Estimation device, estimation method, and estimation program

Abstract

To provide an estimation device capable of easily grasping abnormal estimation.SOLUTION: An estimation device includes: an estimation model generation unit for generating an estimation model using a neural network with data directly related to calculation of estimated unit price of an estimate response, data not directly related to calculation of the estimated unit price of the estimation response, and the estimated unit price of the estimation response being learning data; and an estimation unit for estimating a unit price of a product / component using the estimation model generated by the estimation model generation unit.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] Embodiments of the present invention relate to an estimate estimation device, an estimate estimation method, and an estimate estimation program.

Background Art

[0002] Supply chain management (SCM) systems are widely used. A supply chain is a concept that views processes such as parts procurement as a single supply chain. In a supply chain, the relationship between a purchaser (buyer) and a seller (supplier) is defined. An SCM system manages the supply of products such as parts in a supply chain.

[0003] When a buyer makes an estimate request for parts or the like to a supplier using an SCM system, the supplier uses the SCM system to send an estimate reply to the buyer.

[0004] However, when a buyer obtains a large number of estimates from a supplier all at once, it is not possible to thoroughly examine all the estimates, and there is a possibility of overlooking abnormal estimates.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] An embodiment aims to provide an estimate estimation device, an estimate estimation method, and an estimate estimation program that can easily grasp abnormal estimates.

Means for Solving the Problems

[0007] The quotation estimation device according to the embodiment uses, as learning data, data directly related to the calculation of the unit price of the quotation response, data not directly related to the calculation of the unit price of the quotation response, and the unit price of the quotation response, and generates an estimation model using a neural network. It has an estimation model generation unit and an estimation unit that estimates the unit price of a product / part using the estimation model generated by the estimation model generation unit.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Embodiments for Carrying Out the Invention

[0009] Hereinafter, embodiments will be described in detail with reference to the drawings. FIG. 1 is a configuration diagram showing an example of the configuration of a supply chain management system according to an embodiment. The supply chain management system 1 includes a server 2 as an estimate management device, a plurality of terminals 3 for a plurality of buyers, a plurality of terminals 4 for a plurality of suppliers, and a communication network 5.

[0010] The server 2 and the plurality of terminals 3, 4 are communicably connected to each other via the network 5. The network 5 is the Internet here. Each of the terminals 3, 4 has an input device and a display device 3a. The input device is a keyboard, a mouse, etc., and the display device 3a is a monitor. In FIG. 1, only one terminal 3 is shown with the display device 3a.

[0011] The server 2 has a processor 11 and a storage device 12. The storage device 12 stores various software programs for the supply chain management system described later and stores various information.

[0012] When the processor 11 reads out and executes necessary programs such as an estimate estimation program from the storage device 12, various functions of the supply chain management system and functions for estimate assessment and display of estimate results described later are realized.

[0013] The supply chain management system 1 is a system for a plurality of buyers and a plurality of suppliers to manage the supply chain of products such as parts to be purchased or supplied by each. And, as will be described later, the server 2 can transmit supply chain risk information to the terminals 3, 4 via the network 5 in response to requests from the terminals 3, 4 via the network 5.

[0014] Each buyer can manage its own supply chain by using the supply chain management system 1. Therefore, each buyer can access the server 2 from its own terminal 3 to register its own supply chain information. Each buyer may register all of its own supply chain information, or each supplier may register its own supply chain information, or the primary supplier may register the information of secondary and subsequent suppliers.

[0015] In addition, each buyer can access the server 2 from its own terminal 3 to register information about each supplier (basic supplier information including supplier risk information RI).

[0016] When each buyer and each supplier access the server 2 via the network 5 using their own terminals 3, 4, they can input, display, and output data on a browser-based screen. Note that access from each terminal 3, 4 to the server 2 is possible after various authentications.

[0017] Figure 2 is a schematic diagram for explaining the supply chain. Figure 2 shows an example of the supply chain for parts X and Y. Figure 2 shows a case where a certain buyer purchases part X from supplier A and part Y from supplier D to manufacture and sell its own products. In this case, suppliers A and D are the primary suppliers of parts X and Y, respectively.

