Distribution support device, distribution support method, and program

Abstract

This distribution support device comprises: an acquisition unit that acquires, as information for allocating and distributing supplied merchandise for orders, supply information including a plurality of attributes and a quantity of the merchandise for an i-th supply by a supplier, order information including a plurality of attributes and a quantity of the merchandise for a j-th order by an orderer, and a distribution condition relating to the importance of attributes and of orders; a calculation unit that, on the basis of the supply information, the order information, and the distribution condition that were acquired, acquires an index associated with the distribution condition from a database having prescribed indices relating to attributes stored therein, and uses the acquired index to calculate a first coefficient relating to a supply and order combination taking the importance of the attributes into consideration, and a second coefficient relating to the importance of the orders; a distribution result creation unit that, on the basis of the supply information, the order information, the first coefficient, and the second coefficient, creates a prescribed distribution result including allocated quantities of the merchandise in the supply information for the order information; and a distribution result editing unit that receives an editing operation by a user, said operation comprising an operation instructing that some of the allocated quantities included in the created prescribed distribution result be confirmed, or an operation instructing that some of the allocated quantities be changed.

Description

Loading support device, loading support method, and program

[0001] The disclosed technology relates to a cargo distribution support device, a cargo distribution support method, and a program.

[0002] Conventionally, there are techniques for supporting market trading management.

[0003] Non-Patent Document 1 discloses a system for fruit and vegetable wholesale markets. The system in Non-Patent Document 1 has functions specialized for wholesale markets. Wholesale markets involve tasks such as order management, receipt management, distribution management, and sales management, and this system has the function to input and manage information related to these tasks, thereby contributing to supporting operations at wholesale markets.

[0004] Non-Patent Document 2 discloses another sales management system for the wholesale industry. The system in Non-Patent Document 2 features functions specialized for the wholesale industry, such as handling quality control items unique to the wholesale industry and handling consignment sales, a sales format unique to the wholesale industry. Like the system in Non-Patent Document 1, it is designed to support the operations of the wholesale industry and is equipped with functions to support these operations.

[0005] KitFit Marche Fruit and Vegetable Market System: Features and Functions | Tsuzuki Electric Solutions, https: / / tsuzuki.jp / jigyo / kitfit-marché / features-functions / Fresh Market-kun: A Sales Management System Specialized for Wholesale and Intermediate Wholesale Businesses | Mitsubishi Electric IT Solutions, https: / / www.mdsol.co.jp / products / freshichiba_product-outline /

[0006] One example of market operations is the distribution management of products, which is one of the various operations performed at wholesale markets. Distribution management involves allocating products collected and supplied from production areas across the country to orders from wholesalers and retailers, thereby completing sales between both parties.

[0007] When it comes to distribution, wholesalers' distribution management staff must consider multiple orders simultaneously. Usually, there is often not enough stock to fulfill all orders. Therefore, distribution management staff must appropriately determine the priority of each order and distribute the orders accordingly, resulting in a significant workload. In such cases, even if the systems described in Non-Patent Document 1 and Non-Patent Document 2 are used, it is necessary to make decisions one by one while checking whether the results of allocating orders and supplies satisfy multiple conditions. Traditionally, distribution staff have created distribution results that meet the conditions depending on various situations, based on past distribution performance and experience.

[0008] One possible solution to this problem is to reduce the burden of loading and unloading work by automatically generating loading and unloading results that appropriately reflect the conditions specified by the loading and unloading worker, taking into account various circumstances. However, automatically generating loading and unloading results presents the following challenges. First, it is not always possible to present loading and unloading results that meet all of the loading and unloading worker's requirements or that satisfy the loading and unloading worker. Second, loading and unloading work may begin before all supplies and orders are completed, or may be revised after loading and unloading has begun. Third, it is difficult to determine how to calculate the coefficients used for automatic loading and unloading. Furthermore, it is not always possible to obtain sufficient data in advance to use machine learning, etc. Fourth, the tolerance for discrepancies between the quantity in the order information and the allocated quantity assigned to that order information varies depending on the supply and demand balance and the characteristics of the orderer.

[0009] Even if a system that automatically creates results is used, the results may be of poor quality and unusable. Also, it may take a long time to implement the system.

[0010] The disclosed technology has been developed in consideration of the above points, and aims to provide a loading support device, loading support method, and program that can create high-quality loading results and reduce the workload associated with loading.

[0011] A first aspect of the present disclosure is a loading support device, comprising: an acquisition unit that acquires, as information for allocating and loading a product to be supplied to an order, supply information including a plurality of product attributes and quantities for an i-th supply from the supplier; order information including a plurality of product attributes and quantities for a j-th order from the orderer; and loading conditions related to the importance of each of the attributes and the order; and, based on the acquired supply information, order information, and loading conditions, acquires an index corresponding to the loading conditions from a database in which predetermined indexes related to attributes are stored, and calculates a first coefficient related to a combination of supply and order that takes into account the importance of the attributes and a second coefficient related to the importance of the order using the acquired index. a distribution result creation unit that creates a predetermined distribution result including the allocation quantity of the product in the supply information for the order information based on the supply information, the order information, the first coefficient, and the second coefficient; and a distribution result editing unit that accepts user editing operations including an operation to instruct the confirmation of a portion of the allocation quantities included in the created predetermined distribution result or an operation to instruct a change to a portion of the allocation quantities, wherein the calculation unit changes the first coefficient in accordance with the editing operation, and the distribution result creation unit recreates the predetermined distribution result based on the editing operation, excluding the allocation quantities instructed to be confirmed or the allocation quantities instructed to be changed.

[0012] A second aspect of the present disclosure is a loading support method, which acquires, as information for allocating and loading a product to be supplied to an order, supply information including a plurality of product attributes and quantities for an i-th supply from the supplier, order information including a plurality of product attributes and quantities for a j-th order from the orderer, and loading conditions related to the importance of each of the attributes and the order, and acquires an index corresponding to the loading conditions from a database storing predetermined indexes related to attributes based on the acquired supply information, order information, and loading conditions, and calculates, using the acquired index, a first coefficient related to a combination of supply and order that takes into account the importance of the attributes and a loading condition related to the importance of the order. the first coefficient and a second coefficient related to the allocation quantity of the product in the supply information for the order information based on the supply information, the order information, the first coefficient, and the second coefficient, creates a predetermined distribution result including the allocation quantity of the product in the supply information for the order information, accepts a user editing operation including an operation to instruct the confirmation of a part of the allocation quantity included in the created predetermined distribution result or an operation to instruct the change of a part of the allocation quantity, changes the first coefficient in accordance with the editing operation, and executes a process of re-creating the predetermined distribution result based on the editing operation, excluding the allocation quantity instructed to be confirmed or the allocation quantity instructed to be changed.

