JP2000137615A - Case-based reasoning method and system and readable recording medium for incorporating program for executing case-based reasoning - Google Patents

Abstract

PROBLEM TO BE SOLVED: To extract a case further matched with words and phrases inputted by a user by adopting a procedure based on the different points of view. SOLUTION: In this case-based reasoning method, when words and phrases are inputted by an inputting part 41, a word and phrase case retrieval processing part 21 retrieves a case including words and phrases by using a case file 51, and a scoring part 22 scores the words and phrases in a range up to the maximum score 40 according to the level of matching including words and phrases, and a rearranging part 23 rearranges the words and phrases in the order of high scores, and a monitor 42 displays them with a related question. When the displayed question is answered, the score is re-calculated so as to increase or decrease by a re-scoring part 25 according to the answered result. When the maximum score reaches a certain value 80 by the re-scoring, an evaluation judging part 26 ends inference processing for recommending the event of the maximum score.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a case-based inference method for executing a case-based inference on an input phrase,
The present invention relates to a readable recording medium storing a system and a program for executing case-based reasoning.

[0002]

2. Description of the Related Art Conventionally, in a case-based inference system having a plurality of similar cases in a case base, case-based inference is performed on an input case, and information on a set of matching cases for a problem template is provided. An inference engine that seeks advice from an operator for one set of matching cases has been proposed (Japanese Unexamined Patent Publication No. 6-508702). In other words, the case-based inference system described in Japanese Patent Application Laid-Open No. 6-508702 executes case-based inference to extract a case that matches the customer description input and displayed on the display panel. For example, an advisory message may be provided to the customer and in response to the advisory, the application 601 may perform the case matching step 202 using the question and answer pair 608 as additional matching attribute and value pairs. (U.S. Pat. No. 8, upper right column), the best match case is extracted by repeatedly executing the above case-based inference.

Further, as an inference system on the premise of case-based inference, those described in JP-A-3-266031 and JP-A-3-148727 have been proposed.

[0004]

[Problems to be Solved by the Invention]
The case-based inference system described in Japanese Patent Publication No. 2 can repeatedly extract the best match case for the description from the customer as a result of repeatedly executing the case-based inference in “Step 202” in response to the response to the advice message. , Since it is merely a matter of repeatedly executing case-based reasoning on customer statements and advice,
Matching cases extracted in each execution of case-based inference are inconsistent, and as a result of executing the same case-based inference, there is naturally a limit in extracting the best matching case. Further, the inference systems described in JP-A-3-266031 and JP-A-3-148727 only deduce cases from input words and phrases and devise similarities at the time of extraction. It is not about methods.

[0005] The present invention has been made in view of the above,
In case-based inference, by adopting a procedure based on different viewpoints, it is possible to extract a case that is more matched to the phrase input by the user, or a case-based inference method, a system, and a system that can support evaluation of the inference result. It is an object of the present invention to provide a readable recording medium storing a program for executing case-based reasoning.

[0006]

According to a first aspect of the present invention, a case-based inference method performs a case search between at least one or more input words and a case file and determines a degree of matching with the words and phrases. In response, executing the scoring for each of the searched cases and reporting the result, and reading and notifying a question associated with the searched case from the question file, and And performing a re-scoring for each searched case.

A case-based reasoning system according to a twelfth aspect of the present invention includes a case file, a question file, and a file in which a question and a case are associated with each other. Case searching means for performing scoring, scoring means for scoring each searched case in accordance with the degree of matching with the word, and outputting the searched case and score to the display means in association with each other. A first display control unit, a second display control unit that outputs a question associated with the searched case to the display unit, and a score for each searched case in accordance with an answer to the displayed question. Rescoring means for rescoring.

[0008] At least one readable recording medium having a program for executing the case-based inference according to claim 16 incorporated therein.
A case search is performed between the input phrase and a case file, and according to the degree of matching with the phrase, scoring is performed on each searched case and the result is reported, and the search is performed. A question associated with the searched case is read from the question file and reported, and re-scoring is performed for each of the searched cases according to the answer result to the question.

According to these, at least one word or phrase including a sentence is input as an assignment from a user using a man-machine interface or the like, and the input word is input as a first stage using a case file. Is searched. According to the degree of matching with the phrase for the searched case, for example, if there are a plurality of phrases, according to the number of inclusions of how many words are included, scoring is given to each case, and The result is notified to a display device or the like. Then, the question associated with the searched case is read from the question file and reported to a display device or the like. When the user answers the question, rescoring is performed for each of the searched cases according to the answer result. Then, the case with the highest score is presented to the user in a form that recommends the case with the highest likelihood for the task.

According to a second aspect of the present invention, the notification is displayed on display means. In a readable recording medium incorporating a program for executing case-based inference according to claim 17, the notification is displayed on a display means. As a result, many pieces of information can be recognized at the same time, and the user's operation work becomes easy and efficient.

According to a third aspect of the present invention, a measure corresponding to each retrieved case is reported together. In the case-based inference system according to claim 15, the first display control means displays on the display means a measure corresponding to each case. A readable recording medium incorporating a program for executing the case-based reasoning according to claim 18 is for notifying a measure corresponding to each searched case together. As a result, not only high-probability cases can be recommended but also direct measures as actions to be taken are presented, so the function as support is effectively demonstrated.

In the case-based reasoning method according to the fourth aspect, words for search are extracted from a sentence input from input means. A readable recording medium incorporating a program for executing case-based reasoning according to claim 19 is configured to extract a phrase for search from a sentence input from input means. Thus, it is not necessary to memorize the input terms individually, and the burden on the user is reduced, so that the versatility is easy to use.