[0018] However, in order to manufacture and sell parts X and Y respectively, suppliers A and D purchase parts x1 and y1 from suppliers B and E. Furthermore, in order to manufacture and sell parts x1 and y1 respectively, suppliers B and E purchase parts x2 and y2 from suppliers C and F. Furthermore, in order to manufacture and sell parts x2 and y2 respectively, suppliers C and F purchase parts x3 and y3 from other suppliers. That is, the supply chain includes multiple layers. Therefore, each supplier can also be a buyer.

[0019] Note that, as shown by the dotted line in FIG. 2, the supply chain may branch to a plurality of suppliers. For example, the buyer may receive component supplies from two suppliers with respect to component X, and supplier D may also receive component supplies from two suppliers.

[0020] FIG. 3 is a block diagram showing the configuration of programs and data executed in the server. Server 2 has a software program for managing the operations between the buyer and the supplier by SRM (Supplier Relationship Management). The supply chain management system 1 stores the software program for SRM in the storage device 12.

[0021] In FIG. 3, a procurement analysis unit PA, a BCP (Business Continuity Planning) management unit, a general document exchange unit, an electronic quotation unit, a workflow management unit, a document management unit, a business partner basic information unit, an estimation model generation unit MG, and an estimation unit EP for SRM in supply chain management are shown, and it also has various other processing units. Furthermore, it also has a portal program for access management to the supply chain management system 1. In FIG. 3, the buyer portal is a processing unit for the buyer to access server 2 and performs various authentication processes. When properly authenticated, the buyer can use server 2. The supplier portal is a processing unit for the supplier to access server 2 and performs various authentication processes. When properly authenticated, the supplier can use server 2.

[0022] The procurement analysis unit PA, the BCP management unit, etc. are stored in the storage device 12 as software programs and are read out when necessary and can be executed by the processor 11.

[0023] The Procurement Analysis Department PA conducts an analysis of procurement from suppliers for each buyer. The Procurement Analysis Department PA can generate analysis reports such as the number of quotations, quotation results, and various evaluations for each supplier related to itself based on, for example, the quotation history information ERI.

[0024] The BCP Management Department collects the latitude / longitude information of the bases of each supplier constituting the supply chain, and identifies the bases existing within the affected area in the event of an emergency such as a disaster.

[0025] The General Document Exchange Department conducts document exchange between buyers and suppliers.

[0026] The Electronic Quotation Department manages requests for quotations to suppliers and responses to quotations from suppliers. A buyer can use the Electronic Quotation Department to request quotations for products / parts from one or more suppliers. The quotation request is sent to the one or more suppliers, and the suppliers can send quotation responses to the buyer.

[0027] The Workflow Management Department manages the workflows of various processes between buyers and suppliers.

[0028] The Document Management Department manages documents created by buyers, documents received from suppliers, etc.

[0029] The Counterparty Basic Information Department registers and manages basic information (such as capital, president, counterparty risk information RI, etc.) about primary suppliers, etc. as counterparty information BAI in the storage device 12. If a buyer cannot register information about all suppliers, the primary supplier may be able to register basic information about secondary and lower-level suppliers.

[0030] The supplier risk information RI included in the customer information includes a quantified risk level r regarding disaster prevention measures and the like in the supplier. The risk level r is a value quantified based on a predetermined evaluation criterion. For example, the risk level r has four levels. Level 1 has the lowest risk level r. For example, it is a case where disaster prevention measures are sufficient and product inventory is sufficiently secured. Level 4 has the highest risk level r. For example, it is a case where disaster prevention measures are insufficient. For example, the buyer interviews each supplier and determines the level based on a predetermined evaluation criterion. For example, the risk level r of secondary or lower suppliers is determined by the primary supplier.

[0031] The estimation model generation unit MG reads the estimation responses such as the estimation history information ERI and learns by deep learning to generate an estimation model. The estimation model generation unit MG performs learning by deep learning for each product / part and type of processing, such as resin molding processing, cutting processing, and assembly processing, to generate an estimation model. The estimation model generation unit MG stores the generated plurality of estimation models in the storage device 12 as an estimation model group M.