[0013] A third aspect of the present disclosure is a program for causing a computer to function as each part of the distribution support device described in the first aspect.

[0014] The disclosed technology makes it possible to create high-quality distribution results and reduce the workload associated with distribution.

[0015] FIG. 1 is a block diagram showing the hardware configuration of a loading support device. FIG. 2 is a block diagram showing the functional configuration of a support system including the loading support device of this embodiment. FIG. 3 is a flowchart showing the flow of support processing as a loading support method using the loading support device. FIG. 4 is a flowchart showing the calculation process of each coefficient. FIG. 5 is a flowchart showing the loading result creation process. FIG. 6 is an example of a screen for registering order information. FIG. 7A is supply information for input example (1). FIG. 7B is order information for input example (1). FIG. 7C is loading conditions for input example (1). FIG. 8 is an example of an index to be read out. FIG. 9 is an example of an interpretation result of order information. FIG. 10 is an example of a first coefficient. FIG. 11 is an example of a first coefficient. FIG. 12 is an example of a second coefficient. FIG. 13 is an example of a loading result for input example (1). FIG. 14 is another example of a loading result for input example (1). FIG. 15 is a first example of an editing screen. FIG. 16 shows an example of accepting an editing operation to add a link between a supply and an order. FIG. 17 shows an example of accepting an editing operation to instruct the deletion of a link between an order and a supply. FIG. 18 shows an example of accepting an editing operation to manually correct an allocation amount. FIG. 19 shows an example of accepting an editing operation to automatically correct an allocation amount. FIG. 20 shows an example of accepting an editing operation to confirm a portion of the distribution results and set NG. FIG. 21 shows an example of accepting an editing operation to specify each category for a specified link and allocation amount. FIG. 22 shows an example of notifying a user that a supply or order has been updated. FIG. 23 shows an example of accepting an editing operation to instruct the re-creation of a distribution result. FIG. 24 shows an example of accepting an editing operation to instruct filtering and sorting by attribute value. FIG. 25 shows an example of the operation of the distribution result creation unit when an editing operation is accepted. FIG. 26 shows an example of a process for learning important attributes. FIG. 27 shows another example of the importance of attribute pairs. FIG. 28 shows another example of the importance of attribute pairs. Fig. 29 is another example of the importance of attribute pairs. Fig. 30 is an example of a conversion table between attributes and ranks. Fig. 31 is an example of a table of the importance of attribute pairs using ranks. Fig. 32 is another example of an edit screen. Fig. 33 is another example of an edit screen. Fig. 34 is another example of an edit screen.

[0016] An example of an embodiment of the disclosed technology will be described below with reference to the drawings. Note that the same reference numerals are used to designate identical or equivalent components and parts in each drawing. Also, the dimensional proportions in the drawings are exaggerated for the sake of explanation and may differ from the actual proportions.

[0017] The configuration of an embodiment of the present disclosure will be described below.

[0018] FIG. 1 is a block diagram showing the hardware configuration of a load distribution support device 100. As shown in FIG. 1, the load distribution support device 100 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage 14, an input unit 15, a display interface (I / F) 16, and a communication interface (I / F) 17. Each component is connected to each other via a bus 19 so as to be able to communicate with each other. The load distribution support device 100 may be configured with one or more servers equipped with these hardware components, or may be configured with multiple servers.

[0019] The CPU 11 is a central processing unit that executes various programs and controls each part. That is, the CPU 11 reads the programs from the ROM 12 or the storage 14 and executes the programs using the RAM 13 as a work area. The CPU 11 controls the above components and performs various arithmetic processing in accordance with the programs stored in the ROM 12 or the storage 14. In this embodiment, the programs are stored in the ROM 12 or the storage 14.

[0020] The ROM 12 stores various programs and various data. The RAM 13 temporarily stores programs or data as a working area. The storage 14 is configured by a storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive), and stores various programs including an operating system and various data.

[0021] The input unit 15 includes a pointing device such as a mouse and a keyboard, and is used to input various types of information.

[0022] The display interface 16 is, for example, a liquid crystal display, and displays various information. The display interface 16 may be a touch panel type and function as the input unit 15.

[0023] The communication interface 17 is an interface for communicating with other devices such as terminals, etc. For this communication, for example, a wired communication standard such as Ethernet (registered trademark) or FDDI, or a wireless communication standard such as 4G, 5G, or Wi-Fi (registered trademark) is used.

[0024] Next, each functional configuration of the load distribution support device 100 will be described. Figure 2 is a block diagram showing the functional configuration of a support system including the support device of this embodiment. Each functional configuration is realized by the CPU 11 reading a program stored in the ROM 12 or storage 14, expanding it into the RAM 13, and executing it. Below, embodiments will be described separately according to the processing mode. Note that the hardware configuration and functional configuration are similar in each embodiment, so the same reference numerals will be used in the description.

[0025] First Embodiment As shown in FIG. 2, the support system 1 includes a distribution support device 100 and a user terminal 110 .

[0026] The user terminal 110 includes a first input unit 112 , a second input unit 114 , a third input unit 116 , and a fourth input unit 118 .

[0027] The first input unit 112 accepts input, addition, update, or deletion of supply information specified by the user directly or via a file, and outputs the information to the distribution support device 100. The supply information consists of a supply ID, multiple attributes, supply quantity, and update date and time. The multiple attributes include the desired sales date, shipping origin, grade, class, and packaging style. The first input unit 112 also outputs the information to the distribution support device 100 so that it can be seen that supply information has been added, updated, or deleted after the distribution work has begun.

[0028] The second input unit 114 accepts input, addition, update, or deletion of order information directly input by the user or specified via a file, and outputs the information to the distribution support device 100. The order information consists of an order ID, an orderer, multiple attributes, an order quantity, a key order, and an update date and time. The multiple attributes include a desired purchase date, a designated place of origin, a designated grade, a designated class, and a designated packaging style. The multiple attributes and the order quantity may be specified by the user in addition to the orderer. The order information may include key attributes for each orderer or order. The order information may also include a tolerance for deviations from the allocated order quantity for each orderer or order. The orderer and user specifications for each attribute, including the designated place of origin, designated grade, designated class, and designated packaging style, as well as the key attributes and deviation tolerance are recorded, and patterns in which values ​​based on the frequency of occurrence or the number of days since the last occurrence exceed a threshold are presented as templates, which can be used or modified for input.

