JPH07110832A - Device and method for retrieving merchandise information - Google Patents

Abstract

PURPOSE:To fetch information relating to a merchandise appropriate for a user or a condition by retrieving a previously stored data base when a selective input for the sort of a merchandise and a selective input at least either one of a user sort and a using condition sort exist. CONSTITUTION:When data relating to user's condition are inputted in addition to a merchandise section and a merchandise item, steps 33 to 36, 41 to 44 are executed, and in a merchandise master, the code and merchandise name of a merchandise having a user code, a daily life condition code and a physical condition code which are selectively inputted out of merchandise having selectively inputted merchandise section code and merchandise item code are read out. In the case of retrieving needs, steps 51 to 55 are executed and the code and merchandise name of a merchandise having selectively inputted needs section code and needs code are read out from a needs/merchandise connection file. Consequently an operator can select an optional merchandise from merchandise names displayed on a display screen.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a product information retrieval apparatus and method for retrieving information on a desired product from a data base in which information on many products is stored.

[0002]

BACKGROUND OF THE INVENTION One of the easiest ways to come up with information about a desired product from a database that stores information about many products would be to enter the product name. However, according to this method, when the product name is unknown, it cannot be dealt with at all, and when many products having the same product name exist, it is not possible to quickly reach the desired product.

The next thing that comes to mind is a search from the point of view of product classification. It will be easier to search if the categories are hierarchical, such as large categories, medium categories, and small categories.

[0004]

DISCLOSURE OF THE INVENTION However, these search methods use the attributes (or properties) of the product such as the structure, performance, and suitability of the product.
It completely lacks the perspective of considering compatibility with users. In particular, regarding nursing care equipment such as medical care beds, walking aids, wheelchairs, equipment that eliminates steps, bedsore prevention equipment, urine collection, stool collection, etc. The possibility of product search from a new perspective must be explored. Further, by reflecting the knowledge of the expert on the product, a product search that allows a user who does not have the specialized knowledge about the product to easily and accurately search for a desired product must be sought.

The present invention provides a product information search device and method that can search product information from the viewpoint of a product desired by a user or suitable for the user.

[0006] The first invention proposes an apparatus and method capable of retrieving information about a user or a product suitable for the situation.

The product search device according to the first aspect of the present invention is associated with each of a plurality of products and matches the product category to which the product belongs, the classification of users suitable for using the product, and the use of the product. A display device for displaying a plurality of types of screens including a database in which at least one of the categories relating to the user's situation and product information is stored in advance, a list for searching product information, and a display device. An input device for selecting at least one from the list displayed on the display screen, and a screen for inputting a product category, which is displayed on the display device and waits for selection input from the input device. And a screen for inputting at least one of the classification of the user and the classification relating to the situation of the user are displayed on the display device, and the input device is displayed. Input control means for performing at least a first processing of the second processing to wait for al selection input,
Also, when there is a selection input for the product classification from the input device, the database is searched, a list of products belonging to the selected and input product classification is created, and is displayed on the display device. When there is a selection input for the product classification and a selection input for at least one of the user classification and the usage status classification from the input device, the database is searched, and the product classification selected and input. Of the merchandise belonging to, the merchandise is provided with search processing means for extracting merchandise applicable to at least one of the user classification and the user situation classification and displaying the list on the display device. .

The product information retrieval method according to the first aspect of the present invention is associated with each of a plurality of products, and matches the product category to which the product belongs, the classification of users suitable for using the product, and the use of the product. At least one of the categories relating to the user's situation and the product information are stored in the memory in advance, a screen for inputting the category of the product is displayed on the display device, and the selection input from the input device is performed. A first process for waiting for the user to display a screen for inputting at least one of the user's classification and the user's status classification, and wait for the selection input from the input device. At least first of the processing
When the input device receives a selection input regarding the product classification, the above memory is searched, a list of products belonging to the selected product classification is created and displayed on the display device, When there is a selection input from the input device regarding the product classification and at least one of the user classification and the user situation classification, the above memory is searched and the selected product classification is entered. Out of the commodities to which the user belongs, the commodities applicable to at least one of the user classification and the user situation classification are extracted and the list is displayed on the display device.

According to the first invention, it is possible to retrieve product information from the viewpoint of product classification. Preferably, the commodities are hierarchically classified in two or more stages, and the range of commodities can be narrowed down sequentially from the major classification to the minor classification.

In particular, according to the first aspect of the present invention, the range of products is narrowed down to some extent according to the product classification, and at least one of the user classification and the user situation classification is input, whereby the details are further detailed. Moreover, it becomes possible to find a product suitable for the user.

When the product is a nursing care device, the user may be classified as a care receiver or a carer.
In addition, the situation of the user will include daily living situations and physical conditions.

In particular, the physical condition includes paralysis, contracture, bed sore, etc., and when these are selected, it is preferable to further input the part of the human body. Nursing care devices often have a form adapted to suit each body part. By specifying the body part, it is possible to search for products that are perfectly suited to the user's situation.

The second aspect of the invention is to limit the products in view of the needs of users.

The product information retrieval apparatus according to the second invention is a product data base in which a product code, a product classification code indicating a classification to which the product belongs, and product information are stored in advance for a plurality of products, and needs regarding the product. Displayed on the display screen of a display device or a display device for displaying a screen including a list of needs / product combination files for associating the product classification code with the product classification code or product classification code An input device for selecting at least one from the list of needs classifications that have been selected, and when there is a selection input for classification of needs from the input device, the needs / product combination file is searched and selected. Read the product code or product classification code that corresponds to the entered needs classification,
It is provided with search processing means for creating a list of products corresponding to the read product code or products included in the product classification code and displaying the list on the display device.

A product information retrieval method according to the second invention relates to a product data base in which a product code, a product classification code indicating a classification to which the product belongs, and product information are stored in advance for a plurality of products, and a product. A needs / product combination file that correlates the needs classification code with the product code or the product classification code is created in advance, and a screen including a list regarding the needs classification is displayed on the display device for searching product information. When a selection input for the needs category displayed on the display device is made from the input device, the needs / product combination file is searched and the product code or the product classification code corresponding to the needs classification selected and input is displayed. List of products that are read and that correspond to the read product code or products that are included in the product classification code Create those to be displayed on the display device.

According to the second invention, from the viewpoint of the user's needs, that is, what can be done about what, what kind of use, what kind of use, etc. Access to suitable products becomes possible. Therefore, information about a product that meets the needs of the user can be obtained even if there is little knowledge about the product.

The third invention is to search for products by making inferences based on the knowledge of experts.

A product information retrieval apparatus according to the third invention is a product data base in which a product code, a product classification code indicating a product classification to which the product belongs, and product information are stored in advance for a plurality of products. A display device for displaying a screen including a list of product categories for search, an input device for selecting at least one from the list of product categories displayed on the display screen of the display device, and a product for each product category A plurality of inference rule groups for product search created based on expert knowledge are preset, and an inference rule is selected according to the selection input for the product classification given from the input device, and the selected inference is performed. Inference means for executing inference processing according to rules to obtain inference results, inference results for classifying inference results by the inference means When an inference result / product combination file that correlates a class code with a product code or a product classification code and an inference result by the inference means are obtained, the inference result / product combination file is searched to find the inference result. A search processing unit is provided which reads out a product code or a product classification code corresponding to a classification, creates a list of products corresponding to the read product code or a product included in the product classification code, and displays the list on the display device. .

A product information retrieval method according to the third invention is a product data base in which a product code, a product classification code indicating a product classification to which the product belongs, and product information are stored in advance for a plurality of products. Based on the inference result classification code that classifies the results of the inference processing performed for searching, the inference result / product combination file that correlates the product code or the product classification code, and the expert knowledge of the product for each product classification A rule data base consisting of a plurality of inference rule groups for product search created in advance is created in advance, and a screen including a list regarding product classification for product information search is displayed on the display device, and the above display is performed. When there is a selection input regarding the displayed product classification of the device from the input device, it is inferred according to the above selection input. The inference result according to the selected inference rule, the inference result is obtained by executing the inference process according to the selected inference rule. The product code corresponding to the classification or the product classification code is read, a list of products corresponding to the read product code or the products included in the product classification code is created and displayed on the display device.

According to the third invention, the product can be found by performing the inference according to the rule based on the knowledge of the expert about the product.

The rule group based on the knowledge of the expert includes a rule group regarding product classification, a rule group regarding users, a rule group regarding products, and the like. As a group of rules regarding the user, particularly when the product is a nursing care device, a group of rules regarding the daily living condition of the care recipient (whether bedridden, can walk indoors, etc.), caregiver (presence or absence of caregiver, caregiver) Rules, etc., regarding whether a woman is a woman, etc.).

Since the product is searched based on the knowledge of such an expert, even a user who has little knowledge about the product can search for a product suitable for the user.

In one embodiment of the third invention,
The inference means executes a first inference process according to a rule group related to product classification and a rule group related to a user, selects a rule group related to a product based on a result of the first inference process, and selects the selected rule group. The second inference processing according to the above is executed.

In the first inference process, the inference process is executed by the rule group related to the product classification, so that the product classification can be narrowed down. In addition, in the first inference process, the inference process is executed by the rule group relating to the user, so that it is possible to select the product that suits the daily living situation of the user. Further, in the second inference process, the inference process is executed by the rule group related to the product, so that the product suitable for the user can be selected with respect to the product characteristics such as price and size of the product.

Preferably, the inference means causes the display device to display the question data generated in the inference process by the inference rule group, and executes the inference process using the answer data to the question data obtained through the input device. To do.