[0032] When the supplier sends a new estimation response based on an estimation request from the buyer, the estimation unit EP inputs the estimation response into the corresponding product / part estimation model of the estimation model group M to estimate the unit price of the product / part. The estimation unit EP can calculate the difference between the estimated unit price described in the estimation response and the estimated unit price estimated by AI, and display the calculation result on the display device 3a.

[0033] In addition to software programs, various information is also stored in the storage device 12 of the server 2. In FIG. 3, only the customer information BAI, the supply chain information SCI, the estimation history information ERI, and the estimation model group M are shown.

[0034] The customer information BAI is basic information (such as capital, president, customer risk information RI, etc.) about the primary supplier and the like as described above.

[0035] Supply chain information SCI is the location information of primary suppliers, secondary suppliers, tertiary suppliers, etc. for each incoming item (parts, products, etc.) in the supply chain. The location information of primary suppliers, secondary suppliers, tertiary suppliers, etc. is registered. Each location information includes a location name and location information. Each location is a place where there is a factory such as a subcontractor. The location information includes latitude / longitude information. The location information about secondary suppliers, tertiary suppliers, etc. is registered by the primary supplier.

[0036] Estimation history information ERI is information such as estimation requests, estimation responses, and estimation results. The estimation model group M includes information on a plurality of estimation models generated by the estimation model generation unit MG.

[0037] Figure 4 is a diagram showing an example of an estimation response sent from a supplier to a buyer. When the buyer makes an estimation request to the supplier via the electronic estimation department, after the supplier fills in the estimation response E in Figure 4, it is sent to the buyer via the electronic estimation department. The buyer examines the estimation response E and checks whether there is any abnormality in the estimated unit price (estimated amount) of the product / part.

[0038] The estimation response E is an estimation regarding a molded product and includes entry columns such as material cost, coloring cost, and molding processing cost. Note that the estimation response E has not only entry columns for material cost, coloring cost, and molding processing cost, but also entry columns not shown in the illustration, such as material information including items such as material manufacturers and unit price details including items such as management fees and profits.

[0039] The item of material cost includes material code, material name, specification, quantity (g), material unit price, and loss (%). The item of coloring cost includes coloring material, magnification, material name, quantity (g), coloring unit price, and loss (%). The item of molding processing cost includes molding machine size, quantity, machine type unit price / day, number of shots (times / day), and loss (%).

[0040] Among the items of material costs, the items for which numerical data is input and which are directly related to the calculation of the estimated unit price are the quantity (g) of item A1, the material unit price of item A2, and the loss (%) of item A3. On the other hand, among the items of material costs, the items for which numerical data is not input and which are not directly related to the calculation of the estimated unit price are the material code of item B1, the material name of item B2, and the specification of item B3.

[0041] Also, among the items of coloring costs, the items directly related to the calculation of the estimated unit price are the quantity (g), the coloring unit price, and the loss (%). The items not directly related to the calculation of the estimated unit price are the coloring material, the magnification, and the material name. Furthermore, among the items of molding processing costs, the items directly related to the calculation of the estimated unit price are the molding machine size, the quantity, the machine type unit price per day, the number of shots (times per day), and the loss (%). There are no items not directly related to the calculation of the estimated unit price.

[0042] Figure 5 is a diagram showing an example of input items to be input to the neural network from the item breakdown of the estimate response. In Figure 5, item A is the item name where numerical data such as item A1, A2, A3, etc. of the estimate response E is input, and item B is the item name where non - numerical data (pull - down data) such as item B1, B2, B3, etc. of the estimate response E is input. That is, item A is the item name where numerical data such as "quantity (g)", "material unit price", and "loss (%)" directly related to the calculation of the estimated unit price is input, and item B is the item name where non - numerical data such as "material code", "material name", and "specification" not directly related to the calculation of the estimated unit price is input. The item name of item A and the numerical data corresponding to the item name, and the item name of item B and the non - numerical data corresponding to the item name are given to the input layer of the neural network NN described later.