[0029] In addition, the second input unit 114 outputs to the distribution support device 100 so that it can be seen that order information has been added, updated, or deleted after the distribution work has started.

[0030] The third input unit 116 accepts input of distribution conditions specified by the user and outputs them to the distribution support device 100. The distribution conditions include important attributes and designation of important orders. The designation format is, for example, [Place of Origin / Grade, None]. In this case, it indicates that the attribute with the first priority of importance is "Place of Origin," the attribute with the second priority of importance is "Grade," and it also indicates that important orders are not included in the distribution conditions.

[0031] The fourth input unit 118 accepts editing operations from the user on the distribution result editing screen described below, and outputs the operations to the distribution support device 100 .

[0032] The distribution support device 100 includes an acquisition unit 120, a database 122, a calculation unit 124, a distribution result creation unit 126, a distribution result editing unit 128, and a learning unit 130.

[0033] The acquisition unit 120 acquires supply information, order information, and distribution conditions as various information for allocating and distributing the products to be supplied to the orders, through input from the user terminal 110. Note that products on the market are an example of products.

[0034] The database 122 stores multiple indices related to attributes. The multiple indices are the importance of attribute pairs and weights corresponding to important attributes. The importance of attribute pairs is determined for attribute pairs on the supply side and the order side, and the importance is determined according to the degree of match, for example, "for each attribute, if the supply and the order are the same, the value is 2, and if they are different, the value is 1." Different values ​​may also be set for each attribute. Weights are determined according to the importance. For example, a value of 25 is set for an attribute with the highest priority, a value of 5 for an attribute with the second highest priority, and a value of 1 for the others.

[0035] The calculation unit 124 acquires indicators corresponding to the distribution conditions from the database 122 based on the acquired supply information, order information, and distribution conditions. Furthermore, the calculation unit 124 uses the acquired indicators to calculate a first coefficient related to the combination of supply and order taking into account the importance of attributes, and a second coefficient related to the importance of the order. The calculation unit 124 changes the first coefficient in response to an editing operation received by the distribution result editing unit 128. Furthermore, the calculation unit 124 changes the first coefficient in response to a user specification input by the second input unit 114. Details of the processing by the calculation unit 124 and each coefficient will be described later in the explanation of operation.

[0036] The distribution result creation unit 126 creates a predetermined distribution result including the allocation quantity of the product in the supply information for the order information based on the supply information, order information, first coefficient, and second coefficient. At that time, the distribution result creation unit 126 re-creates the predetermined distribution result based on editing operations from the distribution result editing unit 128, excluding the allocation quantity instructed to be confirmed or changed. In addition, the distribution result creation unit 126 performs distribution reflecting the supply and demand balance. Details of the processing by the distribution result creation unit 126 will be described later in the explanation of operation.

[0037] The distribution result editing unit 128 accepts user editing operations for the distribution results created by the distribution result creation unit 126. Specifically, the distribution result editing unit 128 provides editing functions such as manual correction, saving, and reading of the distribution results, as well as a function to instruct corrections to the distribution results. It also supports changes to supply information and order information. Details of the processing by the distribution result editing unit 128 will be described later in the explanation of its operation.

[0038] The learning unit 130 learns the most important attribute of a plurality of attributes for a user based on the user's past editing operations. The learning unit 130 also learns the tolerance for deviation based on the user's past editing operations. Details of the processing by the learning unit 130 will be described later in the explanation of the operation.

[0039] Next, the operation of the loading distribution support device 100 will be described. Figure 3 is a flowchart showing the flow of support processing as a loading distribution support method by the loading distribution support device 100. The support processing is performed by the CPU 11 reading a program from the ROM 12 or storage 14, expanding it into the RAM 13, and executing it. The loading distribution support device 100 executes the support processing in response to various information input from the user terminal 110.

[0040] In step S100, the CPU 11 acquires supply information, order information, and distribution conditions as the acquisition unit 120. At this time, the CPU 11 may acquire updated contents of supply information, added supply information, updated contents of order information, or added order information as the acquisition unit 120.

[0041] Furthermore, the CPU 11 may function as the acquisition unit 120 to acquire order information in which the order information input by the orderer has been overwritten with attributes and quantities input by the user.

[0042] Furthermore, the CPU 11, as the acquisition unit 120, may present to the orderer patterns of past order information input by the orderer, patterns whose appearance frequency or a value based on the number of days since the last appearance exceeds a threshold, and acquire order information using those patterns.

[0043] Furthermore, the CPU 11 may function as the acquisition unit 120 to present the learned tolerance for deviation to the user and acquire order information including the tolerance for deviation.

[0044] In step S102, the CPU 11, as the calculation unit 124, calculates a first coefficient for a combination of supply and order that takes into account the importance of attributes, and a second coefficient for the importance of the order, based on the acquired supply information, order information, and distribution conditions, using the indexes acquired from the database 122. In step S100 above, when updated content of supply information, added supply information, updated content of order information, or added order information is acquired, the CPU 11, as the calculation unit 124, calculates a first coefficient and a second coefficient for the updated or added supply information or order information.

[0045] In step S104, the CPU 11, as the distribution result creation unit 126, creates a predetermined distribution result including the allocation quantity of the product in the supply information for the order information based on the supply information, the order information, the first coefficient, and the second coefficient.

[0046] In step S106, the CPU 11, as the distribution result editing unit 128, presents an editing screen to the user terminal 110 for accepting user editing operations for the specified distribution result created in step S104. Here, if there is supply information or order information that has been updated or added after the specified distribution result was created, the CPU 11, as the distribution result editing unit 128, notifies the user of the combination of supply and order to be allocated based on the first coefficient, and accepts an instruction from the user as to whether or not to adopt the combination of supply and order to be allocated, or an instruction as to whether or not to create the specified distribution result again.

[0047] In step S108, the CPU 11, functioning as the distribution result editing unit 128, determines whether or not an editing operation has been accepted on the editing screen presented in step S106. If an editing operation has been accepted, the process proceeds to step S102. On the other hand, if an editing operation has not been accepted, the support process ends.

[0048] Next, the calculation process of each coefficient in step S102 will be described with reference to the flowchart of FIG.