In the process of inference, when information such as the user's daily living condition, the user's physical condition, the caregiver's condition, etc. is needed, a question is asked to request these items from the user. Is displayed. The inference process proceeds as the user inputs an answer to this question.

The user can search for products suitable for the living condition, physical condition, etc. of the user simply by inputting the answer to the question. The user only has to answer the question, and does not need to create and enter the request for the product by himself. Therefore, the user can relatively easily perform the product search operation.

[0028] In a preferred embodiment of the third aspect of the present invention, the product information retrieving apparatus stores a product advice file for storing advice on a product, and an inference result for correlating the inference result classification code with the product advice. / Provides advice combination file,
The search processing means, when the inference result is obtained,
The inference result / advice combination file is searched, the advice corresponding to the classification of the inference result is read, and the read advice is displayed on the display device.

The displayed advice includes the outline of the searched product, precautions when using the searched product,
If multiple products are searched, it may include advice on what to select products from among them.

By displaying such advice, the user can easily confirm the outline of the searched product, whether the searched product is the desired product, and the like.

More preferably, different codes are assigned to the inference result classification codes respectively, or different codes are respectively assigned to the inference result classification codes corresponding to the inference results obtained by the inference means. Is assigned by the inference means, and the search processing means uses the inference result classification code for the product or the list of products corresponding to the inference result classification code obtained by the inference means and the advice corresponding to the inference result classification code. The product or the list of products and the advice are displayed on the display device with the same reference numerals assigned to.

In this preferred embodiment, the displayed product or list of products and the advice corresponding to the product or list of products are displayed with the same reference numerals. Therefore, even if a user displays multiple products or product lists and multiple advices,
It is possible to easily determine which product (or list of products) corresponds to which advice.

In any of the first, second and third inventions, when one or a plurality of products are selected and input from the list of products displayed on the display device, they are selected from the data base. The information about the selected product is read and displayed on the display device, or printed and output according to the request.

As typical examples of product information, the price, maker name, product description, etc. may be mentioned.

Preferably, an image showing a product, an image or an illustration showing how to use the product, advice on how to select, etc. are also displayed. It is preferable to prepare a plurality of images or illustrations when the form (shape, structure) of the product changes depending on the usage. These make it easier to understand the content of the product. It will be easier to understand if the product description (including usage) is output by voice.

It is also possible to combine the first invention and the third invention. Also, the second invention and the third invention can be combined.

[0037]

Description of Embodiments Embodiments (first embodiment, second embodiment and third embodiment) in which the present invention is applied to a care device information retrieval system will be described in detail below. Nursing equipment is a person with physical disadvantages such as bed, bedding, wheelchair, elevator, mat, simple toilet, deodorant absorbent, disposable diaper, simple bathtub, bath accessories, kitchen utensils, hand accessory. Machines and equipment used to assist or care for

[First Embodiment] FIG. 1 shows the overall construction of a care device information retrieval system according to the first embodiment.

The care device information retrieval system is composed of a computer system, the main part of which is the computer main body, that is, the processing device 10. The processing device 10 is composed of a computer programmed to execute input / output control processing, search processing, and the like, which will be described later.

As the input / output device of the processing device 10, C
RT display device 12, mouse 14, speaker 15 and printer
16 are connected. Under the control of the display control circuit 11, various screens are displayed on the display device 12 based on the display data given from the processing device 10 and held in the buffer memory 13, as will be described later. The operator inputs a request, an intention, or a selection result by clicking various buttons appearing on the display screen and desired ones among listed items with the mouse 14. The input data is held in the buffer memory 13 and taken into the processing device 10. The speaker 15 is for outputting the product description by voice. The printer 16 is used to print out product information that is a search result.

An image file 17, a sound file 18, a supplementary file 19 and a master file group 20 are connected to the processing device 10. These files are realized by a disk memory or a semiconductor memory. Image files 17
In advance, image data representing the appearance, usage state, etc. of each product is stored in advance. In the audio file 18, audio data for outputting the text of the product description by voice is stored in advance for each product. Supplementary file 19
Stores the data necessary for notifying the user how to use the product and how to select it by using illustrations, etc., for products that are difficult to understand only by image display based on image data and voice output based on audio data. Master file group
Twenty contains many master files described below.

In this embodiment, products (nursing care equipment) are roughly classified by product category, each product category is moderately classified by product item, and a specific product belongs to each product item. .

As shown in FIG. 2, the product category master has
Stored are preset product category codes and corresponding product category names (collection of character codes for display). The product category names include medical care bed related items, wheelchair related items, bath related items, meal related items, clothing items, and the like.

In the product item master, as shown in FIG.
A product classification code, a product item code, and a product item name (collection of character codes for display) are stored in association with each other. The same product category code is assigned to the product items belonging to the same product category. In FIG. 3, in particular, some of the product items belonging to the product classification code 02 are shown.

As shown in FIG. 4, the product master has, for each product, a product category code to which the product belongs, a product item code to which the product belongs, a product code, and a product name (character code for display. , A description of the product (collection of character codes for displaying or printing the description), the price of the product, the name of the manufacturer who manufactured the product, a user code to be described later, a daily life status code,
The physical condition code, data for accessing the storage location of the image file 17 in which the image data representing the product is stored (for example, address, etc.), and voice data representing the product description of the product by voice are stored. It stores data for accessing the storage location of the audio file 18 and data used for accessing the storage location of the supplement file 19 that stores supplement data for the product.

The information about the product stored in the product master can be narrowed down by the information about the user and the situation of the user. A user master (FIG. 5), a daily life situation master (FIG. 6) and a physical situation master (FIG. 7) are provided to define detailed search conditions from the user's point of view.

The user master stores the user code and the corresponding user. The user code 01 means that the user himself / herself uses the nursing care device, and the user code 02 means that the caregiver, not the user, uses the nursing care device.

The data stored in advance in the daily life situation master includes the daily life situation code of the user himself,
It is a code of a character corresponding to the daily life situation.

Further, the physical condition code of the user himself and the corresponding character code for displaying the physical condition are stored in the physical condition master.

In the columns of the user code, daily life status code and physical status code of the above-mentioned product master, for each product, a code suitable for use of the product, a daily life status and a physical status, respectively. (One product may correspond to two or more codes) is set in advance.

In order to enable searching from the viewpoint of needs, a needs classification master (FIG. 8), a needs master (FIG. 9) and a needs / commodity combination file (FIG. 10) are provided.

Various needs are classified according to needs. The needs classification master stores needs classifications and needs classification names (collection of character codes for display, etc.) in association with each other.

The needs master stores individual needs. For each need, a code of the needs classification to which the needs belong, a code of the needs, and a collection of the character codes expressing the needs (need name) are stored. Has been done.

The needs / commodity combination file links needs with products that satisfy the needs. Corresponding to each needs, the needs code, the needs classification code to which the needs belong, the code of the product satisfying the needs, the product classification code and the product item code to which the product belongs, and the product name are stored. One or more products correspond to one need.

11 and 12 show the operation of the processing device 10 in the care device information retrieval system, particularly the flow of the input / output control process and the retrieval process. Examples of typical display screens displayed on the display device 12 are shown in FIGS. 13 to 20.

First, the main menu screen as shown in FIG. 13 is displayed on the display device 12 (step 31). As can be seen from this figure, it is possible to perform a search based on product categories (product search) and a search based on needs (needs search).
When the product search is clicked by the mouse 14 and the OK button is pressed (clicked by the mouse), the process proceeds from step 33. When the needs search is clicked by the mouse 14 and the OK button is pressed, the process starts from step 51. (Step 32).

In the product search, first, a list of product category names (care device information product search screen) is displayed on the display device 12 (step 33). When one of these product category names is selected (step 34), a list of product items belonging to the selected product category is displayed (step 35). This state is shown in FIG. Wheelchair after product category is selected. In the product category list and the product item list, all the product categories and product items stored in the product category master and the product item master can be displayed. It is possible to move up and down on the display screen by pointing to ().

The operator selects one or more of the displayed product items (a plurality of items can be designated) with the mouse 14
By selecting and inputting with (step 36) and pressing the search button (clicking the search character on the display screen with the mouse 14) (step 37), the search in the product master is started (step 61).

In the selection input of product items (step 36), the intake button is pressed without pressing the search button (the character of intake on the display screen is the mouse
14) and (step 41), the intake screen (user situation selection screen) as shown in FIG. 16 is displayed on the display device 12 based on the data of the user master, the daily life situation master and the physical situation master. Is displayed (Step 4
2).

This is to give more detailed search conditions from the viewpoint of the situation of the user, in addition to the search conditions by specifying the product category and the product item described above. According to the display screen shown in FIG. 16, it is the user or the caregiver who uses the nursing care device, and the user's own daily living condition and physical condition select one or more from the list. Is input (step 43). After this, if the OK button is pressed, the original care device information product search screen is returned, and if the search button is pressed (step 44),
Proceed to search for the product master (step 61).

When the needs search is selected in step 32, a care device information needs search screen is displayed (step 51). This needs search is for selecting a device suitable for the user from the viewpoint of what the user wants when it is not known what device is suitable for the user.

On the care device information needs search screen, a list of needs categories is first displayed based on the needs category master data. When the operator selects and inputs one of these needs categories (step 52), a list of needs belonging to the selected needs category is displayed based on the contents of the needs master (step 53). FIG. 17 shows an example of the display screen when the move is selected from the need categories.