[0043] Generally, when generating an estimation model from the estimate response E in Figure 4, for the neural network NN, data directly related to the calculation of the estimated unit price (the item name of item A and the input data (numerical data) corresponding to the item name) is given to the input layer, and the estimated unit price described in the estimate response is given to the output layer. Note that since the item name of item A is non - numerical data, pre - processing for numerical conversion is performed.

[0044] In contrast, in the present embodiment, in addition to data directly related to the calculation of the estimated unit price (the item name of item A and the input data (numerical data) corresponding to the item name), data not directly related to the calculation of the estimated unit price (the item name of item B and the input data (non-numerical data) corresponding to the item name) is used to generate an estimation model. Specifically, for a neural network, the item name of item A directly related to the calculation of the estimated unit price and the input data (the first data set) corresponding to the item name, and the item name of item B not directly related to the calculation of the estimated unit price and the input data (the second data set) corresponding to the item name are given to the input layer, and the estimated unit price is given to the output layer to generate an estimation model. Note that since the item name of item A, the item name of item B, and the input data corresponding to the item name of item B are non-numerical data, preprocessing for numerical conversion is performed.

[0045] FIG. 6 is a diagram for explaining an example of the flow of the generation process of the estimation model and the similar product search process. The procurement analysis unit PA reads the quotation history information ERI and performs preprocessing of the learning data. Specifically, the procurement analysis unit PA extracts the item name of item A directly related to the calculation of the estimated unit price and the numerical data corresponding to the item name, and extracts the item name of item B not directly related to the calculation of the estimated unit price and the non-numerical data corresponding to the item name. Then, the procurement analysis unit PA performs preprocessing for numerical conversion of the item name of item A, the item name of item B, and the input data corresponding to the item name of item B, which are non-numerical data. The estimation model generation unit MG gives the data corresponding to the item name of item A, which is numerical data, and the item name of item A, the item name of item B, and the data corresponding to the item name of item B, which are converted into numerical data by preprocessing, to the input layer, gives the estimated unit price to the output layer, performs learning processing, and generates an estimation model.

[0046] Based on the quotation history information ERI, the procurement analysis unit PA generates the analysis report shown in FIG. 10. Also, when the estimation model generation unit MG generates an estimation model, the procurement analysis unit PA acquires the importance items and importance coefficients shown in FIG. 11, which will be described later. The procurement analysis unit PA can search for similar products based on the similarity of items with high importance coefficients and the similarity of importance coefficients.

[0047] FIG. 7 is a diagram showing a configuration example of a neural network. The estimation models constituting the estimation model group M are generated using the neural network NN shown in FIG. 7.

[0048] The neural network NN has an input layer 31, a hidden layer 32, and an output layer 33. In FIG. 7, the input layer 31 has input units 31a shown as circles for the item names of item A directly related to the unit price in the breakdown of the quotation and the data corresponding to the item names, and for the item names of item B not directly related to the unit price in the breakdown of the quotation and the number of elements of the data corresponding to the item names.

[0049] Input to the input layer 31 are the item names of item A directly related to the unit price in the breakdown of the quotation and the data corresponding to the item names, and the item names of item B not directly related to the unit price in the breakdown of the quotation and the data corresponding to the item names.

[0050] The hidden layer 32 has a multi-layer structure including a plurality of hidden layers 32a. The output layer 33 has one output unit 33a, and the unit price of the breakdown items of the quotation is given to the one output unit 33a, and an estimation model is generated. The estimation models are generated for each product / part, and the generated plurality of estimation models are stored in the storage device 12 as an estimation model group.

[0051] FIG. 8 is a flowchart showing an example of the flow of the estimation model generation process. The procurement analysis unit PA reads the quotation history information ERI (S1). The procurement analysis unit PA performs preprocessing of the learning data from the quotation answer E of the quotation history information ERI (S2). In the preprocessing, as described above, processing for quantifying the item names of item A, which is non-numerical data, the item names of item B, and the input data corresponding to the item names of item B is performed.

[0052] Next, the estimation model generation unit MG generates an estimation model by learning the items of the breakdown of the quotation through deep learning (S3). The items of the breakdown of the quotation are the item names of item A and the input data (numerical data) corresponding to the item names that are directly related to the calculation of the unit price of the quotation, the item names of item B and the input data (non-numerical data) corresponding to the item names that are not directly related to the calculation of the unit price of the quotation, and the unit price of the quotation described in the quotation answer E. The estimation model generation unit MG saves the estimation model in the storage device 12 (S4), and ends the process.