[0049] In step S120, based on the acquired distribution conditions, the CPU 11 acquires the importance of the attribute pair and the weight of each attribute corresponding to the important attribute of the distribution conditions as indicators corresponding to the distribution conditions from the database 122. Examples of the indicators will be described later.

[0050] In step S122, the CPU 11 calculates a first coefficient c for the i-th supply in the supply information and the j-th order in the order information using the acquired supply information, order information, importance of the attribute pair, and weight of each attribute according to the following equation (1): ｉ，ｊ The first coefficient c ｉ，ｊ is the weight p of each attribute for each supply and order pair (i, j). ｋ The importance of attribute pairs is calculated using ｋ，ｉ，ｊ The weighted sum c ｉ，ｊ is obtained by taking

[0051] ...(1)

[0052] Here, P = [p １ , ..., p K ], p ｋ : weight for the kth attribute. In this way, for each combination of the ith supply and the jth order, the first coefficient is calculated by calculating a weighted sum using the importance of the attribute pair and the weight of each attribute.

[0053] In step S124, the second coefficient d ｊ The second coefficient d ｊ In calculating this, for example, 1 is set for important orders specified as distribution conditions, and 0 is set for other orders. In this way, when an order that is designated as important in the distribution conditions is specified, the second coefficient is found by calculating a coefficient that corresponds the allocation quantity to the order quantity of the designated order.

[0054] Next, the distribution result creation process in step S104 will be described with reference to the flowchart of FIG.

[0055] In step S140, the CPU 11 calculates the allocation amount to be calculated by the variable q of the allocation amount from the i-th supply to the j-th order. ｉ，ｊ Define

[0056] In step S142, the CPU 11 calculates a first coefficient c ｉ，ｊ and the allocation variable q ｉ，ｊ Define the objective function as a linear sum of

[0057] ...(2)

[0058] In step S144, the CPU 11 sets (or updates) the tolerance for deviation of the allocation amount from the order quantity.

[0059] In step S146, the CPU 11 calculates the allocation variable q ｉ，ｊ The constraints that must be satisfied are defined as follows [1] to [4].

[0060] [1] Allocation variable q ｉ，ｊ is greater than or equal to zero for all i and j.

[0061] [2] Allocation variable q ｉ，ｊ The sum of the quantities for j is equal to the i-th supply. However, in the case of excess supply where the sum of the orders is less than the sum of the supply, the allocation variable q ｉ，ｊ It may be permitted to relax the constraint so that the sum of the quantities for j is less than the i-th supply.

[0062] [3] Allocation variable q ｉ，ｊ The difference between the total amount of i related to the jth order, that is, the amount allocated to the jth order, and the order quantity is equal to or less than the tolerance for deviation.

[0063] [4] Second coefficient d ｊ If is 1, the allocation to the jth order is set equal to the order quantity.

[0064] Regarding [4], the constraint may be a quantity corresponding to the specification of an important order. For example, if the specification of the distribution condition is "important order (at least 90% of the order quantity)", the second coefficient d ｊIn this way, the constraint [4] can be set to satisfy the allocation of the order quantity according to the second coefficient.

[0065] In step S148, the CPU 11 calculates the allocation variable q that maximizes the objective function by a mathematical optimization technique. ｉ，ｊ is derived.

[0066] In step S150, the CPU 11 determines whether a solution has been found. The determination of whether a solution has been found can be made, for example, by using a method that can determine whether constraints are satisfied, and by referring to the solution derived result obtained by software used in the optimization method, a so-called solver, or the like. If a solution has been found, the process ends. If a solution has not been found, the process returns to step S144, where the deviation tolerance is increased and the process is updated, or in step S146, the constraints that the allocation amount variables must satisfy are relaxed and the derivation is repeated. The found solution is output as the allocation result of the allocation result creation unit 126.

[0067] Next, an example of a screen for registering order information will be described below: Fig. 6 shows an example of a screen for registering order information.

[0068] For each attribute and order quantity, specifications from the user as well as from the orderer can be registered. Important attributes and tolerances for deviations can be set for each orderer and each order. For items other than quantity, frequently used patterns for each orderer are displayed as templates, and order information can be registered simply by entering the quantity for the pattern that was actually ordered, streamlining the registration process.

[0069] For example, it displays patterns where the weighted sum q_1×f+q_2×r+q_3×c of the appearance frequency in the most recent fixed period f, the number of days since the last appearance r, and the appearance frequency c in the same period last year exceeds a threshold. [q_1, q_2, q_3] is the weight, and if [q_1, q_2, q_3] = [1, -1, 1], the greater the appearance frequency f in the most recent fixed period or the appearance frequency c in the same period last year, or the smaller the number of days r since the last appearance, the greater the contribution. Note that the symbol "_" indicates a subscript.

[0070] Figure 6 shows an example of overwriting or complementing "orderer's specification" with "user's specification." For order ID "4," the orderer's specification of "good" and "medium" is overwritten with the user's specification of "B" and "M." For order ID "3," the place of origin, which was not specified by the orderer, is complemented with the user's specification of "Ishikawa."

[0071] Next, an input example (1) of supply information, order information, and distribution conditions will be explained. Fig. 7A is the supply information of input example (1). Fig. 7B is the order information of input example (1). Fig. 7C is the distribution conditions of input example (1). Fig. 8 is an example of the indexes to be read out.

[0072] The following describes the loading conditions. Loading conditions include important attributes and deviation tolerance, and as shown in FIG. 7B, loading conditions can be entered for each order and each orderer. The table in the upper part of FIG. 7B shows an example of entering loading conditions for each order. The table in the lower part of FIG. 7B shows an example of entering loading conditions for each orderer. Furthermore, as shown in FIG. 7C, loading conditions for the entire product can be entered.

[0073] The indexes that are read out will be explained. The weight of each attribute that is read out in the calculation unit 124 according to the distribution conditions is the weight obtained by integrating multiple attributes. For example, if the important attribute of the distribution conditions is designated as "origin", and the weight of each attribute is P = [p_origin, p_grade, p_class], the weight of each attribute that is read out will be P = [4, 2, 1]. In this example, since the packaging style and quantity are all the same, the importance of attribute pairs and the weight of each attribute are not set, but they may be set in the same way as the attributes above.

