic network generation means of the present invention.

  In step S15, the FMEA analysis support function 7 displays the semantic network on the display device. FIG. 8 shows a display example of a semantic network generated based on the co-occurrence frequency table of (3) of FIG. In this example, words having a co-occurrence frequency of 0.5 or more are connected by a link. Although FIG. 8 illustrates an undirected link, a directed link may be used. By viewing the analysis result of the case data in the form of a semantic network, the user can easily grasp the relationship between the phrases and the appearance frequency (importance) of the phrases.

  In the examples of FIGS. 5 to 8, the number of case data and the number of extracted words / phrases are very small for the sake of simplicity. However, since the number of case data can be several tens to several hundreds in a real scene, the number of extracted phrases becomes enormous, and the scale of the co-occurrence frequency table and semantic network that are generated also increases. If so, there is a risk that the intuitive understanding and operation of the semantic network may be hindered. Therefore, the FMEA analysis support function 7 provides various functions described below (these functions correspond to the selection support means of the present invention).

(A) Enlarged display function The enlarged display function is a function for enlarging and displaying a part of the semantic network. When this function is used, as shown in FIG. 9, the network around the word (node) selected by the user is enlarged and displayed. This makes it easy to focus on words that are closely related to the selected word.

  There are various ways of enlargement display. For example, only a selected node and a node whose distance from the node (the number of links constituting the shortest path) is a predetermined value (for example, two links) or less may be expanded. Alternatively, the display magnification may be simply increased.

(B) Highlight Display Function The highlight display function is a function for making a phrase (matching phrase) that matches a registered phrase in the phrase dictionary 10 stand out compared to a phrase (mismatching phrase) that does not match a registered phrase. The reason why the matched phrase is highlighted is that the phrase registered in the phrase dictionary 10 is more likely to be more appropriate as the content of the item in the FMEA sheet than the unregistered phrase.

The left side of FIG. 10 shows a display example when the highlighting function is off, and the right side of FIG. 10 shows a display example when the highlighting function is on. When the highlighting function is turned on, only the matching phrase nodes are displayed in a special shape, color, and size. And the shape and color differ according to the item of the FMEA sheet corresponding to a registered word. In the example of FIG. 10, “capacitor” and “transistor” as words corresponding to the item “component”, “chip standing defect” and “chip floating defect” as items corresponding to the item “defective mode”, and item “cause”. “Under solder” is shown as the corresponding word and “adjustment review” is shown as the word corresponding to the item “measure”. Furthermore, when two linked phrases are both matched phrases, and the corresponding items are adjacent in the order of the item list (selected in step S10), the two phrases Matched phrase links are represented by arrows (directed links). The direction of the arrow indicates the order of items defined in the item list.

  By highlighting as shown in FIG. 10, the correspondence between words and items and the arrangement order of items can be known at a glance, so that it is easy to select a word to be adopted for the FMEA sheet.

(C) Concealment function The concealment function is a function for concealing (not displaying or thinly displaying) other words in order to make only some words particularly conspicuous. The left side of FIG. 11 shows a display example when the concealment function is off, and the right side of FIG. 11 shows a display example when the concealment function is on. When the concealment function is turned on, the nodes and links of the mismatched phrase are concealed. In addition, even a link between matching phrases is concealed in the case of an undirected link such as between “capacitor” and “under solder”. Furthermore, even if it is a matching phrase, it is concealed if it does not have a directed link. That is, when the concealment function is on, only words that match the registered words and are linked in the order of the item list are displayed. There is a high possibility that a phrase corresponding to such a condition is an appropriate phrase as the contents of each item of the FMEA sheet.

  Furthermore, it is also preferable to display only words / phrases corresponding to some items and hide words / phrases corresponding to other items. For example, when the “capacitor” is selected in the state where the concealment function on the right side of FIG. 11 is turned on, only “transistor”, which is a word corresponding to the same item as the selected word, is displayed.