[0053] FIG. 9 is a flowchart showing an example of the flow of the estimation process of the unit price of the quotation. The estimation unit EP inputs the quotation answer E into the estimation model (S11). Specifically, the estimation unit EP inputs the item names of item A and the input data (numerical data) corresponding to the item names that are directly related to the calculation of the unit price of the quotation, and the item names of item B and the input data (non-numerical data) corresponding to the item names that are not directly related to the calculation of the unit price of the quotation among the items of the breakdown of the quotation answer E into the estimation model. As described above, since the item names of item A, the item names of item B, and the input data corresponding to the item names of item B are non-numerical data, preprocessing for quantification is performed. Note that the preprocessing is performed by the procurement analysis unit PA. The estimation unit EP estimates the unit price of the product or part (S12), and ends the process. By inputting the item names of item A and the input data (numerical data) corresponding to the item names that are directly related to the calculation of the unit price of the quotation, and the item names of item B and the input data (non-numerical data) corresponding to the item names that are not directly related to the calculation of the unit price of the quotation into the estimation model, the unit price of the quotation (estimated unit price) estimated by the estimation model is output from the estimation unit EP. The estimated unit price estimated in this way is used for generating an analysis report by the procurement analysis unit PA.

[0054] As described above, the estimation model generation unit MG generates an estimation model using the item name of item A that is directly related to the calculation of the estimated unit price and the input data (numerical data) corresponding to the item name, and the item name of item B that is not directly related to the calculation of the estimated unit price and the input data (non-numerical data) corresponding to the item name.

[0055] Then, when a new estimate response E is input, the estimation unit EP estimates the estimated unit price using the estimation model thus generated. As a result, the estimation unit EP can estimate the estimated unit price considering items that would be overlooked if the buyer is not a veteran (the item name of item B that is not directly related to the calculation of the estimated unit price and the input data corresponding to the item name), so the buyer can easily grasp an abnormal estimate.

[0056] FIG. 10 is a diagram showing an example of an analysis report indicating the relationship between the estimated unit price and the estimated price. In FIG. 10, the horizontal axis represents the estimated unit price estimated by the estimation model, and the vertical axis represents the estimated unit price estimated in the estimate response E. The procurement analysis unit PA can generate an analysis report based on the estimated unit price and the estimated unit price estimated by the estimation unit EP using the estimation model. Specifically, as shown in FIG. 10, the relationship between the estimated unit price of the estimate response E and the estimated unit price obtained by the estimation model is displayed by the marker MK.

[0057] The straight line L is a line connecting the points where the estimated unit price and the estimated unit price are the same. The marker MK above the straight line L indicates that the estimated unit price is higher than the estimated unit price, and the marker MK below the straight line L indicates that the estimated unit price is lower than the estimated unit price. That is, the greater the distance from the straight line L to the marker MK, the greater the difference between the estimated unit price and the estimated unit price. When the buyer selects a marker MK that is far from the straight line L using an input device or the like, an abnormal estimate response E with a large difference between the estimated unit price and the estimated unit price can be displayed on the display device 3a.

[0058] The priority cost-down areas in the analysis report are products / parts with high unit prices and areas where the estimated unit price is higher than the estimated unit price. Therefore, significant cost reduction can be achieved by prioritizing the reduction of the unit price of products / parts in the priority cost-down areas.

[0059] In addition, since the estimated unit price in the area for considering improvement measures is lower than the estimated unit price, it is an area where it can be estimated that the supplier has know-how for cost reduction. By applying such know-how to the products / parts in the priority cost-down areas, continuous cost reduction can be achieved.

[0060] Furthermore, the automation area is an area where the unit price of the product / part is low and the estimated unit price is lower than the estimated unit price, so the product / part is automatically adopted without manual checking of the estimated reply E.