[0074] From the order information and distribution conditions in input example (1), each attribute is interpreted as shown in Figure 9. If there is a user specification, the specification from the orderer is supplemented or overwritten with the user specification. For example, for "Order ID" = "D3", the orderer's specification of "Place of Origin" = "Not Specified" is supplemented with "Place of Origin" = "Ishikawa" as specified by the user. For "Order ID" = "D4", the "Grade" = "Good" and "Class" = "Medium" are overwritten with the user's specification of "Grade" = "B" and "Class" = "M". Priority for important attributes and deviation tolerance is given in the order of order, orderer, and distribution conditions.

[0075] For the orderer "Wholesaler X," the "important attribute" for each orderer is "grade" and the "deviation tolerance" is "10%." However, the "important attribute" of order ID "D1" = "class" and the "deviation tolerance" of order ID "D3" = "20%," which are set for each order, take priority.

[0076] In the case of an orderer "Wholesaler Y" who has not specified either, the "Important Attribute" = "Place of Origin" and "Tolerance of Deviation" = "5%" specified in the third input section 116 are used.

[0077] 10 and 11 are examples of the first coefficient. The example shown in FIG. 10 is an example in which the first coefficient is calculated for each supply and order pair (supply / order pair). The example shown in FIG. 11 is an example in which the weight p of each attribute is calculated for each supply and order pair (supply / order pair) as shown in the process of step S122 described above. ｋ The importance of attribute pairs is calculated using ｋ，ｉ，ｊ The first coefficient c ｉ，ｊ For each supply / order pair, the importance v for each attribute pair is calculated. ｋ，ｉ，ｊis read out. An example will be given for the supply / order pair "S1, D1" (see the dashed line portion in FIG. 10). As shown in FIG. 11, for "S1, D1", the attribute pair of origin is "Aichi-Aichi", which is the same for supply and order, so the value is 1. For "S1, D1", the attribute pair of grade is "A-B", which is different for supply and order, so the value is 0.5. For "S1, D1", the attribute pair of class is "L-L", which is the same for supply and order, so the value is 1. The weighted sum of the importance of each attribute pair above can be calculated as "1*1.0+2*0.5+4*1.0=6.0".

[0078] 12 shows an example of the second coefficient. The calculation unit 124 calculates the second coefficient representing the order importance. For example, since the important order is checked for the order ID "D2", the "second coefficient" is calculated as "1". Since the important orders are not checked for the order IDs other than "D2", the "second coefficient" is calculated as "0".

[0079] Fig. 13 is an example of the distribution result for input example (1). As shown in Fig. 13, the distribution result is obtained by allocating the quantity for each pair of order and supply.

[0080] Order IDs "D2" and "D5" have the same first coefficient for supply ID "S2", but because order ID "D2" is an important order, it is allocated an allocation of "30" that satisfies the order quantity. Here, an allocation that does not exceed the order quantity is performed, and in step S146 the constraints that the allocation quantity variables must satisfy are relaxed, so that supply IDs "S3" and "S4" have remaining quantities of "2" and "5", respectively.

[0081] If we summarize Figure 13 by order (D1 to D7), the allocation results from each supply for each order ID are obtained as distribution results, as shown in Figure 13. In the case of input example (1), order ID "D1" is allocated from supply ID "S1", and order ID "D2" is allocated from supply ID "S2".

[0082] 14 is another example of the distribution result for input example (1). If the maximum allowable deviation is allowed, the distribution result creation unit 126 creates the following distribution result with the remaining quantity of 0.

[0083] The order ID "D3" is allocated an allocation amount of "24" (=20×120%) from the supply ID "S4" (see the dashed line portion in FIG. 14(1)).

[0084] The order ID "D4" is allocated an allocation amount of "11" (=10 x 110%) from the supply ID "S3" (see the dashed line portion in Figure 14 (2)).

[0085] The order ID "D6" is allocated an allocation amount of "21" (=20×105%) from the supply ID "S3" (see the dashed line portion in FIG. 14(3)).

[0086] The order ID "D7" is allocated an allocation amount of "31" (=30×105%) from the supply ID "S4" (see the dashed line portion in FIG. 14(4)).

[0087] Next, we will explain the editing operations accepted by the distribution result editing unit 128. The distribution result editing unit 128 receives supply information, order information, a first coefficient, a second coefficient, and distribution results as input, and displays an editing screen on the user terminal 110. The user checks the distribution results displayed on the user terminal 110 and performs editing operations, and the distribution result editing unit 128 outputs distribution result editing information, which is the result of the editing operations.

[0088] The editing operation includes an operation to instruct the confirmation of a part of the allocation quantity included in the created predetermined distribution result, or an operation to instruct the change of a part of the allocation quantity, and the user can change the linkage between supply and order and the allocation amount. Specifically, the following editing operations (1) to (5) are possible.

[0089] (1) Adding or deleting supply and order linkages.

[0090] (2) Automatically set allocation amounts when linking supply and orders.

[0091] (3) Modifying quotas, including automatic and manual adjustments of quotas.

[0092] (4) Confirming part of the distribution results or setting part of the distribution results as NG.

[0093] (5) Cancellation of the created partial shipment results, including cancellation of the remaining portion excluding the confirmed portion.

[0094] The user can specify the following categories (1) to (4) for the created distribution results. It is not necessary to explicitly specify all of them, and if no individual specification is made, it may be determined as one of the following.

[0095] Category (1): Items that the user has determined to be final. This includes items that the user has corrected and finalized.

[0096] Category (2): The user indicates that the linking of supply and order is OK, but requests that the allocation amount be changed.

[0097] Category (3): The user has requested that the link destination be changed.

[0098] Category (4): Items that the user has indicated are not applicable.

[0099] If the date and time of the supply or order update is later than the date and time of the distribution result creation, the user is notified that the update has been made.

[0100] Furthermore, when a supply or order is added and the first and second coefficients are calculated, if there is a combination of unlinked supply and order whose first and second coefficients are higher than the coefficients of the existing linking, the system presents it as a linking proposal and receives instructions from the user as to whether to replace the existing linking, or whether to create the distribution results again.

[0101] You can also save the results of your shipments and load the results you have saved.

[0102] A first example of the editing screen is shown in Fig. 15. The dashed line portion (1) in Fig. 15 is a supply information addition / update button, which transitions to the first input section 112.

[0103] The dashed line portion in (2) of FIG. 15 is an order information addition / update button, which transitions to the second input section 114.

[0104] The dashed line portion (3) in FIG. 15 is a display area for the total number and remaining number, and displays the total number of supplies and orders and the remaining number.