(D) FMEA sheet prototype function The FMEA sheet prototype function is a function for supporting the trial production of the FMEA sheet. As shown on the left side of FIG. 12, when the user sequentially selects words / phrases linked on the semantic network, the selected words / phrases are sequentially applied to the items defined in the item list. To go. The FMEA sheet being prototyped is displayed as shown on the right side of FIG. Of course, the selection can also be canceled. For example, when the selection of “adjustment review” is canceled, the item “measure” of the FMEA sheet being prototyped returns to the blank.

  A simple selection function may be provided so that a plurality of words can be easily selected. For example, when a user designates a word / phrase as a starting point, a series of word / phrase groups connected in order by a directed link from the word / phrase may be automatically selected. According to this function, the same result as selecting “chip lift failure” → “undersolder” → “adjustment review” can be obtained with only one action of selecting “chip lift failure”. Alternatively, when the user specifies two words, a start point and an end point, a series of words connected in order from the start point to the end point may be automatically selected. According to this function, it is the same as selecting “Transistor” → “Chip floating defect” → “Solder under-sizing” → “Adjustment review” in two actions: “Transistor” and “Adjustment review”. Results are obtained.

  In order to prevent words from being selected in an order different from the order of the item list, it is possible to select other than adjacent nodes, to select only in the same order as the direction of the directed link, It is also possible to make it impossible to select non-corresponding nodes.

(E) Case Data Display Function The case data display function is a display device that extracts case data including all of one or more selected words as words are selected or deselected in the prototype. It is a function to display. In the example of FIG. 13, case data search is executed using one or more words or phrases in the selected state as an AND condition, and the result of the search is displayed as a list. In other words, when the phrases are selected in order, the displayed case data is gradually narrowed down. Conversely, when the selection of words is canceled, the displayed case data is gradually increased. This function enables drill-up analysis and drill-down analysis of case data at the same time as the FMEA sheet is being prototyped.

(F) Statistical information display function When the case data extracted by the case data display function includes information represented by numerical values, the statistical information display function obtains statistical information of those numerical values and displays the information It is a function to display. In the example of FIG. 14, since the impact level is included in the case data, the average of the impact levels is calculated and displayed. Such information serves as a reference for drill-up analysis and drill-down analysis of case data, and by extension, selection of words and phrases in the FMEA sheet. For example, when there are a plurality of candidate words to be selected, it is preferable to select the word having the largest influence degree.

  As information represented by numerical values, in addition to the influence level, the occurrence frequency of the failure mode, the detection difficulty level, and the like are assumed. Statistical information includes variance, histogram, etc. in addition to average.

<Generation of FMEA sheet>
The FMEA sheet generation function 8 can generate an FMEA sheet based on the selected phrase using the FMEA analysis support function 7. The selected words and items are fitted with the same logic as in the trial production of the FMEA sheet. In addition, when the prototype FMEA sheet is recorded in the storage device, the FMEA sheet can be generated using the data.

  The FMEA sheet generation function 8 can output the FMEA sheet in the form of electronic data (electronic document), can be displayed on a display device, and can also be output to a printing device.

  According to the FMEA sheet creating apparatus of the present embodiment described above, a large number of case data (raw data) is automatically analyzed, and words in the case data are displayed in the form of a semantic network. This semantic network makes it easy to grasp the relationship between phrases and the frequency of occurrence (importance) of phrases. Then, an FMEA sheet can be created by an intuitive operation of selecting a linked phrase while looking at a semantic network. Therefore, if this apparatus is used, even a person who does not have experience and knowledge can easily and efficiently analyze a large number of cases and create FMEA sheets.

FIG. 1 is a block diagram illustrating a functional configuration of the FMEA sheet creation apparatus. FIG. 2 is a diagram illustrating an example of a phrase dictionary. FIG. 3 is a diagram illustrating an example of the item list. FIG. 4 is a flowchart showing the flow of the case data analysis process. FIG. 5 is a diagram illustrating an example of case data. FIG. 6 conceptually illustrates morphological analysis. FIG. 7 is a diagram conceptually showing how to obtain the co-occurrence frequency table. FIG. 8 is a diagram illustrating a display example of the semantic network. FIG. 9 is a diagram illustrating an operation example of the enlarged display function. FIG. 10 is a diagram illustrating an operation example of the highlighting function. FIG. 11 is a diagram illustrating an operation example of the concealment function. FIG. 12 is a diagram illustrating an operation example of the FMEA sheet prototype function. FIG. 13 is a diagram illustrating an operation example of the case data display function. FIG. 14 is a diagram illustrating an operation example of the statistical information display function. FIG. 15 is a diagram illustrating an example of an FMEA sheet.