[0061] Figure 11 is a diagram showing an example of the importance of the items in the breakdown of the estimate. When the estimation model generation unit MG generates an estimation model, it calculates the importance (importance coefficient) indicating how much each item (variable) affects the estimation of the estimated price. As shown in Figure 11, the procurement analysis unit PA displays, for example, 15 items with a large importance (importance coefficient) calculated by the estimation model generation unit MG on the display device 3a. As a result, the top 15 important items (variables) contributing to the calculation of the estimated unit price are displayed on the display device 3a.

[0062] The buyer can grasp the items (variables) with high importance contributing to the unit price of the product / part, which helps the buyer to know which items should be focused on when evaluating the estimated reply E. In addition, by grasping the items (variables) with high importance contributing to the unit price of the product / part, the buyer can obtain information useful for price negotiation and achieve continuous cost reduction.

[0063] In addition, the procurement analysis unit PA can search for similar products from the estimated history information ERI for which estimates were made in the past, using the importance (importance coefficient) calculated by the estimated model generation unit MG.

[0064] FIG. 12 is a diagram showing an example of a similar product search screen when searching for similar products using importance. The similar product search screen 40 has input fields 41, 42, and 43 into which the importance coefficients A, B, and C are input (selected). The order of the importance coefficients A, B, and C is in descending order of importance. In addition, the similar product search screen 40 has an input field 44 for inputting the upper limit value of the importance coefficient A and an input field 45 for inputting the lower limit value of the importance coefficient A. A selection button 46 for selecting the importance coefficient is provided in the input field 41 into which the importance coefficient A is input (selected). The same configuration applies to the importance coefficients B and C.

[0065] When the user presses the selection button 46 using an input device or the like, a list of importance coefficients is displayed. The user can select a desired importance coefficient from the displayed list of importance coefficients. In addition, the user can input the upper limit value of the selected importance coefficient A into the input field 44 and the lower limit value into the input field 45. In the example of FIG. 12, as the importance coefficient A, [material cost] quantity (g) is selected, and the upper limit value is 24 and the lower limit value is 12 are input.

[0066] In addition, the similar product search screen 40 has a search button 47. After the user selects the importance coefficients A, B, and C and inputs the upper limit value and the lower limit value of each coefficient, by pressing the search button 47, similar products can be searched from the estimated history information ERI for which estimates were made in the past.

[0067] Note that the selected importance coefficients are not limited to three, namely importance coefficients A, B, and C, and may be one, two, or four or more. In addition, it is not necessary to input both the upper limit value and the lower limit value of the selected importance coefficient, and either one may be sufficient. Also, when the selected importance coefficient is the non-numerical data item B, it is not necessary to input the upper limit value and the lower limit value.

[0068] The user can search for similar products from the quotation history information ERI for which quotations were made in the past while adjusting the search conditions by arbitrarily setting the order of importance (coefficient), the upper limit value, and the range of the lower limit value for each coefficient using the similar product search screen 40.

[0069] FIG. 13 is a flowchart showing an example of the flow of the similar product search process using importance. The procurement analysis unit PA acquires the items with high importance and importance coefficients calculated by the estimation model generation unit MG (S21). The procurement analysis unit PA extracts, from the past quotation responses E stored as the quotation history information ERI, the order of the items with high importance set on the similar product search screen 40 and the quotation responses E with similar importance coefficients (S22). The procurement analysis unit PA outputs the products / parts of the extracted quotation responses E as similar products to the display device 3a (S23) and ends the process.

[0070] Through the above processing, when the buyer receives a quotation response E for a certain product / part from a certain supplier, the buyer can search for past similar products and request a quotation for the product / part from another supplier that offers the similar products, thereby comparing the quotation unit prices among multiple suppliers. In addition, the buyer can easily compare the quotation unit price of the quotation response E with the quotation unit prices of past similar products and can grasp abnormal quotations.

[0071] FIG. 14 is a diagram showing an example of a standard unit price table. When the estimation model generation unit MG generates an estimation model by learning the items in the breakdown of the quotation using the neural network NN, the estimation model generation unit MG selects the items that serve as the basis for estimating the quotation unit price and calculates the standard unit price for each of the selected items. The calculated standard unit prices are associated with the breakdown name of the quotation, item name, unit / reference conditions, and currency as shown in FIG. 13 and are stored in the storage device 12 as a standard unit price table.