[0013] In the case-based reasoning method according to the fifth aspect, the question can be selectively answered. A readable recording medium incorporating a program for executing the case-based reasoning described in claim 20 is capable of selectively answering the question. As a result, the degree of freedom to appropriately answer the question expected by the user is increased.

According to a sixth aspect of the present invention, the display of the cases is rearranged in the order of the score at each rescoring. A readable recording medium incorporating a program for executing the case-based reasoning according to claim 21 is arranged so that the display of cases is rearranged in the order of scores every time rescoring is performed. As a result, inference work can be performed without overlooking a case with high probability. Further, in a display mode in which the number of display lines is small in the display means and scroll display is required, the highest score is always displayed, which is very useful.

In the case-based reasoning method according to the present invention, the questions associated with the displayed cases are displayed in correspondence with the display order of the cases. A readable recording medium incorporating a program for executing the case-based reasoning according to claim 22 is configured to display questions associated with the displayed cases in a display order of the cases. As a result, the questions are displayed in the order corresponding to the case, so that the questions can be displayed more accurately.

In the case-based reasoning method according to the present invention, when the highest score reaches a predetermined value by re-scoring, the reasoning process is terminated. Also,
A readable recording medium incorporating a program for executing case-based inference according to claim 23 terminates the inference processing when the highest score reaches a predetermined value by re-scoring. These enable faster inference work.

According to a ninth aspect of the present invention, there is provided a case-based reasoning method, wherein the case and a question associated with the case are tabulated and displayed in a form that can be analyzed. Also,
A readable recording medium incorporating a program for executing a case-based reasoning according to claim 24, tabulates and displays the case and a question associated with the case in an analyzable manner. By these, we analyze the relationship between the question and the case and the question and examine the validity of the questions included in the case (expression method, number, order, etc.), and obtain hints for improvements to the field work from the entire case Becomes easier.

A case-based inference method according to claim 10 is:
Searching and displaying a question including at least one or more input words in the question file, reading and displaying a case associated with the displayed question from the case file, and displaying the answer to the displayed question. And re-scoring each of the displayed cases accordingly.

A case-based reasoning system according to a thirteenth aspect includes a case file, a question file, and a file in which a question and a case are associated with each other.
A question search unit for searching a question including a phrase entered in a question file; a unit for reading a case associated with the searched question from a case file; and a display unit for searching the searched question together with the read case. A first display control means for outputting and a means for re-scoring each of the displayed cases according to the answer result to the displayed question.

A readable recording medium incorporating a program for executing the case-based reasoning described in claim 25 searches and displays a question including at least one or more input words in a question file, and displays the displayed question. Read the case associated with the question from the case file and display it,
The re-scoring is executed for each case being re-displayed in accordance with the result of answer to the displayed question.

According to these, the questions including the input words and phrases are displayed together with the examples associated with these questions, and the user can answer the displayed questions, so that the user is familiar with the questions. In the case of a field, for example, the possibility of arriving at a case with higher certainty more quickly is improved.

A case-based inference method according to claim 11 is:
Prior to the step of executing the re-scoring, a step of setting a condition for the displayed question and performing a search, and re-displaying the question satisfying the search condition together with a related case is provided. The case-based reasoning system according to claim 14 performs a search by setting a condition for the question displayed by the first display control means, and re-executes a question satisfying the search condition together with the associated case. It has means for displaying. Further, the readable recording medium incorporating the program for executing the case-based reasoning according to claim 26 allows the displayed question to be set and searched before executing the re-coring. The question that satisfies the condition is redisplayed together with the related case. As a result, when the number of questions is large, questions can be narrowed down under appropriate desired conditions, so that operational efficiency and workability are improved.

[0023]

FIG. 1 is a block diagram showing an embodiment of a case-based reasoning system. In the figure, the system includes a processor 10 for executing case-based reasoning and the like, and a case-based reasoning engine 20, and also includes a man-machine interface unit 40 and a case base unit 50.

The man-machine interface section 40 comprises an input section 41 and a monitor 42. Input unit 41
A keyboard, such as a keyboard used in a normal computer, in which characters, numerical values, and the like can be input and keys for instructing a specific function are arranged as necessary is employed. The monitor 42 employs a display device that is capable of displaying at least characters and includes a CRT, an LCD, and various other display elements. As long as information can be exchanged between the user and the user, the man-machine interface unit 40 may employ a mode other than key input, such as voice recognition means, as the input unit 41. Is the monitor 42
Is great at recognizing large amounts of information simultaneously,
The present invention is not limited to this, and voice guidance may be performed using voice synthesis means.

The case base unit 50 includes a case file 51,
It is composed of a phrase case index file 52, a question file 53, and a question case index file.

The case file 51 stores, for example, in the case where the present system is employed for quality control support, examples of various claims and the like as past results (including those that can be predicted if necessary). The details of the countermeasures and, if necessary, the details of the countermeasures and the like are stored in association with the case numbers. The case file 51 includes, in addition to the contents of the case, the countermeasure thereof, a detailed description of the case or the countermeasure, and a detailed diagram, as shown in FIG. (Stored in the identification number field) and a question number (stored in the question field) for identifying one or a plurality of questions associated with each case and stored in a table format. In addition, it is preferable that each of the question numbers associated with each case number is hierarchically arranged and stored as shown in the question field. For example, in case numbers 5 and 7, question number 2 is positioned at the top, while in case number 10, the number is several from the top.

In this embodiment, the case means that the system is necessary or effective to receive the quality control support from the system, such as the details of the abnormality in the manufacturing process of the metal material and the symptoms caused by the abnormality. Means a questionable expression), and is mainly expressed in sentences. On the other hand, words and idioms are mainly used as phrases for searching cases. For example, "compressor", "pump",
The terms include "crack", "oil leak", "vibration", and the like, but may also include negative forms such as "do not start" from the viewpoint of expressing symptoms. Further, the term "measures" may include general conclusions including causes.