After that, one need (may select a plurality of needs) from the displayed list of needs is selected and input by the mouse 14 (step 54),
If the search button is pressed (step 55), the search using the needs / product combination file is started (step 6).
2).

It goes without saying that all the needs categories and needs stored in the needs category master and the needs master can also appear in the needs category and the list of needs. When the needs are selected over a plurality of needs categories, it is advisable to provide a selection status button as shown by the chain line in FIG. 17 in order to display a list of the selected needs.

When a search condition consisting of only product categories and product items is input (steps 33 to 37), the product code and product name having the product category code and product item code that have been selected and input in the product master. The product name is read out and displayed on the display screen (step 61). An example of this display is shown in FIG.

When data relating to the user's situation is input in addition to the product category and product item (step
33-36, 41-44), in the product master, of the products having the selectively input product category code and product item code, the products having the selectively input user code, daily life situation code, and physical condition code. The code and the product name are read, and the product name is displayed on the display device 12 as in the screen shown in FIG. 15 (step 61).

In the case of needs search (steps 51 to 5)
5) In the needs / product combination file, the code and product name of the product having the selectively input needs classification code and needs code are read out, and the product name is displayed on the display screen (step 62). An example of this display is shown in FIG.

By looking at the product name displayed on the display screen in this way, the operator can select an arbitrary product from among them using the mouse 14 (step 63). When the product is selected and the inquiry button is pressed, the product master is referenced again, the product name and price of the selected product are read, and the corresponding image data of the image file 17 is read and displayed on the display device 12. It is displayed (inquiry screen display) (step 64). An example of the inquiry screen is shown in FIG. When the print button is pressed (step 65), the display screen is printed by the printer 16 (step 66). When the return button is pressed (step 67), the product list is displayed again.

When the displayed image is clicked with the mouse in this inquiry screen display (step 6
8) The product master is referred to again, the price, maker name, and product description of the product are read and displayed, and the product description is displayed on the speaker 15 based on the audio data of the product stored in the audio file 18. Is output by voice (step 69). An example of this product description screen is shown in FIG. As shown in Fig. 20, since up to 4 types of image data are created in advance for one product,
Explains how to use the product and details. Supplemental data about the product is read from the supplemental file 19 as necessary, and additional information is displayed. If the back button is pressed (step 70), the display returns to the inquiry screen display.

At step 63, a plurality of products are selected,
When the inquiry button is pressed, inquiry screens for a plurality of products are displayed all at once. If an explanation request is made, explanation screens for a plurality of products are displayed at once. It is also possible to print these inquiry screens and explanation screens by designating the products or by using the printer 16 for all the products.

[Second Embodiment] An improved second embodiment will be described with reference to FIGS. 21 to 32. In the second embodiment, when the body condition such as paralysis, contracture, or bed sore is selected on the intake screen (user condition selection screen) (FIG. 16), the body parts of those persons are further input, and the user himself / herself More appropriate products are selected based on more detailed information about the physical condition of the (care recipient). The description of the same configuration as that of the first embodiment is omitted, and the points different from the first embodiment (especially added functions) will be described below.

21 and 22 show a product master. As can be seen from the comparison with FIG. 4, the product master additionally stores a Mahi code, a contraction code, and a bed slip code for each product.

FIG. 23 shows a physical condition master. Figure 7
As can be seen from the comparison with, the separate codes are attached to the paralysis and the paralysis. Therefore, each physical condition code after the bed slip code is deviated by one from the corresponding code shown in FIG.

FIG. 24 shows a Mahi master, FIG. 25 shows a contracting master, and FIG. 26 shows a bed sore master. These masters are newly added to the master file group 20. The Mahi master stores a Mahi code and a corresponding character code for displaying a part of the human body. Similarly, the contracture master and the bed sore master respectively store the contracture code and the bed sore code and the corresponding character code representing the part of the human body in advance. The Mahi code, contracture code, and bed sore code are associated with the display position of each part of the human body diagram displayed on the display device 12.

There are products that are suitable for a specific physical condition and products that are not. Conversely, a product may be suitable for any physical condition, or may be suitable only for a specific physical condition. For example, "Pad with beads"
Is right side of the body (right side), left side of the body (left side) and paralysis of the lower limbs (paralysis code 01, 02 and 08), hip contracture (contract code 06), and coccyx and rump bed sores (bed pressure codes 07, 08) ) Is suitable for. Therefore, in the product master (Figs. 21 and 22), "Pad with beads F
Corresponding to the product "2", the Mahi code is 01,0
2, 08, 06 as the contraction code, and 07, 08 as the bed slip code are stored. As described above, a code representing a paralysis site, a code representing a contracture site, and a bed pressure site that are suitable for the product corresponding to the product code are stored in advance in the product master.

FIG. 27 shows the flow of input / output control processing corresponding to FIG. Although not shown in FIG. 27,
The processing of steps 31 to 35 and 51 to 55 shown in FIG. 11 is applied as it is. In addition, the processing of steps 36, 37, 41 to 44 is also shown in FIG.
Is the same as that shown in. Steps 71 to 76 are added between steps 43 and 44.

After selecting and inputting the product item (step 36), when the intake button is pressed (step 41),
An intake screen as shown in FIG. 16 is displayed on the display device 12 (step 42). This intake screen (user situation selection screen) differs from that shown in FIG. 16 only in that the window showing the body situation shown in FIG. 23 reflects the body situation master.

Either or both of the list of users (the person and the caregiver) and one or more of the list of daily life situations are selected and input, and then "mahi" in the list of physical situations is clicked with the mouse. Then, the screen shown in FIG. 28 is displayed so as to overlap a part of the display screen (step 71).

The screen of FIG. 28 is for inputting the part of the paralysis, and is composed of a human body diagram (front) and a list of characters representing the part of the paralysis. Since the part of the paralysis is defined in a large area, not in a small part of the human body, contracture,
Unlike the case of bedsore, a list of characters that represent the part of the paralysis is displayed. When one of the lists is clicked with the mouse (step 72), the screen shown in FIG. 29 is displayed. The character clicked with the mouse is highlighted and the part of the human body diagram represented by the character is displayed in blinking. When the OK button is pressed, the selected parasite (mahi code representing the site) is added to the search conditions. By pressing the cancel button, the previously selected part can be canceled and another part can be newly selected.

When "contract" in the list of physical conditions is clicked on with the mouse, a screen as shown in FIG. 30 is displayed (step 73). This screen consists of a front human figure and a rear human figure. In any of these human body diagrams, if you select a contracted part and click it with the mouse, that part will blink as shown in FIG. 31 (step 7).
Four). FIG. 31 shows a state where the knee (contract code = 07) is clicked. When the OK button is pressed, the contraction code corresponding to the clicked part is added to the search condition.

Similarly, when "bed sore" is selected, a screen as shown in FIG. 32 is displayed, and a bed sore region can be selected on the human body diagram, and the selected region (the region is displayed). Bedsore code) is added to the search conditions (steps 75 and 76).

When any item other than paralysis, contracture, and bedsore is selected from the physical condition list, a code representing the selected physical condition is added to the search condition.

After these, if the search button is pressed,
Proceed to the search process shown in 12 (step 44).

When a Mahi code, contracture code, or bed sore code is entered (steps 72, 74, 76), the Mahi code entered in the product item in the selected product category in the product master search, It goes without saying that only the products for which the contraction code or the bedsore code are registered are extracted from the product master and displayed (step 61). The subsequent processing is the same as in the first embodiment.

As described above, since it is possible to input a specific part such as paralysis, contracture or bed sore, it is easy to select a product suitable for the input part.

[Third Embodiment] In the third embodiment, a search for finding a product by performing inference according to a rule created based on the knowledge of an expert in the product search and the needs search in the first embodiment ( Hereinafter, it is referred to as AI search).

(1) System Configuration FIG. 33 shows the overall configuration of the care device information retrieval system in the third embodiment. The same components as those of the nursing care device information retrieval system in the first embodiment shown in FIG. 1 are designated by the same reference numerals.

The difference from FIG. 1 is that the master file group 20 does not include a user master, a daily life situation master, and a physical situation master. Instead,
The master file group 20 includes an AI classification master, an AI / commodity combination file, and an advice file, which will be described in detail later. These masters and files are
It is used when performing an AI search which will be described in detail later.

Further, the point that the knowledge base 21 is connected to the processing device 10 is different from FIG. This knowledge base 21 is generally realized by a magnetic disk (or optical disk) device. Many inference rule groups (a part of which will be described later) are stored in advance in the knowledge base 21.

The product category master included in the master file group 20 is the same as that shown in FIG. 2, and the product item master is the same as that shown in FIG. The product master is partially different from that shown in FIG. 34 and 35 show a product master in the third embodiment. The product master shown in these figures does not include a user code, a daily life situation code, and a physical situation code as compared with those shown in FIG. The product search using these codes in the first embodiment can be performed in the AI search in the third embodiment. In this product master, products belonging to the product classification code 02 are introduced in a little more detail than in FIG.

The needs classification master is shown in FIG. 8, the needs master is shown in FIG. 9, and the needs / commodity combination file is the same as that shown in FIG.

In the AI classification master, as shown in FIG. 36,
An AI code (described later), an AI classification name (described later), and an advice code are stored in correspondence with each other.

In the AI / commodity combination file, as shown in FIG. 37, an AI code, a commodity classification code, a commodity item code and a commodity code are stored in correspondence with each other.

As shown in FIG. 38, the advice file, the advice name (collection of character codes for display) and the advice (collection of character codes for display) correspond to each other in the advice file. Remembered

(2) Search Processing FIGS. 39 to 43 are processing devices in the care device information search system.
11 is a flowchart showing a flow of processing of 10.