Explanation of symbols

1 Phrase Registration Function 2 Item List Selection Function 3 Case Data Input Function 4 Morphological Analysis Function 5 Co-occurrence Frequency Calculation Function 6 Semantic Network Generation Function 7 FMEA Analysis Support Function 8 FMEA Sheet Generation Function 10 Phrase Dictionary 11 Item List DB

Claims (10)

An FMEA sheet creating apparatus for creating an FMEA sheet from case data in which cases are recorded in sentences,
Item list storage means for storing an item list in which a plurality of items of the FMEA sheet are defined together with their arrangement order;
A phrase extracting means for extracting a phrase from a plurality of given case data;
Semantic network generation means for generating a semantic network having each phrase as a node from the phrase group extracted by the phrase extraction means;
A selection support means for displaying the semantic network on a display device, and allowing a user to select a plurality of phrases linked on the displayed semantic network;
FMEA sheet generating means for generating an FMEA sheet having each item selected by the selection support means as the content of each item;
FMEA sheet creating apparatus. A dictionary further storing a plurality of registered words together with corresponding FMEA sheet items;
The FMEA sheet creating apparatus according to claim 1, wherein the selection support unit changes a display mode of the node on the semantic network between a phrase that matches the registered phrase and a phrase that does not match the registered phrase. The FMEA sheet creating apparatus according to claim 2, wherein the selection support unit changes a display mode of a node of a phrase that matches the registered phrase according to an item of the corresponding FMEA sheet. The selection support means is configured to select the two when the two linked phrases match the registered phrase, and the items of the FMEA sheet corresponding to each of the linked phrases are adjacent in the arrangement order of the item list. The FMEA sheet creating apparatus according to claim 2 or 3, wherein a link between two words is represented by a directed link indicating the arrangement order. The FMEA sheet creation apparatus according to any one of claims 2 to 4, wherein the selection support unit displays only words that match the registered words and that are linked in the arrangement order of the item list. The selection support means extracts case data including all of one or more selected words as the words are selected or deselected on the semantic network, and displays them on the display device. The FMEA sheet creation apparatus of any one of 1-5. 7. The FMEA sheet creating apparatus according to claim 6, wherein when the extracted case data includes information represented by numerical values, the selection support means displays statistical information of the numerical values on the display device. An FMEA sheet creation method for creating an FMEA sheet from case data in which cases are recorded in sentences,
Computer
A process of reading an item list in which a plurality of items of the FMEA sheet are defined together with their arrangement order;
A phrase extraction process for extracting phrases from a plurality of given case data;
A semantic network generation process for generating a semantic network having each phrase as a node from the phrase group extracted by the phrase extraction process;
A selection support process for displaying the semantic network on a display device, and allowing a user to select a plurality of words linked on the displayed semantic network;
An FMEA sheet generation process for generating an FMEA sheet having the contents of each item selected in the selection support process;
FMEA sheet creation method to execute. An FMEA sheet creation program for creating an FMEA sheet from case data in which cases are recorded in sentences,
On the computer,
A process of reading an item list in which a plurality of items of the FMEA sheet are defined together with their arrangement order;
A phrase extraction process for extracting phrases from a plurality of given case data;
A semantic network generation process for generating a semantic network having each phrase as a node from the phrase group extracted by the phrase extraction process;
A selection support process for displaying the semantic network on a display device, and allowing a user to select a plurality of words linked on the displayed semantic network;
An FMEA sheet generation process for generating an FMEA sheet having the contents of each item selected in the selection support process;
FMEA sheet creation program for executing   A computer-readable recording medium on which the FMEA sheet creation program according to claim 9 is recorded.

Images