The phrase case index file 52 is a content inputted by the user, and is shown in FIG. 2 in which the phrase expected to be included in the case that the user wants to search is associated with the case stored in the case file 51. Each phrase (stored in the phrase field) is stored in the form of a table with the case number (stored in the case number field) of the corresponding case. A mode in which a word and a case are stored in association with each other via an index formula or the like may be stored. This has an advantage that the storage capacity of the case can be reduced.

FIG. 2 is an explanatory diagram showing an outline of a procedure for creating a phrase case index file 52 from a case file 51 using the database construction unit 20a. The database construction unit 20a creates a phrase case index file 52 from the case contents and the case number field of the case file 51.
That is, the database construction unit 20a extracts a phrase expected to be used by the user as a phrase from the content of the case, and creates a case number including each phrase in a table format.
In the example of FIG. 2, the phrase case index file 52 indicates that the phrase “compressor” is included in case numbers 3, 5, and 7, and the phrase “pump” is included in case numbers 5 and 5.
7 and 10.

The question file 53 is, as shown in FIG. 3B, a question sentence (stored in a question sentence field) which can be answered to the user, for example, “Yes, No”.
Is associated with a question number (stored in a question number field) and stored. Further, in the question file 53, in addition to the above recorded contents, weight values for determining whether to increase or decrease the score for each case according to the answer result of each question are stored as attribute values. The weight value may be a coefficient by which the score is multiplied, or a value that is added or subtracted.

The question case index file 54 is shown in FIG.
Question number (stored in question number field) as shown in
Is associated with a case number (stored in a related case number field) in a table format.

An example of the recording mode of the case file 51, the question file 53, and the question case index file 54 will be described in detail with reference to the explanatory diagram of FIG. Focusing on the case number 5, the case number 7, and the case number 10, all of them include the question number 2 (the question corresponding to the case number 7 is one of the question numbers 2, 10 includes a plurality.). In the question file 53, a question sentence corresponding to each question number is stored such that a question sentence "Is the model A?" Then, in the question case index file 54, related case numbers for each question number are recorded in an index format such that related case numbers 5, 7, and 10 are stored based on the question number 2.

The processor 10 is a case-based inference engine 2 including a function unit for executing a search step as a first stage and a function unit for executing an inquiry step as a second stage.
0 and man-machine interface 4
An interface 31 and a display control unit 32 are provided between the display control unit 32 and the interface 0.

The data extraction unit 11 extracts a phrase for inference processing (text code) from one or more input sentences (contents of a task including a phrase expected to be included in a case to be searched) input from the input unit 41. For) extraction. From the input unit 41, for example, in response to an input sentence such as "the pump is leaking oil" or "the motor does not start", the words "pump", "oil leak", "motor", and "do not start" are displayed. The phrase is extracted, and a phrase (term, idiom, negative form predicate) intended by the user is extracted from the text by using a grammatical analysis process. The input sentence is prepared so that a plurality of lines can be displayed on the monitor 42 (see 421 in FIG. 4). When the input sentence is plural, each sentence can be identified by line feed for each sentence. Alternatively, the break of each sentence can be determined based on the period. Note that if it is known that a part or all of the input content from the input unit 41 is a word (for example, a line is broken in a state where no punctuation is entered in the line), the word is not included in the data. Through the extraction unit 11, the case file 51 can be directly accessed.

The operation judging section 12 is for judging whether or not execution of a specific function has been instructed, especially for other than text input. Specific functions, which will be described later on the monitor screen shown in FIG. 4, include, for example, input determination and answer determination as conditions for shifting to the next process, and an instruction to perform a logical sum search or a logical product search. .

The case-based inference engine 20 includes a phrase case search processing unit 21, a scoring unit 22, and a rearrangement processing unit 23 as functional units for executing a search step as a first stage, and an inquiry step as a second stage. Includes a question case search processing unit 24, a re-scoring unit 25, and an evaluation determination unit 26.

The detail display processing section 27 is a functional section for displaying the detailed contents of the best search case on the monitor 42, and the analysis processing section 28 analyzes the relation between the question and the case and the question to form the case. When examining the validity of the included questions (expression method, number, order, etc.), make a table in an appropriate form and display it, and analyze the relation between measures and cases, measures This is a function unit for similarly examining the validity of the included measures and displaying the same, and the details will be described later.
The case correction processing unit 29 is a functional unit that makes a correction to the searched case.

The phrase case search processing section 21 is a data extraction section 1
The related case is searched by using the case file 51 and the phrase case index file 52 by using the case file 51 and the phrase case index file 52. The phrase case search processing unit 21 reads, for example, a case including a phrase from the phrase case index file 52 by a case number, and reads the case contents of the case number from the case file 51 together with the countermeasure and the like. As will be described later, when a plurality of words are input, processing can be performed by the respective inference methods depending on whether inference by logical OR or logical product is instructed.

The scoring unit 22 assigns a score according to the degree of coincidence (degree of coincidence) between the phrase and the case according to a preset method. In the first step, the degree of coincidence is 1 (% ) (Hereinafter, referred to as a score 1), a score is preferably given with a predetermined value preferably equal to or less than 50 as the maximum value (score 40 in the present embodiment). By setting the likelihood of the search step in the first stage to preferably 50 or less, the reliability of the first stage is ensured, the significance of the second inquiry step is increased, and the case with higher certainty is finalized. It can be reasonably inferred.