(2-1) Product Search Process First, an initial screen as shown in FIG. 59 is displayed on the display device 12 (step 81). On this initial screen, a list of product category names is displayed in the product category selection field at the upper left of the screen together with the display. When one of the product category names is selected from the list of product category names (here, the wheelchair is selected) (steps 82 and 83), the list of the product items belonging to the selected product category name is shown in FIG. Are displayed in the product item column as shown in (step 84). Furthermore, one or more of the product items are selected (here, the lightweight / compact type is selected) (NO in step 85, step 86),
When the search execution button is pressed (YES in step 87), the product name belonging to the selected product item is displayed in the product name column (step 88) (FIG. 60). The details of the processing of the processing device 10 performed in this product search are the same as those in the first embodiment, and therefore the detailed description is omitted here.

The operator can select one or more arbitrary product names from the displayed product names (step 89). When a product is selected and the inquiry button is pressed, the product master is referenced, and the product name, price, and corresponding image data of the image file 17 for the selected product are read. Then, these read data are displayed on the display device 12 (inquiry screen display) (step
90).

FIG. 67 shows an example of the inquiry screen. Here, if the return button is pressed (YES in step 91), the display returns to the product list display (FIG. 60). When the product description button is pressed (YES in step 92), the product master is referenced again, and the price, maker name, and product description of the product are read,
It is displayed on the display device 12 (display of product description screen) (step 93). Further, the product description can be output from the speaker 15 based on the audio data of the product stored in the audio file 18.

An example of the product explanation screen is shown in FIG. Except that the display form is slightly different from that of the first embodiment, the contents of the product image and product description displayed on this screen are as follows.
It is the same as that of the first embodiment. When the print button is pressed (YES in step 94), the display screen is printed by the printer 16 (step 95). When the return button is pressed (step 96), the screen returns to the inquiry screen (Fig. 67).

Similar to the first embodiment, a plurality of products can be selected in step 89, and the inquiry screens for the selected plurality of products can be displayed all at once. It is also possible to display the explanation screens for a plurality of products at once.

(2-2) Needs Retrieval Processing When the needs button in the upper part of the screen is pressed in the initial screen shown in FIG. 59 (step 82), the care device information needs retrieval screen (window) is displayed ( Step 5
1). FIG. 61 shows an example of a care device information needs search screen. The needs search screen is displayed at a position overlapping a part of the initial screen (window). Then, the operator can perform a needs search on this needs search screen. Step 51 to Step 55 and Step 62
The need search process of the first embodiment is similar to that of the first embodiment (see FIG.
Since it is the same as 12), its explanation is omitted here.
A list of product names searched by the needs search is displayed in the product name column of the initial screen (window).

(2-3) AI Search Process In the initial screen shown in FIG. 59, after one of the product category names is selected (step 83) and the product item is displayed (step 84), the AI button (screen When the upper left part of is pressed (YES in step 85), the AI search is started. A
The I search is performed as follows.

First, by pressing the AI button,
The AI search screen as shown in FIG. 62 is displayed (step
101). Then, when the inference start button is pressed (step 102), the AI search start processing inference instruction list causes the internal memory (RAM or the like) of the processing device 10 (hereinafter, "the internal memory of the processing device 10" to be simply referred to as the internal memory). Is created) (step 103). An example of the AI search start processing inference instruction list is shown in FIG. The inference instruction list is a collection of pointers for mainly linking the inference rule groups to be processed with each other and indicating the storage location of the inference result. The meaning of each pointer will be described later. The inference number # 100 is an inference number that identifies the AI search start process. In the AI search start process, which product category the rule group will be inferred from now on,
It is determined based on the selected product category.

Next, the AI search start processing rule group (rule group corresponding to the rule number # 100) is stored in the knowledge base 21.
Read from. Then, the AI search start rule list is expanded in the internal memory (step 104).

As shown in FIG. 46, a large number of inference rule groups corresponding to inference numbers are stored in advance in the knowledge base 21. These inference rule groups are created based on the knowledge of specialists (especially specialists regarding care equipment). The inference rule group corresponding to the inference number can be loaded into the internal memory using the inference number as a key.

FIG. 56 shows the AI search start rule list expanded on the internal memory. The rule group for inference includes j (here, the same as the number of product categories, j
= 17) rule is included. Each rule is A1, ...,
Aj. In the AI search start rule, a rule instruction list, a condition part list, and a consequent part list are created for each rule.

The inference rule is a so-called if, then rule, which includes a condition part (starting with if) and a consequent part (starting with then).
Composed of and. The conditional part may contain multiple conditional propositions, in which case each conditional proposition is connected by and (and). The conclusion may also include multiple conclusions.

The rule instruction list is an area for storing a pointer indicating the start address of the "previous rule", a pointer indicating the start address of the "next rule", and a description of the "condition part" of the rule (condition part list). And a pointer indicating the start address of the area (conclusion part list) that stores the description of the "conclusion part" of the rule. Since there is no rule before the first rule A1, there is N as a pointer indicating the first address of the previous rule in the instruction list of the rule A1.
ULL is stored. Since there is no rule next to the last rule Aj, in the instruction list of the rule Aj, the pointer indicating the start address of the next rule is NUL.
L is set. These "previous rule" pointers and "next rule" pointers link each rule instruction list with each other.

The condition part list stores data representing one or a plurality of conditional propositions of the condition part, and the consequent part list stores data representing one or a plurality of conclusions.

In addition, a fact list and a non-fact list are created in the internal memory, and data representing facts and non-facts are stored in these lists based on the product category selected in step 84, respectively.

For example, in the initial screen shown in FIG. 59,
It is assumed that the product category "wheelchair" has been selected. In this case, it is a fact that the product category "wheelchair" has been selected, so that "the selected product category is a wheelchair" is stored in the fact list. Product categories other than "wheelchairs" such as "medical bed related" and "bath related" are not selected, so "selected product category related to medical care bed" and "selected product category related to bath" are not It is a fact. Such non-fact items are stored in the non-fact list. When an item is stored in the fact list, the non-fact item to be stored in the non-fact list for this fact item is predetermined. The processing device 10 performs the data storage process of the non-fact list according to such a predetermined setting.

In the inference instruction list, the fact list pointer indicates the start address of the fact list,
The non-fact list pointer indicates the start address of the non-fact list. The rule list pointer indicates the top address of the first rule instruction list in the rule list. When creating the AI search start rule list, fact list, and non-fact list, the processor 10 sets each head address data as the pointer of the AI search start processing inference instruction list.

After the above processing, the inference processing is started (step 105). 44 and 45 are flowcharts showing the detailed flow of the inference process.

First, the inference processing is started for the first rule A1 instruction list indicated by the rule list pointer of the AI search start processing inference instruction list (step 131). This inference process is to check whether or not the same thing as the conditional proposition of the conditional part list designated by the "conditional part" pointer C111 of the rule A1 instruction list is in the fact list (step 132). If the conditional proposition is not in the fact list, then the processes from step 133 onward are performed (these processes will be described in detail in the explanation of other rule groups), and after step 141 to step 14
Moving to 2, the same processing is performed for the rule A2 indicated by the "next rule" pointer in the rule A1 instruction list. In this way, the condition part of each rule and the contents of the fact list are compared.

As described above, if the fact list includes the input item "selected product category is wheelchair", the determination in step 132 is YES in the process of rule A2. Therefore, the inference intermediate result list is created using the conclusion data stored in the consequent part list of the rule (step 139). Further, the contents of the consequent part (inference number # 500) are added to the fact list (step 139).

In the consequent part of the AI search start rule, an inference number indicating a rule group to be executed next is stored as a conclusion. An inference intermediate result list is created in an empty area of the internal memory, and the inference number of the consequent part is written in this inference intermediate result list. In addition, the head address (BB00 in this example) of the created inference intermediate result list is
It is written in the column of inference intermediate result list pointer of the inference instruction list for AI search start processing. An example of the inference intermediate result list and inference instruction list created in this way is shown in the figure.
Shown in 48.

Then, the inference process ends. In general, A
In the I search start rule processing, since NO is obtained in step 132 for all the rules, there is no abnormal termination (step 143) through step 141.

Returning to FIG. 41 again, when the inference is completed, the inference instruction list for all inferences to be executed in the future is created in the internal memory based on the created inference intermediate result list (FIG. 48) ( Step 106).

49 to 51 show examples of the created inference instruction list for the inference to be executed. In the above-described AI search start inference, only the inference number # 500 is stored in the inference intermediate result list. Therefore,
After the AI search start inference, only the inference instruction list in the daily life situation evaluation phase (the phase in which the general status of the care recipient is judged) is created in FIG.
Caregiver evaluation phase following the daily life situation evaluation phase,
The inference instruction list for the wheelchair evaluation phase 1 and the like is created as a result of the inference processing in the daily life situation evaluation phase and the inference processing subsequent thereto, as described later.

The address AA00 of the inference instruction list in the AI search start phase is stored in the "pointer of previous instruction list" in the inference instruction list in the daily life situation evaluation phase. The address AB00 of the inference instruction list created by the inference processing of the daily life situation evaluation phase is stored in the "pointer of the next instruction list" (at the present time (before the inference processing of the daily life situation evaluation phase), Since "the pointer of the next instruction list" is in an undecided state, NULL is stored). # 5 for the "inference number"
00 is stored.