[0105] The dotted line portion (4) in FIG. 15 is a display area for a list of supply information, and is displayed as a list by shipper and class.

[0106] The dotted line portion (5) in Figure 15 is a display area for supply information, which displays a supply list for each shipper. The "remaining quantity / arrival quantity" is displayed for all shippers and for each rank.

[0107] The dashed line portion (6) in Fig. 15 is a confirmation check, which is a mark for fixing the allocation amount at that time. For example, it is a mark such as a check box that indicates that the allocation amount has been confirmed.

[0108] The dashed line portion (7) in FIG. 15 is a display area for a list of order information, and displays a list of orders by orderer.

[0109] The dashed area (8) in Figure 15 is the display area for wholesaler information, displaying a list of orders by orderer. In this example, some orders have been split up, and orders that meet the required quantity are grayed out. The "allocation quantity / order quantity" is displayed for the entire wholesaler and for each class.

[0110] The dashed line portion (9) in FIG. 15 is a quantity allocation button, which is a button for instructing automatic calculation of the allocation number for the combination of linked supply information and order information.

[0111] The dashed line portion (10) in Figure 15 is the automatic distribution button, which is a button for instructing the creation of distribution results.

[0112] The dashed line portion (11) in FIG. 15 is a reset button, which is a button for canceling the linking operation.

[0113] The dashed line portion (12) in FIG. 15 is a save / read button, which is a button for saving and reading the distribution results.

[0114] The dashed portion (13) in FIG. 15 is a sales date selection button, which is a button for selecting the sales date.

[0115] Next, an example (1) of an editing operation on the first example of the editing screen will be described. Fig. 16 shows an example of accepting an editing operation.

[0116] In Figure 16, a link between supply and order is added. For example, by clicking on (1) Order and (2) Supply in Figure 16 in that order, the link is created, and the order information corresponding to the link between supply and order is entered in the position indicated by the dashed line in (3). Note that the order of clicking on (1) Order and (2) Supply can be reversed.

[0117] Also, in FIG. 16, allocation amounts may be automatically set at the time of pegging. In the example of FIG. 16, "0" is input as the initial value of the allocation amount for (4). The smaller of the order quantity "50" and the remaining supply quantity "30" may be input as the initial value. Here, it is preferable that the remaining supply quantity excludes the allocation amount for orders with the confirmed check box checked. For example, in the case of the supply of the dashed line portion of (5), the allocation amount "50" for (6) out of the quantity "80" has been confirmed, so the initial value is the smaller of the quantity "30" excluding that and the order quantity of the newly pegged order.

[0118] Next, an example of an editing operation (2) on the first example of the editing screen will be described. FIG. 17 shows an example of accepting an editing operation to instruct the deletion of a link between an order and a supply. In this editing operation, for example, by clicking on the order (1) and then pressing the Delete key on the keyboard, the link shown in (1) of FIG. 17 is deleted. Alternatively, by leaving the allocation amount (2) blank and pressing the Enter key on the keyboard, the link shown in (2) of FIG. 17 is deleted. Alternatively, by pressing an icon or button such as a trash can icon (3), the link shown in (3) of FIG. 17 is deleted.

[0119] Next, an example of an editing operation (3) on the first example of the editing screen will be described. Figure 18 shows an example in which an editing operation for manually correcting an allocation amount is accepted. For example, an editing operation for editing the allocation amount for a specified order (allocation amount (1) in Figure 18) is accepted. For example, the allocation amount (1) is clicked and then edited in the text input area. Alternatively, the amount can be increased or decreased by pressing the up and down cursor keys or by clicking the ▲ or ▼ button.

[0120] Next, an example of an editing operation (4) on the first example of the editing screen will be described. FIG. 19 shows an example of accepting an editing operation that automatically corrects the allocation amount. For example, if you click on the supply (1) and then on the button (2), the allocation amount for the order (3) linked to the specified supply is automatically calculated and registered. The allocation amount may be calculated using mathematical optimization, or may be distributed based on the ratio of the number of orders. It is also possible to specify the operating conditions (1) to (3) as shown below.

[0121] Operating condition (1): The number of orders is the upper limit of the allocation.

[0122] Operating condition (2): The upper limit of the allocated amount is set to the amount plus the deviation tolerance.

[0123] Operation condition (3): Calculate the allocation amount so that the remaining supply amount becomes 0.

[0124] Next, an example of an editing operation (5) on the first example of the editing screen will be described. FIG. 20 shows an example of accepting an editing operation to confirm a portion of the distribution results and set NG. Specifically, an editing operation to instruct the linking of a specified order and supply and to confirm the allocation amount is accepted. For example, clicking the check box (1). Also, an editing operation to specify that the linking of an order and supply is NG is accepted. For example, as shown in the dashed line portion of (2), setting the allocation amount of (3) to 0 and clicking the check box.

[0125] In addition, if the created distribution results do not match the user's intentions, an editing operation to cancel the distribution results is accepted. For example, by clicking button (4), all linkages between orders and supplies that have not been confirmed are deleted.

[0126] Next, an example of an editing operation (6) on the first example of the editing screen will be described. Fig. 21 shows an example of accepting an editing operation to specify the following categories (1) to (4) for the specified linkage and allocation amount.

[0127] Category (1): Items that the user has determined to be confirmed. This includes items that the user has corrected and confirmed. For example, this is indicated by clicking the check box (1) in FIG. 21.

[0128] Category (2): The user indicates that the linking of orders and supplies is OK, but that the allocation amount should be changed. For example, this is indicated by leaving the linking with the checkbox unclicked, as shown in the dashed line area (2) in Figure 21.

[0129] Category (3): The user requests a change in the link destination. For example, as shown in the dashed line portion of (3) in Figure 21, this is done by setting the allocation amount of (4) in Figure 21 to 0 and clicking the checkbox.

[0130] Category (4): Items that the user has specified as being excluded from the scope of the distribution results. For example, by marking a supply or order with an X, as shown in the dashed line areas (5) and (6) in Figure 21, the item will be excluded from the distribution results.

[0131] Next, an example (7) of an editing operation on the first example of the editing screen will be described. Figure 22 shows an example of notifying the user that a supply or order has been updated if the date and time of the update of the supply or order is later than the date and time of the distribution result creation. For example, icons (1) and (2) in Figure 22 are displayed for the supply and order, and the user can click the icon to confirm and make it disappear.