As one embodiment of the scoring method, a score 40 is set for a case including all the words for a given number of words, and thereafter, the number of words to be included (minimum 1) is set. ), The scores lower than the maximum value are sequentially calculated and set according to a preset calculation method (including a tabulated mode). For example, when inference is performed on two words, a score 40 is given to a case containing two words, and a score 20 is given to a case containing one word. Note that all the searched cases may be displayed on the monitor 42 as search results,
In consideration of the fact that the retrieved cases are rather complicated, such as when a large number of cases are found, those having a score higher than a predetermined score, for example, score 1
Only those that clear 0 or more may be the results of the first search step.

The rearrangement processing section 23 rearranges the searched cases in the order of the scores assigned to the cases, and
In the case display column No. 2 (see 423 in FIG. 4), the display is performed in order from the first row to the lower row.

The question case search processing unit 24 searches the monitor 42 for the question sentences extracted in the first search step and individually associated with the cases having the highest scores. The system reads out the question number assigned to the relevant case from 54 and reads out the question sentence of the question number from the question file 53.
Usually, there are a plurality of extracted cases, but one or more question sentences associated with each case are read from the question file 53 and displayed in the question display column of the monitor 42 in a line-by-line manner in the order in which the cases are displayed. Is done. If there is a common question number for the searched cases, the display of the second and subsequent same question texts is omitted in order to reduce the complexity of the display.

The re-scoring unit 25 recalculates the score as the certainty of the case according to the answer of “Yes” or “No” to the question sentence corresponding to the case. As described above, the re-scoring calculation method uses a weight value corresponding to the answer “Yes” or “No” to the question, and is given to the corresponding case for the answer “Yes”, for example. The current score is "Ye
s "(e.g., in the case of a coefficient, by a ratio) according to the weight (match weight) set for the" s "(when the calculation result exceeds 100, the score is 10).
0 or or score 99), "No"
(“Yes”) according to the weight (mismatch weight) set for “No” for the current score assigned to the corresponding case
(Which may be the same as the weight value set for). The rescoring may be addition or subtraction instead of the ratio calculation, or another method may be adopted. For cases not associated with a question (that is, cases not having the question), the weight value as the absolute weight is similarly set corresponding to “Yes” or “No”, In this way, the score is increased or decreased even for a case that is not associated with the question.

In the second stage of the inquiry step, the evaluation judging section 26 selects, among the scores obtained by the re-scoring,
It is determined for each re-scoring whether the highest score has reached a preset value, for example, a value 80, and the highest score is
When the number exceeds the limit, the inference procedure is automatically terminated.

The detailed display processing section 27 displays the countermeasure contents corresponding to the searched case in the countermeasure display column (FIG. 4) on the screen of the monitor 42.
424). By specifying the top-level measure or any measure among these measures, specifying a more detailed measure content and a detailed measure reference diagram, and specifying a display line, the case file is displayed. The information is read from the display 51 and displayed in a predetermined area of the monitor 42 (for example, a new window is opened in the screen) so as to show the user the details of the countermeasure. This detailed display processing unit 27
Is also driven in response to an instruction to display case details and case reference diagrams.

FIG. 4 is a diagram showing one display example of the screen of the monitor 42. As shown in FIG. As shown in the figure, at the top of the monitor screen 420, there are a plurality of menus such as "search", "clear", "answer evaluation", ..., "case information", "case reference diagram", "measure details", "measures" “Reference diagram” and “Exact search” are prepared in the form of icons.
1 has a cursor pointing member 411 (for example, a mouse)
By moving the cursor using and clicking (or double-clicking as appropriate), the function of the display menu is selectively executed. In addition, below these display areas, a “detailed description (keyword)” column 421, a question display column 422, a case display column 423, and a countermeasure display column 424 for sequentially displaying input sentences by predetermined lines are displayed. Is provided. Each of the columns 421 to 424 is not limited to the number of lines displayed in FIG. 2, and a larger number of lines is prepared, and the cursor indication member 411 is used to scroll and display in a line direction for each column. This allows a desired amount (row) of content to be entered or displayed.

The menu "search" is for instructing a search by a logical sum (first search step: coarse search) for a plurality of input words and phrases. The menu “Clear” is for instructing the deletion of the input content for the search displayed in the “detailed description (keyword)” column 421, and the selective deletion of the content of the other columns. The menu “response evaluation” is for instructing evaluation of a case, that is, re-scoring, after input operation of one or more answers to the displayed question (determination of the answer). In addition, 1
It is also possible to selectively specify the mode of evaluating each answer one by one,
In this case, the menu "response evaluation" becomes unnecessary.

The menu "case information" is for instructing the display of "detailed description" for the case specified as shown in FIG. The menu “case reference diagram” instructs to open and display a new window such as a chart (reference diagram) captured as an image for the purpose of showing the details of the case as shown in FIG. The menu “exact search” is for instructing a search by a logical product (first search step: coarse search) for a plurality of input words.

Subsequently, the monitor screen 4 shown in FIGS.
20 will be described. FIG. 5 is a screen showing a data input state. In this display example, the word “scratch” is displayed on the first line of the “detailed description (keyword)” column 421, and the word “crack” is displayed on the second line. I have.

FIG. 6 shows a display screen at the time when the first search step is completed. The search results are further displayed in a question display column 422, a case display column 423, and a countermeasure display column 424. In the first line of the case display column 423, the score 4
0, the case number C6 indicates "crack during cold working", the second line indicates the score 40, and the case number C9 indicates "crack during cold forging". In the first line of the countermeasure display column 424, as a countermeasure corresponding to the case of the first and second lines, the first line is "Thorough inspection of steel billet" and the second line is "Do not leave alumina-based inclusions under the skin. Is displayed.