When the inference instruction list is created, the inference instruction list for the daily life situation evaluation phase is accessed according to the pointer AA88 of the next instruction list in the inference instruction list for the AI search start phase and stored therein. The rule group of inference number # 500 is read in. A part of the read rule group is shown below.

[0122]

The rule list for these rules is shown in FIG.
It is created in the internal memory in the same manner as shown in (step 107). In addition, the fact list and non-fact list for the daily life situation evaluation phase are created in the internal memory. Then, the contents of the fact list of the AI search start phase (“selected product category is wheelchair” and inference number # 500) are copied to this fact list. The contents of the non-fact list of the AI search start phase ("the selected product category is related to medical care bed", etc.) are copied to the non-fact list (step 108). Further, the head address D800 of the fact list is stored in the fact list pointer of the inference instruction list in the everyday life situation evaluation phase.
The head address DC00 of the non-fact list is stored in the non-fact list pointer of the inference instruction list in the daily life situation evaluation phase (see FIG. 49).

Next, if the inference number is # 500 or more and #
It is judged whether or not it is 5000 or less (step 109).
If the inference number is within this range (in step 109
YES), the screen corresponding to the inference number is displayed on the display device 12 (step 110). This screen data (image data)
Are stored in advance in the knowledge base 21 corresponding to the inference number. The reasoning number for the everyday life evaluation phase is #
Since it is 500, the screen corresponding to the inference number # 500 is displayed. FIG. 63 shows a screen corresponding to the daily life situation evaluation phase (inference number # 500). The operator selects any one of the items displayed on the screen and presses the enter button. Here, “movable (indoor walking)” is selected.

When the confirm button is pressed, the selected item (“move (indoor walk)”) is added to the fact list of the daily life situation evaluation phase. All the items that are not selected (such as "all help by bedridden" and "mainly sitting in bed") are added to the non-fact list (step 111). As a result, the fact list stores "selected product category is wheelchair", inference number # 500, and "might be able to move in everyday life (indoor walking)".
In the non-fact list, "selected product category is medical bed related", "selected product category is bath related", "daily living condition is bedridden and all assistance", "daily living condition is mainly bed life""Keepsitting" is stored.

Then, inference processing is performed (step 11).
2). The inference process is performed in the same manner as the process of the AI search start phase described above. That is, the above rule B1
-Bj-Bk and the like are compared with the contents of the fact list (step 132 in FIG. 44). If all items in the conditional part are in the fact list (YES in step 132), the consequent part of the rule is copied to the inference intermediate result list and added to the fact list (step 139). When "Move (indoor walk)" is selected in Figure 63, rule B above
Since all the items of the condition part of k exist in the fact list, the inference numbers # 27300 and # 25400 of the consequent part are copied to the inference intermediate result list of the daily life situation evaluation phase and added to the fact list. Then, the inference process ends. 52 to 54 show the content of the inference instruction list and the content of the inference intermediate result list in each inference phase (the caregiver evaluation phase after the daily life situation evaluation phase, etc. will be described later). The inference intermediate result list pointer of the inference instruction list in the everyday life situation evaluation phase is set to the start address BB8 of the inference intermediate result list.
8 is stored.

Based on the contents of the inference intermediate result list, as shown in FIG. 49, the caregiver evaluation phase (inference number #
27300) and wheelchair evaluation phase 1 (inference number #
25400) inference instruction list is created (step
113 to 117 and step 106) (steps 113 to 116)
Will be described later).

In the caregiver evaluation phase, the operator is informed of the presence or absence of a caregiver and whether the caregiver is a woman or an elderly person. Then, the data input by the operator is used as a judgment material for selecting a wheelchair in the wheelchair evaluation phase 1 etc. following this phase. The wheelchair evaluation phase 1 is to inform the operator of the wheelchair usage place and the like. The data entered by the operator is
Used to select a suitable wheelchair for the cared person.

The inference instruction lists for the daily life situation phase, the caregiver evaluation phase, and the wheelchair evaluation phase 1 are as follows:
These are linked to each other by the "previous instruction list pointer" and the "next instruction list pointer". Since "the pointer of the next instruction list" of the inference instruction list in the wheelchair evaluation phase 1 is undecided at this stage, NULL is stored in this pointer. Then, # 27300 and # 25400 are stored in the inference numbers of the inference instruction lists in the caregiver evaluation phase and the wheelchair evaluation phase 1, respectively.

Subsequently, the rule group corresponding to the inference number # 27300 stored in the inference instruction list in the caregiver evaluation phase is read from the knowledge base 21. Then, a rule list is created (step 107). A part of the read rule group is shown below.

[0131]

Of the character strings prefixed with $, such as $ TYPE0000 included in the consequent part, the character string after $ (TYPE0000) is the AI classification name included in the AI classification master (FIG. 36). . As described in detail later, the product is searched based on the AI classification name obtained from the inference result.

Based on these rules, a rule list as shown in FIG. 57 is created in the internal memory (step 107). Then, the fact list and non-fact list of the caregiver evaluation inference instruction list are created in the internal memory.
Contents of the fact list and the non-fact list in the daily life situation evaluation phase are copied to the fact list and the non-fact list, respectively (step 108). Figure 58
Shows the contents of the fact list and non-fact list in the caregiver evaluation phase. These fact list and non-fact list include the fact list and non-fact (indicated by COPY1) obtained in the AI search start phase, respectively. Further, the fact list and the non-fact list (indicated by COPY2) obtained in the daily life situation evaluation phase are also included. The list indicated by "addition" in the fact list is added to this fact list in the inference processing of the caregiver evaluation phase to be described below.

Then, the inference numbers are # 500 and # 5000.
It is determined whether or not it is included in the range (step 109 in FIG. 42). The reasoning number for the caregiver evaluation phase is # 273.
00 (># 5000) (N at step 109)
O), the inference processing is immediately performed (step 112).

The inference processing will be described in detail with reference to the flowcharts of FIGS. 44 and 45, and the rule list of FIG. 57 and the fact list and non-fact list of FIG. 58.

First, inference is started for the first rule C1 (step 131). All the items in the condition list of rule C1 are compared with all the contents of the fact list to determine whether all the items in the condition list are in the fact list (step 132). Since "caregiver is a woman or elderly person" among the items of the illustrated condition section list of rule C1 is not in the fact list, the determination in step 132 is NO.

Since there is an item that is not in the fact list among all the items of the condition section list of rule C1, whether or not there is an item that is not in the fact list and is listed in the non-fact list. Is next examined (step 133).

If any one of the items in the condition list that was not in the fact list is in the non-fact list, then this condition is never satisfied. In this case,
Rules with that condition are no longer worth inferring. Therefore, if YES at step 133,
The process moves to the inference process for the next rule (step 141, then step 142). Looking at the item “caregiver is female or elderly”, this item is not included in the non-fact list. Therefore, step 133
Then, the determination result is NO, and the process proceeds to step 134.

For an item that is not included in either the fact list or the non-fact list, it is necessary to check whether the item is a fact or a non-fact. Therefore, in step 134, the operator is queried for items not included in either the fact list or the non-fact list. The question is asked by displaying a question screen as shown in FIG. 64 on the display device 12.

In response to this question, the operator selects and presses either the "Yes" (affirmative, YES) button or the "No" (negative, NO) button. If the pressed button is "yes", the item is added to the fact list (step 136). If the button pressed is no, then the item is added to the non-fact list (step 13).
7).

After that, the process returns to step 132 and the above-mentioned processing is repeated. If there are two or more items that are not included in the fact list or the non-fact list in the conditional part of one rule, the above question processing is
It will be repeated more than once (as many times as there are items in neither the fact list nor the non-fact list).

In this way, items (or items for which the operator has forgotten to enter) that are not included in the screen displayed in the AI search start phase, the daily life situation evaluation phase, etc. but exist in the condition section list ， The operator's answer will be obtained.

When the operator presses the “No” button on the question screen of FIG. 64, “caregiver is female or elderly” is added to the non-fact list. As a result, the determination result of step 132 is NO, and the determination result of step 133 is Y.
Each becomes ES, and the next rule (rule C2 ... Ci
…) Are processed.

If there is no applicable rule (YES in step 141) even though it is the last rule in the rule list (the rule whose "next rule" pointer is NULL), the inference process is abnormal. The process ends (error end) (step 143). In case of abnormal termination,
The same inference process is repeated again as shown in step 113 of FIG. This is because there is a possibility that there is an input error by the operator as the reason that the applicable rule does not exist. For this reason, the items added to the fact list and the non-fact list in the last inference are cleared (step
114). For example, when the reasoning in the caregiver evaluation phase is abnormal termination, the items except the items COPY1 and COPY2 in the fact list and non-fact list shown in FIG. 58 are cleared. After that, the process is repeated from step 109 (FIG. 42).

Also, the operator can re-execute the same inference with his / her will. When the "Return" button on the AI search screen shown in Fig. 62 is clicked, if the previous inference was successful, the previous inference is executed again (the fact item and non-fact item obtained by the previous inference are cleared. If the previous inference is abnormally terminated or the inference is being executed, the previous inference is executed again (the fact item or the non-fact item obtained by the immediately preceding inference or the inference being executed). , And the factual and non-factual items obtained by the preceding inference are cleared). When the "return" button is pressed twice in succession, it is more preferable to return to the previous inference and execute the re-inference processing. In this way, if there is an input error, the operator can press the "return" button to redo the inference.

Returning to the flowchart of FIG. 44 again, the processing of the rule C1 will be described. In the following, rule C
The explanation will be given under the assumption that the “Yes” button is pressed on the question screen 1 (FIG. 64).