[0132] Furthermore, if there is a combination including an added or updated supply or order, and a combination with an unlinked supply or order, in which the first coefficient and the second coefficient are higher than the coefficient of the existing link, the corresponding combination is presented as an alternative linking plan, and an editing operation is accepted to indicate whether or not to adopt it, or whether or not to subject it to the re-creation of the distribution results.

[0133] For example, icons (3) and (4) in Fig. 22 are displayed for supply and order, and when the user clicks on the icon, a dialog like that in Fig. 22 (5) is displayed, asking whether to adopt another candidate, and when the user clicks the "Adopt" button, the other candidate replaces the existing link. The user can also remove the link in Fig. 22 (3) to make it the target for creating a new distribution result.

[0134] Next, an example of an editing operation (8) on the first example of the editing screen will be explained. Figure 23 shows an example of accepting an editing operation to instruct the re-creation of the distribution results. Specifically, this instructs the creation of distribution results that reflect the corrections and instructions described above. For example, the "Auto Distribution" button (1) in Figure 23 is clicked.

[0135] It also accepts editing operations to instruct the user to save the distribution results and to load the saved distribution results. Specifically, if the distribution results do not meet the user's intentions, the user can instruct the system to return to the saved state. For example, by clicking the "Save / Load" button (2) in Figure 23, a menu is displayed (not shown) that instructs the user to save or load the distribution results, and the user can instruct the system to save or load the distribution results from the menu.

[0136] Next, an example (9) of an editing operation on the first example of the editing screen will be described. Figure 24 shows an example of accepting editing operations to instruct filtering and sorting by attribute values. Specifically, editing operations to instruct filtering by attribute values ​​such as place of origin, grade, class, packaging style, and weight, and editing operations to instruct sorting are accepted. For example, the value to be narrowed down is selected in the [ ] parts of (1) and (2) of Figure 24, and sorting is instructed using the ▲ and ▼ buttons.

[0137] Furthermore, an editing operation to instruct the presentation of linking candidates is accepted. For example, by clicking on a supply and then on the "Linking Candidates" button in (3) of FIG. 24, an instruction is given to sort the orders in descending order of the first coefficient. Alternatively, by clicking on an order and then on the "Linking Candidates" button in (4) of FIG. 24, an instruction is given to sort the supplies in descending order of the first coefficient.

[0138] The display also accepts editing operations to narrow down incomplete supplies and orders. For example, clicking the "Incomplete" button as shown in (5) of Fig. 24 instructs the display of only supplies with remaining quantities. Alternatively, clicking the "Incomplete" button as shown in (6) of Fig. 24 instructs the display of only orders for which the order quantity has not been met.

[0139] Next, the process of the calculation unit 124 when an editing operation is received will be described.

[0140] In step S122, the CPU 11 calculates the first coefficient c for the i-th supply in the supply information and the j-th order in the order information using the acquired supply information, order information, importance of the attribute pair, and weight of each attribute, according to the above formula (1). ｉ，ｊ The first coefficient c ｉ，ｊ is the weight p of each attribute for each supply and order pair (i, j). ｋ The importance of attribute pairs is calculated using ｋ，ｉ，ｊ The weighted sum c ｉ，ｊ is obtained by taking

[0141] In this case, for each attribute including one or more of origin, grade, class, packing style, and weight, if specified by the user, it will take priority over the specification from the orderer. The weight of each attribute is given priority in the order of order, orderer, and total.

[0142] However, for pegging for which the user indicates that pegging is OK but the allocation amount should be changed, the first coefficient for other supplies pegged to the order may be adjusted to a lower value. In this case, if the order cannot be fulfilled using the pegging specified by the user, other supplies are allocated. For example, if the allocation amount for another pegging is manually increased and confirmed, and then automatic distribution is executed again, the allocation amount for the existing pegging is changed to a smaller amount. As a result, it may become impossible to fulfill the order using only the existing pegging. By allowing pegging other than the pegging specified by the user, it is possible to create a distribution result that will fulfill the order even in such cases. If you do not want to peg other supplies, you may use the method described in paragraph

[0153] below. The third input unit 116 may be used to specify which operation to perform.

[0143] Next, the processing of the distribution result creation unit 126 when an editing operation is accepted will be described.

[0144] In step S140, the CPU 11 calculates the allocation amount to be calculated by using the variable q of the allocation amount from the i-th supply to the j-th order. ｉ，ｊ However, supplies and orders that the user has indicated are not included are excluded.

[0145] In step S144, the CPU 11 sets (or updates) the tolerance for deviation of the allocation amount from the order quantity. Here, the tolerance is prioritized in the order of per order, per orderer, and overall. As described below, the third input unit 116 may be configured to allow the user to specify operation conditions (1) to (3).

[0146] Operating condition (1): The deviation tolerance is set to 0, and the number of orders is set as the upper limit of the allocation amount.

[0147] Operating condition (2): The upper limit of the allocated amount is set to the amount plus the deviation tolerance.

[0148] Operating condition (3): Allocate so that the remaining supply amount becomes 0.

[0149] In step S146, the CPU 11 calculates the allocation variable q ｉ，ｊ The constraints that must be satisfied are defined as follows [1] to [4].

[0150] Constraint [1]: Allocation variable q ｉ，ｊ is greater than or equal to zero for all i and j. However, when an editing operation is accepted, the allocation variable q ｉ，ｊ Reflected in.

[0151] Specifically, the items that the user has determined to be finalized and the items that the user has corrected and finalized are the allocation variable q ｉ，ｊ equate to the determined value.

[0152] In addition, if the user requests a change in the link destination, the allocation variable q ｉ，ｊ is set to zero.

[0153] In addition, for the linking where the user has instructed that the linking is OK but the allocation amount should be changed, the variable q ｉ，ｊ may be set to zero. In this case, even if the order cannot be fulfilled using the pegging specified by the user, other supplies will not be allocated. If other supplies can be allocated, it is preferable to do as described in paragraph

[0142] above. The third input unit 116 may be configured to allow the user to specify which operation to perform.

[0154] Constraint [2]: Allocation variable q ｉ，ｊThe sum of the quantities for j is equal to the i-th supply. However, if the number of orders plus the allowance for discrepancies is significantly less than the total supply, resulting in an excess supply, the supply may be allowed to fall below the supply.

[0155] Constraint [3]: Allocation variable q ｉ，ｊ The difference between the total amount of i related to the jth order, that is, the amount allocated to the jth order, and the order quantity is equal to or less than the tolerance for deviation.