In the question display column 422, “Is the Q13 steel type S40C?” Is displayed as a question associated with at least the case on the first line. The first line indicates that the display mode has been changed (black and white inverted display), and that this question has been selected to answer (ie, “unanswered”).

FIG. 7 is a screen showing a state where a question on the first line or a question on a desired line of the question display column 422 is selected, and a window screen 422a for accepting an answer is displayed. Now, "Ye" is
s "is selected. In this state, if the cursor pointing member 411 is clicked, "Yes" has been answered to the selected question.

FIG. 8 is a screen showing an example of answers to a question. The display form of the answered question is different from the unanswered display form. Although not shown in the figure, the display line that has been answered is moved to the bottom line of the question column, thereby facilitating the selection and answering of unanswered questions. For unanswered questions, the display order of the questions is changed according to the rearrangement of the cases by the re-scoring process.

FIG. 9 is a display screen showing an example of the inference result. The case “C6 crack during cold working” in the first line of the case display column 423 has a score of 100. In this case, in the first search step, as shown in FIG.
It was what was. Similarly, the first line of the countermeasure display column 424 has the contents corresponding to the first line of the case, “A5
Thorough inspection of billets "has a score of 100. Also,
As a result of the question being answered to Q13, Q14, and Q15 (Q is the meaning of the question) shown in FIG. 6 and being moved to the lower row, in this display example, Q26 which was in the next row is displayed in the uppermost row. Have been.

FIG. 10 shows a display screen in a state where the menu "case information" is designated, in which information on cases and countermeasures is confirmed. When the menu "case information" is instructed, a case number and a title column are displayed. In this column, "crack during cold working" which was on the first line of the case is extracted as case number "6". I have.

The content to be displayed may be sequentially instructable from the first line of the case, or the user may be irrespective of the line position (without being limited to a score of 80 or more which is a predetermined value or not). It is also possible to select a desired one.

At the position immediately below the title, the details of the case are read from the case file 51, and for example, "A crack occurred during cold forging (screw rolling). The steel type is S40C.
Caused by roll cracking due to pressure flaws. ”Is displayed at the same time, the answer history to the question is displayed with the question number, and the answer to the question can be confirmed. Is displayed.

FIG. 11 shows a display screen in a state in which the menu “case reference diagram” is designated, in which a new window is opened and detailed information of cases and countermeasures is shown in an image such as a chart.

FIG. 12 is a flowchart showing the procedure of the case-based reasoning method. This flow is roughly divided into a search step as a first stage (steps S1 to S11) and an inquiry step as a second stage (steps S13 to S23).
Consists of

First, words and phrases as data input by the user from the input unit 41 are managed, and whether or not the input has been completed is determined by a menu "search" or a menu "strict search".
Is confirmed by the operation determining unit 12, and if there is no instruction, the input permission state is maintained (No in steps S1 and S3). The input content is displayed in a “detailed description (keyword)” column 421 as shown in FIG. For example, if there is an instruction of the menu “search”, the data extracting unit 11 executes a process of extracting a phrase from the input data (step S5). If it is determined that the extraction process is unnecessary because the word is a term or the like, step S
5 is passed through.

When the phrase extraction process is completed or skipped for all the input data, the case-based inference engine 20 executes a case retrieval process including the input phrase as the first retrieval step (step S7). . Then
Scoring (score calculation and addition) is performed on the case searched in the search process according to a calculation method set in advance according to the degree of inclusion of the word (match degree) (step S9), and the score is calculated. Are sorted in the descending order, and the result is guided to the monitor 42 and displayed in the question display column 422, the case display column 423, and the countermeasure display column 424 as shown in FIG. S
11).

Subsequently, the case display field 42 of the searched case
1 associated with the case displayed in conjunction with the display to 3
Alternatively, for a plurality of desired question sentences, selection from the top or by the user's arbitrary selection is made by the cursor instruction member 41
1 (step S13). Then, each answer from the selected one or a plurality of questions, that is, one of “Yes” or “No” is designated as shown in FIG. 7 (step S15), and further, the menu “answer evaluation” is instructed. Then, the answer is determined (step S17), and the re-scoring process is performed (step S19).
Then, a question search consisting of a display sorted in the order of scores (step S21) is executed. Then, it is determined (evaluated) whether or not the highest score has reached the value 80 (step S23),
As shown in FIG. 9, when an example exceeding the value 80 is obtained,
This flow ends.

On the other hand, if the highest score does not reach the value 80 even after re-scoring, it is determined whether the user has performed an answer operation for the next (unanswered) question or a new data input (step S25). If the answering operation has been performed using 411 and the menu "answer evaluation", steps S13 to S23 are repeated.

Conversely, if the user determines that a higher score is unlikely to be obtained even if the inquiry step is continued, input (including deletion) of a new word or phrase can be accepted,
When data is input by the user in this state, the process returns to step S1 and includes a new word, or deletes unnecessary words from the previous input word appropriately using the menu “Clear”, etc. This flow is continued from step S1. Note that, even before the highest score reaches the value 80, the present flow may be forcibly terminated by an operation from the user, and the search result before the termination may be maintained.

Although not shown in FIG. 12, after this flow is completed, the menu “case information” is designated to display the detailed contents of the case together with the countermeasure as shown in FIG.
Further, by specifying the menu “case reference diagram”, a detailed reference diagram of the case can be displayed on the display screen as shown in FIG.