When the "Yes" button is pressed on the question screen of rule C1 (FIG. 64), "caregiver is female or elderly" is added to the fact list. As a result, the determination in step 132 becomes YES, and the consequent part of the rule C1 is copied to the inference intermediate result list and added to the fact list (step 139). As a result, the fact list is
It will be something like 58.

Subsequently, the wheelchair evaluation phase 1 inference is performed. This inference rule includes, for example, the following.

[0149]

Also in the wheelchair evaluation phase 1, the same processing as in the above-mentioned caregiver evaluation phase is performed. That is, the items “use a wheelchair at home” and “use a wheelchair for going out” included in the condition part of the rule D1 are not included in the fact list or the non-fact list. Therefore,
First, as shown in FIG. 65, a question screen about “use wheelchair at home” is displayed on the display device 12. Assume that the operator has pressed the "Yes" button for this question screen. In this case, subsequently, a question screen about "use the wheelchair for going out" is displayed on the display device 12. The operator
Suppose you press the "No" button for this question. In this case, this rule is not satisfied, and the processing for the next rule D2 is performed. As a result of the processing of rule D1, "use wheelchair at home" is stored in the fact list, and "use wheelchair for going out" is stored in the non-fact list.

Subsequently, processing is performed for the rules D2, D3 .... Then, it is assumed that the process for the rule Di is started. The condition part of the rule Di is already satisfied by asking the operator a question and obtaining an answer as described above in the rule D1. As a result, the consequent part of the rule Di is stored in the inference intermediate result list and the fact list (FIG. 52). Then, the inference in the wheelchair evaluation phase 1 ends.

Next, the inference instruction list for the inference number # 25600 in the consequent part of the rule Di is created, and the inference processing of the wheelchair evaluation phase 2 is performed. The wheelchair evaluation phase 2 is for determining the inference number for inference such as the wheelchair evaluation phase 3 and the wheelchair evaluation phase 4 that follow this phase.

The following are examples of rules for the wheelchair evaluation phase 2.

[0154]

The item "use the wheelchair for going out" in the condition part of the rule E1 is already stored in the non-fact list in the above-mentioned wheelchair evaluation phase 1. Therefore, the rule E1 is not satisfied. All the items of the condition part of the rule E2 are included in the fact list. Therefore, the consequence of rule E2 is stored in the inference intermediate result list and fact list (FIG. 53). Then, the inference in the wheelchair evaluation phase 2 ends.

Next, inference numbers # 26200 (wheelchair evaluation phase 3), # 25800 (wheelchair evaluation phase 4), # 27300 (caregiver evaluation phase) and # 2.
An inference instruction list for 7400 (evaluation phase from the wheelchair to the lifting step eliminating device) is created.

In the wheelchair evaluation phase 3, the operator asks how to use the wheelchair in daily life. Wheelchair evaluation phase 4 is to find out the physical condition of the cared person. A wheelchair suitable for the cared person is selected based on the data input by the operator in these phases.

The caregiver evaluation phase (reason number #
27300) is included again because information about the presence or absence of a caregiver is necessary in the later step of evaluating the step difference eliminating device, so that the information regarding the caregiver is confirmed again. The rules and inference contents of this caregiver evaluation phase are the same as those described above, so a description of this phase is omitted below.

The evaluation phase from the wheelchair to the lifting / lowering step eliminating device is for determining the rule group relating to the lifting / lowering step eliminating device based on the inference result regarding the wheelchair.

Wheelchair Evaluation Phase 3 (Inference Number # 262
The following is an example of the rule of (00).

[0161]

In the processing of the rule F1, whether or not the operator is "to face a desk / table in a wheelchair" (question 1
The question is whether to "eat / work in a wheelchair" (Question 2) and "use the wheelchair for sunbathing / nap / break" (Question 3). Assume that the operator has selected “yes” for questions 1 and 2 and “no” for question 3 for these questions. In this case, the contents of the items of questions 1 and 2 are added to the fact list, and the contents of the item of question 3 are added to the non-fact list.

Since the condition part of the rule F1 is not satisfied,
The processing of the rules after the rule F2 is performed. Then, it is assumed that the condition part of the rule Fi is satisfied. As a result, in the inference intermediate result list, the consequent part of the rule Fi (A
The I classification name # TYPE0214) is stored (FIG. 53).

Then, the inference of the wheelchair evaluation phase 4 is performed. Wheelchair evaluation phase 4 (inference number # 25800)
An example of this rule is shown below.

[0165]

In the processing of rule G1, the operator is asked whether or not he or she may use a cane. It is assumed that the operator selects "Yes" to this question. In this case, since the condition part of the rule G1 is satisfied, the result part (AI classification name $ TYP) is included in the inference intermediate result list.
E0259) is stored (FIG. 53).

Next, the processing moves to the evaluation phase (inference number # 27400) from the wheelchair to the lifting step eliminating device. An example of the rules for this phase is shown below.

[0168]

When this rule is applied, the result part (# 2200) of this rule is added to the inference intermediate result list.
0) is stored (FIG. 54).

Subsequently, an inference instruction list of inference number # 22000 (elevation step resolution device evaluation phase) is created (FIG. 51). The elevator step eliminating device evaluation phase is based on the fact input by the operator in the wheelchair evaluation phase and selects the elevator step eliminating device suitable for the fact.

Inference processing is performed on the rule group corresponding to the inference number # 22000. An example of the rule is shown below.

[0172]

It is assumed that the rule Ik has been applied. In this case, the consequence of this rule is inference number # 11.
Is. This inference number is preset as a special number for ending all inferences. With reference to FIG. 42, if a rule having the inference number # 11 as the consequent part is applied, a new inference instruction list is not created and the inference process ends here. Even if an unprocessed inference instruction list has already been created, the inference instruction list is not processed and the inference ends midway.

As a case where all inference processing is completed,
In addition, the "pointer of the next instruction list" in the inference instruction list may be NULL (step 116).

After the completion of all the inference processing, FIG.
When the AI search execution button in the upper part of the screen shown in (1) is pressed (step 118), the AI classification master (FIG. 36) is searched (step 119). This search is performed as follows.

First, the AI classification name stored in the inference intermediate result list (FIG. 52 to FIG. 54) by the above inference processing (the character string beginning with $ and the character string after $ (TYP
E0000))) is extracted. One column of the extracted AI classification name is shown in FIG. Next, the AI classification master is searched for a list having the extracted AI classification name.

The AI classification names TYPE0000 and TYPE9992 shown in FIG. 55 do not exist in the AI classification master. Therefore, the AI code and the advice code corresponding to these classification names are not read. An AI classification name (dummy classification name) that does not exist in such an AI classification master may also be included in the consequent part of the rule.

The AI classification names TYPE0214 and TYPE0259 shown in FIG. 55 exist in the AI classification master.
Therefore, the AI code 21 corresponding to these classification names
4 and 259, and advice code adv2
13 and adv253 are read.

Referring again to FIG. 43, the AI / commodity combination file is searched for a list having the read AI code (step 120). That is, in the AI / commodity combination file, the product category code, product item code, and product code corresponding to the AI codes 214 and 259 are searched and read. Then, the product category code, the product item code, and the product name corresponding to the product code are read from the product master (FIG. 34) and displayed in the product name column as shown in FIG. 66 (step 120). The AI code is for associating the AI classification name with the product classification code, product item code, and product code in this way.

In the advice file,
Read advice codes adv213 and ad
The advice name and advice having v253 are read. Then, the read advice name and advice are displayed in the advice column as shown in FIG. 66 (step 121). The advice code thus associates the AI classification name with the advice name and the advice.

Codes such as ☆, ★, ○, ●, ◇, ◆ are associated with each AI code read from the AI classification master. Here, the AI code 214 is coded ☆
However, the symbol * is associated with the AI code 259. In displaying the product name, the AI corresponding to each product name
The code attached to the code is displayed at the beginning of the product name. That is, as shown in FIG. 66, AI code 2
Product name corresponding to 14 (Wheelchair desk type NA1 etc.)
The symbol ☆ is displayed at the beginning of. AI code 259
The symbol ★ is displayed at the beginning of the product name (wheelchair cane holder) corresponding to. Similarly, when displaying the advice name and the advice, the code attached to the AI code corresponding to each advice name and the advice is displayed at the beginning of the advice name (see also FIG. 66).

This allows the operator to easily associate the product name displayed in the product name column with the advice name and advice displayed in the advice column. For example, the product name "Wheelchair Desk Type N"
The symbol ☆ is attached to the table for wheelchairs from "A1". Therefore, the advice corresponding to these products corresponds to the advice "indoor wheelchair for bedridden people-care for females / elderly people" with the same symbol ☆. Since the product name "Wheelchair cane" is attached with the symbol ★, the advice of "Wheelchair option" with the same symbol ★ corresponds to it.

Since the processing of step 89 and the like following step 121 has been described above, the description thereof will be omitted here.

As described above, in the AI search, the condition of the cared person, the request for the wheelchair, etc. are displayed as selection items or questions. Then, the operator can perform a product search by selecting a selection item or answering a question. Therefore, even a person who has little knowledge about products (nursing equipment) can search for products suitable for the cared person. In addition, the search can be performed simply by selecting the selection item or inputting the answer to the question (pressing the button), so the operation is extremely simple. Therefore, even an operator unfamiliar with the search device can easily operate the search device.

It goes without saying that the product search of the second embodiment (product search in consideration of paralysis, contracture and bed sore) can be added to the third embodiment.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a product information search system in a first embodiment.