[0072] By referring to the standard unit price table, the buyer can obtain the standard unit price of item B that is not directly related to the quoted unit price, so it is possible to determine whether there is any abnormality in the quoted unit price of the quotation response E.

[0073] In addition, each step in the flowchart in this specification may be executed in a different order, executed simultaneously in multiple steps, or executed in a different order for each execution, as long as it does not violate its nature.

[0074] Although several embodiments of the invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and its equivalent scope.

Explanation of Reference Numerals

[0075] 1... Supply chain management system, 2... Server, 3, 4... Terminals, 3a... Display device, 5... Network, 11... Processor, 12... Storage device, 31... Input layer, 31a... Input unit, 32, 32a... Hidden layer, 33... Output layer, 33a... Output unit, 40... Similar product search screen, 41, 42, 43, 44, 45... Input fields, 46... Selection button, 47... Search button.

Claims

1. An estimated model generation unit that uses, as learning data, data directly related to the calculation of the estimated unit price of a quotation response, data not directly related to the calculation of the estimated unit price of the quotation response, and the estimated unit price of the quotation response, and generates an estimated model using a neural network; An estimation unit that estimates the unit price of a product / part using the estimated model generated by the estimated model generation unit; An analysis unit that calculates an importance coefficient indicating the degree of influence on estimating the unit price when the estimated model generation unit generates the estimated model, and searches for similar products to the product / part based on the similarity of items with high importance coefficients and / or the similarity of importance coefficients; A quotation estimation device having the above.

2. The data directly related to the calculation of the estimated unit price of the quotation response includes the item name of the item directly related to the calculation of the estimated unit price and the input data corresponding to the item name of the item directly related to the calculation of the estimated unit price. The data not directly related to the calculation of the estimated unit price of the quotation response includes the item name of the item not directly related to the calculation of the estimated unit price and the input data corresponding to the item name of the item not directly related to the calculation of the estimated unit price. The quotation estimation device according to Claim 1.

3. The analysis unit performs preprocessing to digitize the item name of the item directly related to the calculation of the estimated unit price, the item name of the item not directly related to the calculation of the estimated unit price, and the input data corresponding to the item name of the item not directly related to the calculation of the estimated unit price. The quotation estimation device according to Claim 2.

4. The analysis unit generates an analysis report showing the relationship between the unit price estimated by the estimation unit and the estimated unit price of the quotation response, and displays it on a display device. The quotation estimation device according to Claim 1.

5. When generating the estimated model, the estimated model generation unit selects items that are the basis for estimating the unit price, and calculates a standard unit price for each of the selected items. The quotation estimation device according to Claim 1.

6. The analysis unit accepts the input of one or more items in descending order of the importance coefficient, and searches for similar products to the received one or more items. The quotation estimation device according to Claim 1.

7. The estimated model generation unit generates the estimated model for each product / part, and stores the generated multiple estimated models in a storage device as an estimated model group. The quotation estimation device according to Claim 1. Claim 8. A quotation estimation method executed by a quotation estimation device, wherein data directly related to the calculation of the unit price of a quotation response, data not directly related to the calculation of the unit price of the quotation response, and the unit price of the quotation response are used as learning data, and an estimation model is generated using a neural network; estimating the unit price of a product / part using the generated estimation model; calculating an importance coefficient indicating the degree of influence on estimating the unit price when generating the estimation model, and searching for similar products similar to the product / part based on the similarity of items with a high importance coefficient and / or the similarity of the importance coefficients; A quotation estimation method. Claim 9. On a computer, generating an estimation model using a neural network, where data directly related to the calculation of the unit price of a quotation response, data not directly related to the calculation of the unit price of the quotation response, and the unit price of the quotation response are used as learning data; estimating the unit price of a product / part using the generated estimation model; calculating an importance coefficient indicating the degree of influence on estimating the unit price when generating the estimation model, and searching for similar products similar to the product / part based on the similarity of items with a high importance coefficient and / or the similarity of the importance coefficients; A quotation estimation program for causing the above to be executed.

Images