FIG. 13 is an explanatory diagram showing the contents of "phrase search" in step S7, and corresponds to the example of FIG. In the figure, for example, assuming that the words “compressor” and “pump” are extracted by the data extracting unit 11, the case numbers 3, 5, 7 and the word “pump” including the word “compressor” are extracted from the word case index file 52. Are included, and the degree of matching is extracted. That is, case numbers 5 and 7 are recognized as cases including all search phrase groups, and case numbers 3 and 30 are recognized as cases including only one search phrase. For example, a score 40 is given to the former case numbers 5 and 7, and a score 20 is given (scored) to the latter case numbers 3 and 30, for example.

FIG. 14 is an explanatory diagram showing the contents of “re-scoring” and “score order rearrangement display” in steps S19 and S21, and corresponds to the examples of FIGS. The re-scoring is processed using the case file 51 and the question file 53. In the question file 53, a weight value (match or mismatch) is set for the associated case number as described above for each question number, while the absolute value is also set for cases other than the associated case. Up and down weight values using weights are set (not shown in FIG. 14). Therefore, according to the answer to the question associated with the case displayed in the case display column 423, the current score of the case goes up or down, and the current score of the other case is set. The current score is incremented or decremented in accordance with each weight value at the time of the down and down. In addition, a case where the current score is reduced to be equal to or less than a predetermined threshold value may be deleted in consideration of display capacity and the like. Absence weights for cases that do not have a answered question need not be uniform, and are based on the relationship between questions and cases that are hierarchically constructed, taking into account the similarity of the relevant case or question. It may be decided beforehand.

When the score is changed by the re-scoring, the cases are rearranged in the order of the highest score, the questions are also rearranged in the order of the cases, and the result is displayed on the monitor 42. .

Next, a second embodiment of the present system will be described. The system configuration in the second embodiment is shown in FIG.
The word case index file 52, the word case search processing unit 21, and the scoring processing unit 22 are excluded from the block diagram shown in FIG.

FIG. 16 is a flowchart showing a second embodiment of the present inference system.
The description is omitted because it is the same as steps S1 to S5 in FIG.

When a word is input, the question file 53
(Question sentence field), all question sentences including the phrase are read out, and corresponding cases are extracted using the case file 51 via the question case index file 54, and the monitor 42 asks questions, cases, and countermeasures. Is displayed correspondingly (step S37). At this stage, only the question is extracted, and the score assignment process is not particularly performed on the case read out as the corresponding case. For example, a uniform score is set as the initial value. Alternatively, it may be in an unassigned state.

Next, a search process is executed for the question (step S39). That is, a search keyword can be input in a detailed explanation (keyword) display column for a question category prearranged from the displayed question or for a phrase included in the question sentence, and when a menu “search” or the like is designated, Only the questions that satisfy the search conditions are extracted and displayed in the question display column (step S41), and questions outside the search conditions are deleted. At this time, only the case and the measure corresponding to the question satisfying the condition are displayed in the respective display columns. Next, a question is selected from the top questions or a desired question (step S43), and “Yes” and “N” are selected.
When an answer is given by "o" and the menu "response evaluation" is instructed (steps S45 and S47), rescoring is executed according to the answer result as in the first embodiment (step S49), and the scores are arranged in order. The cases, that is, the questions are rearranged in a corresponding manner and displayed on the monitor 42 (step S51).

In step S41, it can be distinguished according to whether the search condition is satisfied (see FIG.
5 is not shown), each question may be displayed. FIG. 15 is a diagram illustrating an example of the screen after the question search. At a slightly central position on the monitor screen 420, a menu “Question jump” or the like for enabling an answer only to a question satisfying the search condition is provided. The menu “question jump” designates, in response to each answer only to a question that satisfies the search condition, a corresponding row (from the top row) which can be answered sequentially. As shown in the figure, the user can answer a question desired by the user (within a range satisfying the search condition).

Subsequently, it is evaluated whether or not the score exceeds a predetermined value 80 (step S53). If the score exceeds the predetermined value, the flow ends as a search end. If not, the flow returns to step S31 to return to the word data. Is restarted.

FIG. 17 is a flowchart showing a third embodiment of the present inference system.

Steps S61 to S71 are the same as steps S1 to S11 in FIG.
7 is the same as steps S39 to S53 in FIG. Step S89 is for selectively permitting the question search in step S73 and the re-input of data in step S61 when the maximum score of the case does not exceed the value 80 in step S87.

Next, the configuration and operation of the analysis processing unit 28 shown in FIG. 1 will be described. The analysis processing unit 28 includes the input unit 4
The menu “Question analysis” separately prepared in the form of an icon by using one specific key or the cursor indicating member 411.
, The questions stored in the question case index file 54 are read out in association with related cases (by question numbers and case numbers), and hierarchically sorted as shown in FIG. And a screen 42 of the monitor 42 in the format shown in FIG.
It has a tabulation function unit for displaying the number 0. Question file 53
To enable this question analysis, for example, FIG.
As shown in, data is stored in the question field in a form in which a hierarchical relationship is added between the questions in advance, and each question is processed so as to be included in the hierarchical relationship together with an example included in the question. , And display. The display format of the relationship between a case and a question may be such that a list of questions included in each case is displayed based on the case as shown in FIG. In this case, the case file 51 may be used.

Further, the analysis processing unit 28 is provided, for example, in FIG.
(B) to (c) have an analysis function unit for displaying a question included in the case as shown in the example in the inclusion diagram so that the inclusion relationship can be easily recognized. ), A tabulation function unit for displaying an inclusion list is also provided.

FIGS. 18A to 18D show on the monitor screen 420.
By selecting and displaying the hierarchical relationship and the inclusion relationship collectively or individually (including other than the batch) in the form of, analysis is performed so that the preferred hierarchical relationship and the inclusion relationship are obtained, and the user is appropriately corrected (added, Change, move, delete, etc.). Specifically, the correction is to add, change, move, or delete corresponding cases and questions for each of the files 51 to 54. It is also possible to grasp or propose a problem of the entire case.