FIG. 2 shows a product category master.

FIG. 3 shows a product item master.

FIG. 4 shows a product master in the first embodiment.

FIG. 5 shows a user master.

FIG. 6 shows a daily life situation master.

FIG. 7 shows a physical condition master.

FIG. 8 shows a needs classification master.

FIG. 9 shows a needs master.

FIG. 10 shows a needs / commodity combination file.

FIG. 11 is a flowchart showing a flow of product information search processing in the first embodiment.

FIG. 12 is a flowchart showing the flow of a product information search process in the first embodiment.

FIG. 13 shows a main menu screen in the first embodiment.

FIG. 14 shows an example of a care device information product search screen in the first embodiment.

FIG. 15 shows an example of a care device information product search screen in the first embodiment.

FIG. 16 shows an example of a care device information product search screen in the first embodiment.

FIG. 17 shows an example of a care device information needs search screen in the first embodiment.

FIG. 18 shows an example of a care device information needs search screen in the first embodiment.

FIG. 19 shows an example of an inquiry screen in the first embodiment.

FIG. 20 shows an example of a product explanation screen in the first embodiment.

FIG. 21 shows an example of a product master in the second embodiment.

FIG. 22 shows an example of a product master according to the second embodiment.

FIG. 23 shows a physical condition master in the second embodiment.

FIG. 24 shows a Mahi master in the second embodiment.

FIG. 25 shows a contracture master in the second embodiment.

FIG. 26 shows a bed sore master in the second embodiment.

FIG. 27 is a flowchart showing a part of the flow of product information search processing in the second embodiment.

FIG. 28 shows an example of a display screen when Mahi is selected.

FIG. 29 shows an example of a display screen when one of the Mahilists is selected.

FIG. 30 shows an example of a display screen when the contract is selected.

FIG. 31 shows an example of a display screen when a contracture site is selected.

FIG. 32 shows an example of a display screen when a floor slip is selected.

FIG. 33 is a block diagram showing the configuration of a product information search system in the third example.

FIG. 34 shows a product master according to the third embodiment.

FIG. 35 shows a product master according to the third embodiment.

FIG. 36 shows an AI classification master.

FIG. 37 shows an AI / commodity combination file.

FIG. 38 shows an advice file.

FIG. 39 is a flowchart showing the flow of product information search processing in the third embodiment.

FIG. 40 is a flowchart showing a flow of product information search processing in the third embodiment.

FIG. 41 is a flowchart showing a flow of product information search processing in the third embodiment.

FIG. 42 is a flowchart showing the flow of product information search processing in the third embodiment.

FIG. 43 is a flowchart showing the flow of product information search processing in the third embodiment.

FIG. 44 is a flowchart showing the flow of inference processing.

FIG. 45 is a flowchart showing the flow of inference processing.

FIG. 46 shows the structure of a knowledge base.

FIG. 47 shows a configuration example of a created inference instruction list for AI search start processing.

FIG. 48 shows an example of an inference intermediate result list and an inference instruction list created in the AI search start process.

FIG. 49 shows an example of an inference instruction list created in AI search.

FIG. 50 shows an example of an inference instruction list created in AI search.

FIG. 51 shows an example of an inference instruction list created in AI search.

FIG. 52 shows an inference instruction list and an inference intermediate result list when all inferences are completed.

FIG. 53 shows an inference instruction list and an inference intermediate result list when all inferences are completed.

FIG. 54 shows an inference instruction list and an inference intermediate result list when all inferences are completed.

FIG. 55 shows an AI classification name list.

FIG. 56 shows an AI search start inference instruction list, an AI search start rule list, a fact list, and a non-fact list expanded in the internal memory of the processing device.

FIG. 57 shows an example of a caregiver evaluation rule list expanded in the internal memory of the processing device.

FIG. 58 shows an example of the fact list and non-fact list used in the inference of the caregiver evaluation phase.

FIG. 59 shows an example of an initial screen in the third embodiment.

FIG. 60 shows an example of a care device information product search screen in the third embodiment.

FIG. 61 shows an example of a care device information needs search screen in the third embodiment.

FIG. 62 shows an example of an AI search screen.

FIG. 63 shows an example of an inquiry screen for everyday life situations.

FIG. 64 shows an example of a question screen.

FIG. 65 shows an example of a question screen.

FIG. 66 shows an example of an AI search screen after the search processing is completed.

FIG. 67 shows an example of an inquiry screen in the third embodiment.

FIG. 68 shows an example of a product explanation screen in the third embodiment.

[Explanation of symbols]

 10 Processor 12 Display 14 Mouse 20 Master files

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Mihoko Sano 44-7 Kamata, Shizuoka City, Shizuoka Prefecture (72) Inventor Michiko Yamada 439-2, Kojo, Shizuoka City, Shizuoka Prefecture Sun Heights No. 301

Claims (19)

[Claims] 1. In association with each of a plurality of products, the product classification to which the product belongs, the classification of the user who is suitable for using the product, and the classification regarding the situation of the user who matches the use of the product A database for storing at least one of the above and product information in advance, a display device for displaying a plurality of types of screens including a list for searching product information, and a list displayed on the display screen of the display device. An input device for selecting at least one of
At least one of a first process for displaying a screen for inputting a product category on the display device and waiting for a selection input from the input device, and a user category and a category relating to a user situation Input control means for displaying at least a first process of a second process of displaying a screen for inputting the input on the display device and waiting for a selection input from the input device, and a product from the input device. When there is a selection input for classification,
The base is searched, a list of products belonging to the selected and input product classification is created and displayed on the display device, and the input device selects and inputs the user classification and the usage status classification from the input device. When at least either one of the selection inputs is made, the database is searched, and at least one of the user classification and the user situation classification of the products belonging to the selected and input product classification. 2. A product information search device, comprising search processing means for extracting products applicable to the selection input of and displaying the list on the display device. 2. A product data base in which a product code, a product classification code indicating a classification to which the product belongs, and product information are stored in advance for a plurality of products, a product need code, and a product code or products. A needs / product combination file that correlates with a classification code, a display device that displays a screen including a list of needs classification for product information retrieval, and a list of needs classification that is displayed on the display screen of the display device. An input device for selecting at least one of the above, and when there is a selection input regarding the classification of needs from the input device, the above needs / commodity combination file is searched to correspond to the classification of the selected input The product code or product classification code is read, and the product or product classification code corresponding to the read product code A product information search device comprising a search processing means for creating a list of products included in the display device and displaying the list on the display device. 3. A product data base in which a product code, a product classification code indicating a product classification to which the product belongs, and product information are stored in advance for a plurality of products, and a list related to the product classification for searching product information. A display device for displaying the screen including the input device, an input device for selecting at least one from the list related to the product category displayed on the display screen of the display device, and a product created for each product category based on expert knowledge of the product. A plurality of inference rule groups for product retrieval are set in advance, the inference rule is selected according to the selection input for the product classification given from the input device, and the inference processing according to the selected inference rule is executed. Inference means for obtaining inference results, inference result classification code for classifying inference results by the inference means, and product code or product When the inference result / commodity combination file that correlates with the class code and the inference result by the inference means are obtained, the inference result / product combination file is searched and the product code or product corresponding to the classification of the inference result Retrieval processing means for reading the classification code, creating a list of products corresponding to the read product code or products included in the product classification code, and displaying the list on the display device.
Product information search device equipped with. 4. The inference rule group set in the inference means includes a rule group related to product classification, a rule group related to a user and a rule group related to a product, and the inference means includes a rule group related to product classification and a rule group related to product classification. A method of executing a first inference process according to a rule group related to a user, selecting a rule group related to a product based on a result of the first inference process, and executing a second inference process according to the selected rule group The product information search device according to claim 3, wherein 5. A product advice file for storing advice on a product, and an inference result / advice combination file for correlating the inference result classification code with the advice on the product, wherein the search processing means includes the inference result. 5. When the above is obtained, the inference result / advice combination file is searched to read the advice corresponding to the classification of the inference result, and the read advice is displayed on the display device.
Item information retrieval device described in. 6. The inference result classification codes are assigned different codes in advance, or the inference result classification codes corresponding to the inference results obtained by the inference means have different codes. Assigned by the inference means, the search processing means is
The product or the product corresponding to the inference result classification code obtained by the inference means and the advice corresponding to the inference result classification code are attached with the same symbols assigned to the inference result classification code. 6. The product information search device according to claim 5, wherein the list and advice are displayed on the display device. 7. The inference means displays the question data generated in the inference processing by the inference rule group on the display device, and uses the answer data to the question data obtained through the input device to enter the inference rule group. According to claim 3, which executes the inference processing.
Item 1. The product information search device according to any one of 1. 8. The search processing means searches the data base to select the corresponding product when at least one product in the product list displayed on the display device is selected from the input device. 8. The product information search device according to claim 1, wherein the product information about the item is read out and displayed on the display device. 9. The image processing apparatus further comprises an image file in which image data representing a product is stored in association with the product, wherein the search processing means reads out image data of the selected product from the image file and displays the product information on the display device together with the product information. The product information search device according to claim 8, which is to be displayed. 10. The data base stores in advance a classification relating to a physical condition and a classification relating to a body part as one of the classifications relating to a user's condition, and the input control means, in the second processing, A screen is displayed on the display device for inputting a classification regarding a physical condition, and a screen is displayed on the display device for inputting a classification regarding a body part when a specific one of the classifications regarding the physical condition is input. When the search processing means makes a selection input for the classification regarding the physical condition and the classification regarding the body part,
The product information search device according to claim 1, wherein the selected and inputted classification concerning the physical condition and classification concerning the body part are added to the search condition to search the database. 11. The product information search device according to claim 10, wherein the screen for inputting the classification regarding the body part includes a human body figure. 12. A product information search device for performing a product classification search for searching for a product based on a product classification to which the product belongs and a needs search for searching for a product based on needs related to the product. In relation to each, the product code, the product classification code indicating the product classification to which the product belongs, the classification of the user who is suitable for using the product, and the classification of the situation of the user who matches the use of the product A database that stores at least one of the above items and product information in advance, a needs / product combination file that correlates product needs classification codes with product codes or product classification codes, and a list for searching product information A display device for displaying a plurality of types of screens including, and at least from the list shown on the display screen of the display device. An input device for selecting one,
First input control means for displaying a screen for selecting one of the product category search and the needs search on the display device and waiting for selection input from the input device; A first process for displaying a screen for inputting a product classification on the display device when there is a selection input and waiting for a selection input from the input device, and a user classification and a user situation A second screen for displaying at least one of the classifications on the display device and performing at least a first process of waiting for a selection input from the input device. When a selection input for the product category is made from the input control means and the input device, the database is searched and a list of products belonging to the selected product category is created. When the selection input for the product classification and the selection input for at least one of the user classification and the user situation classification are input from the input device, the database is displayed. Of the products belonging to the selected and input product classification, the products to which the selection input of at least one of the user classification and the user situation classification is applied are extracted, and the list is displayed on the display device. First search processing means, when a selection input for a needs search is made from the input device, a screen including a list relating to needs classification is displayed on the display device to input the needs classification, and the input is performed. The third input control means for waiting for a selection input from the device, and the above-mentioned when there is a selection input for classification of needs from the input device, The product code or product classification code corresponding to the needs classification that has been selected and input is retrieved and a list of products corresponding to the read product code or products included in the product classification code is created. And a second search processing means for displaying on the display device. 13. A product information search device for performing a needs search for searching for products based on needs related to products and an inference search for searching for products according to a plurality of inference rules created on the basis of expert knowledge. For a plurality of products, a product data base in which a product code, a product classification code indicating the product classification to which the product belongs, and product information are stored in advance, a product need code, and a product code or product classification. A need / product combination file for associating codes with each other, an inference result classification code for classifying inference results, and an inference result / product combination file for associating a product code or a product classification code, and product classification for product information retrieval Display device for displaying multiple types of screens including lists, displayed on the display screen of the above display device An input device for selecting at least one from the list, and a screen for selecting either the needs search or the inference search are displayed on the display device and wait for selection input from the input device First
Input control means, when a selection input for a needs search is made from the input device, a screen including a list relating to needs classification is displayed on the display device so that the needs classification can be input, The second input control means for waiting for the selection input of the item, when there is a selection input regarding the classification of needs from the input device, the above-mentioned needs / product combination file is searched, and the product corresponding to the selected classification of the needs is input. First retrieval processing means for reading out a code or a product classification code, creating a list of products corresponding to the read product code or products included in the product classification code, and displaying the list on the display device, related to product classification from the input device When there is a selection input and there is a selection input for inference search, it is based on the expert knowledge about the product for each product classification. A plurality of created inference rule groups for product search are set in advance, the inference rule is selected according to the selection input for the product classification given from the input device, and the inference processing is executed according to the selected inference rule. And a product included in the product classification code corresponding to the classification of the inference result by searching the inference result / product combination file when the inference result by the inference device is obtained. Product information retrieving apparatus including second retrieving processing means for creating a list of the above and displaying it on the display device. 14. In association with each of a plurality of products,
At least one of a product category to which the product belongs, a user's classification suitable for the use of the product, and a user's situation matching the use of the product, and
The first process of storing the product information and the product information in the memory in advance, displaying a screen for inputting the product classification on the display device, and waiting for selection input from the input device; At least one of a second process of displaying a screen for inputting at least one of the categories relating to the situation on the display device and waiting for a selection input from the input device, and executing the input device. When there is a selection input for the product classification from, the above memory is searched, a list of products belonging to the selected and input product classification is created and displayed on the display device, and the product classification is performed from the input device. When there is a selection input of and a selection input of at least one of the user classification and the usage status classification, the above-mentioned memory is searched to belong to the selected and input product classification. Of products, and displays the list on the display device extracts the product applies to at least one select input of the user classification and user status classification, product information retrieval methods. 15. A screen for preliminarily storing in memory a category relating to a physical condition and a category relating to a body part as one of the categories relating to a user's condition, and inputting the category relating to a physical condition in the second processing. Is displayed on the display device, and when a specific one of the classifications relating to the physical condition is input, a screen for inputting the classification relating to the body part is displayed on the display device to wait for the selection input, and the body is input from the input device. 15. The memory according to claim 14, wherein when the selection regarding the classification regarding the situation and the classification regarding the body part is made, the selected and inputted classification regarding the body condition and the classification regarding the body part are added to the search condition to search the memory. Product information search method. 16. A product code for a plurality of products,
A product database that stores in advance the product classification code that represents the classification to which the product belongs, and a product needs database that stores the product information base and the product need code that correlates the product code or the product classification code. When a screen containing a list of needs categories for searching product information is displayed in advance on the display device and the selection input for the needs category displayed on the display device is made from the input device, The above-mentioned needs / commodity combination file is searched to read out the product code or product classification code corresponding to the classification of the selected needs, and create a list of products corresponding to the read product code or products included in the product classification code. A method for retrieving product information, which is displayed on the above display device. 17. The product code and the product code for a plurality of products,
A product classification code indicating the product classification to which the product belongs,
Product data base that stores product information in advance,
An inference result classification code that classifies the results of the inference processing performed for product information retrieval, and an inference result / product combination file that correlates the product code or the product classification code, and the expert knowledge of the product for each product classification Create a rule database that consists of multiple inference rule groups for product search based on the following, and display a screen containing a list of product classifications on the display device for product information search, When there is a selection input regarding the displayed product classification on the display device from the input device, an inference rule is selected according to the selection input, an inference process is executed according to the selected inference rule, and an inference result is obtained. When an inference result is obtained by inference processing, the above inference result / commodity combination file is searched to classify the inference result. Product Code or read a product classification code, and create a list of items to be included in the product or product classification code corresponding to the read commodity code is displayed on the display device, the product information search how. 18. A product information search method for performing a product classification search for searching for a product based on a product classification to which the product belongs and a needs search for searching for a product based on needs related to the product, the method comprising: In relation to each of the
A product code, a product classification code indicating a product classification to which the product belongs, and / or at least one of a user's classification suitable for the use of the product and a user status matching the use of the product. , Product information is stored in the memory in advance, and a needs / product combination file that correlates the product needs classification code with the product code or product classification code is created in advance, and product classification search and needs search When a selection input for the product category search is made from the input device, a process for displaying a screen for selecting one of the items 1 and 2 on the display device and waiting for the selection input from the input device is executed. A first process of displaying a screen for inputting a classification on the display device and waiting for selection input from the input device, and And at least a first process of a second process of displaying a screen for inputting at least one of the categories relating to the situation of the user on the display device and waiting for selection input from the input device And when there is a selection input for the product classification from the input device, the memory is searched, a list of products belonging to the selected and input product classification is created and displayed on the display device, When there is a selection input from the input device regarding the product classification and a selection input of at least one of the user classification and the user situation classification, the memory is searched and the selected product classification is input. Out of the products belonging to, the products that meet the selection input of at least one of the user classification and the user situation classification are extracted and the list is displayed on the display device. To be displayed on,
When there is a selection input for the needs search from the input device, a screen including a list regarding the needs classification is displayed on the display device in order to input the needs classification and waits for the selection input from the input device. Execute the process,
When there is a selection input for the needs classification from the input device, the above-mentioned needs / product combination file is searched to read the product code or product classification code corresponding to the selected and input needs classification, and the read product code A product information search method for creating a list of products corresponding to the item or products included in the product classification code and displaying the list on the display device. 19. A product information search method for performing a needs search for searching for products based on needs related to products and an inference search for searching for products according to a plurality of inference rules created on the basis of expert knowledge. For a plurality of products, a product code that stores a product code, a product classification code that indicates the product classification to which the product belongs, a product data base that stores product information, a product needs classification code, and a product code or a product classification code. Needs / product combination files that correlate with each other, and inference result / product combination files that correlate inference result classification codes that classify inference results with product codes or product classification codes, and search for product information A screen containing a list of product categories for When a selection input for the needs search is made from the input device, a process for displaying a screen for selecting one of the above on the display device and waiting for the selection input from the input device is performed, and the needs are classified. A screen including a list regarding needs classification is displayed on the display device to input the input, and a process of waiting for selection input from the input device is executed, and the selection input about the classification of needs is made from the input device. At this time, the above-mentioned needs / commodity combination file is searched and the product code or product classification code corresponding to the classification of the needs that has been selected and input is read, and the product corresponding to the read product code or the list of products included in the product classification code is read. Create and display it on the display device, there is a selection input regarding the product classification from the input device, and a selection input regarding the inference search. When there,
According to the selection input for the product classification given from the input device among a plurality of inference rule groups for product search, which are created based on the expert knowledge about the product for each product classification set in advance. Select an inference rule, execute the inference process according to the selected inference rule to obtain the inference result, and when the above inference result is obtained,
A product information search method for searching the inference result / product combination file to create a list of products corresponding to the classification of the inference result or products included in the product classification code and displaying the list on the display device.