Further, the analysis processing unit 28 sets the case base 5
Measure file 55 and measure case index file 56 as 0
In the embodiment provided with, the following analysis is also possible. The countermeasure file 55 describes the countermeasure contents in the case file 5.
This is necessary when the information is not stored in the storage unit 1, and the countermeasure number and the content of the countermeasure are stored correspondingly. The measure case index file 56 stores a measure number and a case number associated with the measure in a contrasting manner. As described above, by having a file for storing the content of the countermeasure separately from the case file 51, when a plurality of cases are associated with one countermeasure, the total storage capacity can be reduced. There is. Then, the analysis processing unit 28 performs the same processing as in the case of the above-described questions and the relationship between the question and the case also with respect to the relationship between the measures and the relationship between the measure and the case. It can assist in the analysis of the presence or absence of the inclusion relationship and in understanding or proposing the problems of the entire case. Note that in a mode in which the present system is equipped with an output unit such as a printer, the charts of FIGS. 18A to 18D may be output as a hard copy.

The present invention is also applicable to the following modifications.

(1) Although the present embodiment has been described by taking as an example the support for quality control abnormalities, the present invention is not limited to such fields, but can be applied to all inspection services, and furthermore, it can be applied to sales operations, estimate design Business, process design, planning (production planning, process design, scheduling, inventory planning), failure diagnosis, material design, operation, operation,
Like education and other specialized intellectual work in general,
It can be applied to a support system (expert system) for various tasks.

(2) In the present embodiment, as the re-scoring method, weighting coefficients for up and down are provided in the case associated with the question in accordance with the result of answering the question. Only the sign may be reversed.

(3) In the present embodiment, as the re-scoring method, the answer result to the question is made as an alternative. However, the re-scoring method uses a multiple selection method or a numerical value (such as a physical quantity).
Furthermore, the present invention can also be applied to a mode in which a response is made in characters. In these cases, the rescoring method may be adopted in a format that can deal with the contents of the answer.

(4) In the first to third embodiments, the condition for terminating the inference processing is that the case with the highest score reaches the value 80. However, the present invention is not limited to the value 80.
The value may be set to a value higher than the value 80 or a lower value. Also, this value does not need to be fixed, and may be appropriately set according to the case, or may be set to an appropriate value according to the type of business.

(5) In the first to third embodiments, when the case with the highest score reaches a predetermined value (for example, value 80),
The inference process is terminated, but is not limited to this.
The user may be notified that the highest score has reached the predetermined value (or without any special notification), and the end of the inference processing may be left to the user.

[0087]

According to the first, twelfth and sixteenth aspects of the present invention, inference is performed from different viewpoints, ie, word search and question-and-answer by questioning, so that a case with higher certainty can be extracted. A highly workable and versatile system can be provided.

According to the second and 17th aspects of the present invention, a large amount of information can be recognized at the same time, and the operation work by the user can be made easier and more efficient.

According to the third, fifteenth, and eighteenth aspects of the present invention, not only a case with high certainty can be recommended, but also a direct action as an action to be taken can be presented to make the function as a support effective. Demonstrate.

According to the inventions described in claims 4 and 19, it is not necessary to memorize the input terms individually, and the burden on the user can be reduced.

According to the fifth and twentieth aspects of the present invention, it is possible to increase the degree of freedom in appropriately answering the questions expected by the user.

According to the sixth and twenty-first aspects of the present invention, it is possible to perform an inference operation without overlooking a case with high certainty. Is very useful because you can always display the highest score.

According to the inventions as set forth in claims 7 and 22, since the questions are displayed in the order corresponding to the cases, the questions can be displayed more accurately.

According to the eighth and twenty-third aspects of the present invention, a quicker inference operation can be performed.

According to the invention of claims 9 and 24, the relationship between the question and the case and the question is analyzed to determine the validity (expression method, number, order, etc.) of the question included in the case, The task of studying to obtain hints for improvements to the system becomes easier.

According to the tenth, thirteenth, and twenty-fifth aspects of the present invention, it is possible to more quickly reach a highly reliable case, for example, when the user is familiar with the field.

According to the invention as set forth in claims 11, 14 and 26, when the number of questions is large, the questions can be narrowed down under appropriate desired conditions, and the operation efficiency and workability are improved. Can be.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a case-based reasoning system.

FIG. 2 is an explanatory diagram showing an outline of a procedure for creating a phrase case index file from a case file using a database construction unit.

FIG. 3 is a drawing illustrating the contents and structure of each file;
(A) is a case file, (b) is a question file, (c)
Indicates a question case index file.

FIG. 4 is a diagram illustrating a display example of a monitor screen.

FIG. 5 is a diagram showing a screen showing a data input state.

FIG. 6 is a diagram showing a display screen at the time when the first search step is completed.

FIG. 7 is a diagram showing a screen showing a state in which a question on a first line or a question on a desired line in a question display column 422 is selected.

FIG. 8 is a diagram showing a screen showing an example of an answer to a question.

FIG. 9 is a diagram showing a display screen showing an example of an inference result.

FIG. 10 is a diagram showing a display screen in a state where a menu “case information” is designated.

FIG. 11 is a diagram showing a display screen in a state where a menu “case reference diagram” is designated.

FIG. 12 is a flowchart illustrating a procedure of a case-based reasoning method.

FIG. 13 is an explanatory diagram showing the contents of “phrase search” in step S7.

FIG. 14 is an explanatory diagram showing the contents of “re-scoring” and “score order rearrangement display” in steps S19 and S21.

FIG. 15 is a diagram showing a display screen after a question search.

FIG. 16 is a flowchart illustrating a second embodiment of the present inference system.

FIG. 17 is a flowchart illustrating a third embodiment of the present inference system.

18A and 18B are diagrams showing the results of question analysis, wherein FIG. 18A is a hierarchical diagram, FIG. 18B is a list diagram, FIG. 18C is an inclusion diagram, and FIG. 18D is an inclusion list diagram.

[Explanation of symbols]

 10 Processor 11 Data Extraction Unit 12 Operation Judgment Unit 20 Case-Based Reasoning Engine 21 Phrase Case Search Processing Unit 22 Scoring Unit 23 Rearrangement Processing Unit 24 Question Case Search Processing Unit 25 Re-Scoring Unit 26 Evaluation Judgment Unit 27 Detailed Display Processing Unit 28 Analysis processing unit 29 Case correction processing unit 31 I / F 32 Display control unit 40 Man-machine interface unit 41 Input unit 411 Cursor pointing member 42 Monitor 50 Case base unit 51 Case file 52 Word / phrase case index file 53 Question file 54 Question case index File

Claims (26)

[Claims] 1. A case search is performed on at least one or more input words with a case file, and scoring is performed on each of the searched cases according to the degree of matching with the word. Notifying the result, and reading and notifying a question associated with the searched case from the question file, and performing re-scoring on each of the searched cases according to the answer result to the question And a case-based reasoning method. 2. The case-based reasoning method according to claim 1, wherein the notification is displayed on display means. 3. The case-based reasoning method according to claim 1, wherein a countermeasure corresponding to each searched case is reported together. 4. The case-based reasoning method according to claim 1, wherein a word for search is extracted from a sentence input from an input unit. 5. The case-based reasoning method according to claim 1, wherein said question can be selectively answered. 6. The case-based reasoning method according to claim 2, wherein the display of cases is rearranged in the order of scores each time rescoring is performed. 7. The case-based reasoning method according to claim 6, wherein the questions associated with the displayed cases are displayed in correspondence with the display order of the cases. 8. The case-based inference method according to claim 1, wherein the inference processing is terminated when the highest score reaches a predetermined value by re-scoring. 9. The case-based reasoning method according to claim 2, wherein the case and a question associated with the case are tabulated and displayed in a form that can be analyzed. 10. A question including at least one or more input words is searched for and displayed in a question file,
Reading a case associated with the displayed question from the case file and displaying the read case; and re-scoring each of the displayed cases in accordance with the result of answering the displayed question. A case-based reasoning method characterized by: 11. A step is provided in which, prior to the step of executing the re-scoring, a search is performed by setting a condition for the displayed question, and a question satisfying the search condition is re-displayed together with a related case. 11. The case-based reasoning method according to claim 10, wherein: 12. A case file comprising a case file, a question file, and a file in which a question and a case are associated with each other, a display means, a case search means for performing a case search between a case file and an input word, Scoring means for scoring each case according to the degree of matching with the phrase, first display control means for outputting the searched case and score in association with the display means, Second display control means for outputting a question associated with the searched case to the display means, and re-scoring means for re-scoring a score for each searched case according to an answer to the displayed question. A case-based reasoning system comprising: 13. A case file, a question file, and a file in which a question is associated with a case, display means, a question search means for searching a question including an input word in a question file, and association with the searched question. Means for reading the retrieved case from the case file, first display control means for outputting the retrieved question to the display means together with the read case, and displaying the displayed question in response to the answer to the displayed question. Means for rescoring each case. 14. A method for setting a condition for a question displayed by said first display control means, performing a search, and redisplaying a question satisfying the search condition together with an associated case. The case-based reasoning system according to claim 13, characterized in that: 15. The case-based reasoning system according to claim 12, wherein said first display control means displays a countermeasure corresponding to each case on said display means. . 16. A case search is performed for at least one or more input phrases with a case file, and scoring is performed on each retrieved case in accordance with the degree of matching with the phrase. Execute the case-based reasoning that causes the result to be reported, reads out the question associated with the searched case from the question file, and notifies the searched case, and executes the re-scoring for each searched case according to the answer result to the question. Readable recording medium with a built-in program. 17. The readable recording medium according to claim 16, wherein said notification is displayed on a display means. 18. A readable recording medium having a program for executing case-based inference according to claim 16 or 17, wherein a countermeasure corresponding to each searched case is reported together. 19. A readable recording medium having a program for executing case-based inference according to claim 16 or 17, wherein a word for search is extracted from a sentence input from an input means. . 20. A readable recording medium having a program for executing case-based inference according to claim 16 or 17, wherein the question can be selectively answered. 21. A readable recording medium having a program for executing case-based inference according to claim 17, wherein the display of cases is rearranged in the order of scores each time rescoring is performed. 22. A readable recording incorporating a program for executing case-based reasoning according to claim 21, wherein questions associated with the displayed cases are displayed in correspondence with the order of displaying the cases. Medium. 23. The readable recording medium according to claim 21, wherein the inference processing is terminated when the highest score reaches a predetermined value by the re-scoring. 24. The case-based inference according to claim 17, wherein the case and a question associated with the case are tabulated and displayed in a form that can be analyzed. A readable recording medium with a built-in program. 25. Searching and displaying a question including at least one or more input words in a question file,
A program for executing a case-based reasoning for reading a case associated with a displayed question from a case file and displaying the read case, and performing rescoring for each of the displayed cases according to an answer result to the displayed question. A readable recording medium with a built-in. 26. Prior to performing re-coring,
For the displayed question, set a condition and perform a search,
26. The readable recording medium according to claim 25, wherein the question satisfying the search condition is redisplayed together with the related case.