[0156] Constraint [4]: ​​Second coefficient d ｊ If is 1, the allocation to the jth order is set equal to the order quantity.

[0157] Next, an example of the operation of the distribution result creation unit 126 when an editing operation is accepted will be described.

[0158] In the distribution result editing unit 128, as shown in FIG. 25, when a user accepts an editing operation instructing that the link destination of order ID "D1" be changed from supply ID "S1" to supply ID "S4" and the allocation amount be changed, and if this is to be handled by adjusting the first coefficient, for example, the following processes (1) and (2) are performed.

[0159] Process (1): If the first coefficient of supply ID "S1" with the largest allocation amount for order ID "D1" is greater than the first coefficient of supply ID "S4" to which the order will be changed, the first coefficient of supply ID "S1" must be made smaller than the first coefficient of supply ID "S4." Therefore, a ratio r is calculated to make the first coefficient of the supply to a certain percentage (e.g., 80%) of the first coefficient of the supply to which the order will be changed.

[0160] r=5.5×80% / 6.0=0.73

[0161] Process (2): The first coefficients other than those for supply ID "S4" are multiplied by the ratio r to obtain new first coefficients (see the dashed line in Figure 25). First coefficient for supply ID "S1": 6.0 x 0.73 = 4.3 (rounded down to two decimal places) First coefficient for supply ID "S2": 5.0 x 0.73 = 3.6 (rounded down to two decimal places) First coefficient for supply ID "S3": 3.5 x 0.73 = 2.5 (rounded down to two decimal places)

[0162] Next, an operational procedure for presenting important attribute candidates in the second input unit 114 will be described.

[0163] Priority given to attributes such as place of origin, grade, class, packaging style, and weight varies depending on the orderer, and may also differ between orders even for the same orderer. This is because when the orderer is an intermediate wholesaler, the orderer, such as a retailer, may be different.

[0164] Therefore, it would be ideal if important attributes could be set in detail for each orderer or order. However, setting detailed settings for many orderers or orders would be time-consuming.

[0165] Therefore, in this embodiment, important attributes that are applied to the whole as distribution conditions and important attributes that can be set individually for orderers or orders are provided, and priority is given to the order, orderer, and the whole in that order.

[0166] Furthermore, the learning unit 130 learns important attributes for orderers or orders based on past editing operations by the user. Specifically, the learning unit 130 learns important attributes from the history of linkages changed by the user in the distribution result editing unit 128, and presents the learned important attributes to the user. The user may be allowed to decide whether to adopt the presented important attributes, or attributes with a high degree of accuracy may be automatically set and the user may be allowed to cancel the setting.

[0167] Specifically, for each history in which the user has changed the association, the important attribute that maximizes the first coefficient c_i,j in the changed association is selected, and the important attribute that has been frequently selected as the important attribute is presented, or the important attribute selected from the most recent history is presented.

[0168] An example of the process by which the learning unit 130 learns important attributes will be described below with reference to Figure 26. For example, when a distribution result is created for the order ID "D8" in Figure 26 (1) with the important attribute of the distribution conditions set to "Place of Origin" and without specifying the orderer or important order attributes, the order ID is linked to the supply ID "S4" (see the dashed line in Figure 26 (3)) with the maximum first coefficient of 6.0 shown in Figure 26 (2), and then the distribution result editing unit 128 receives an editing operation from the user to change the link to the supply ID "S2" (see the dashed line in Figure 26 (4)).

[0169] For each of the important attributes (place of origin, grade, class), the first coefficient between order ID "D8" and supply ID "S2" in the changed linkage is calculated. Regarding the first coefficient between order ID "D8" and supply ID "S2", the calculation result when the important attribute is "place of origin" is 4.5, the calculation result when the important attribute is "grade" is 5.5, and the calculation result when the important attribute is "class" is 4.5, with the first coefficient when the important attribute is "grade" being the largest.

[0170] When the important attribute is "grade," the first coefficient with supply ID "S2" is the highest among supply IDs "S1" to "S4," and there is a high possibility that it will be linked to supply ID "S2."

[0171] In such cases, the important attribute "grade" is learned and becomes a candidate for presentation. By setting this as an important attribute for the orderer or order, it may be possible to obtain the desired results from the next time without having to modify the distribution results.

[0172] The accuracy of important attributes obtained as a learning result is calculated using, for example, the weighted sum q_1×f+q_2×r+q_3×c of the appearance frequency in the most recent fixed period f, the number of days since the last appearance r, and the appearance frequency c in the same period of the previous year. The pattern with the largest value or exceeding the threshold is presented as the learning result. [q_1, q_2, q_3] are weights, and if [q_1, q_2, q_3] = [1, -1, 1], the greater the appearance frequency in the most recent fixed period f or the appearance frequency c in the same period of the previous year, or the smaller the number of days since the last appearance r, the greater the contribution.

[0173] Next, an operational procedure for presenting candidates for deviation tolerance in the second input unit 114 will be described.

[0174] When there is a shipment of goods exceeding the ordered quantity, the degree to which the excess amount is accepted varies depending on the orderer, and even for the same orderer, it may vary depending on the order. This is because when the orderer is a wholesaler, the orderer, such as a retailer, may be different.

[0175] Therefore, it would be ideal if the allowable deviation from the order quantity for the allocation amount could be set in detail for each orderer or each order. However, setting detailed settings for many orderers and orders would be time-consuming.

[0176] Therefore, as a distribution condition, a deviation tolerance that is applied to the whole and a deviation tolerance that can be set individually for each orderer or order are set, and priority is given to the order, orderer, and the whole in that order.

[0177] Furthermore, the learning unit 130 learns and presents deviation tolerances from the history of editing operations in which the user changes the allocation amount, which are received by the distribution result editing unit 128. The user may decide whether to adopt the presented deviation tolerances, or a highly accurate tolerance may be automatically set and the user may be allowed to cancel the setting.

[0178] Specifically, for each editing operation history in which the user changed the allocation amount, the system calculates the difference between the order quantity and the allocation amount, and learns and presents the tolerance for the difference, calculated taking into account frequency and the most recent time of use.

[0179] Next, an example of the process in which the learning unit 130 learns the tolerance for deviation will be described.

[0180] For example, suppose the order quantity is "40," the tolerance for deviation in the distribution conditions is "5%," and no tolerance for deviation is specified for each orderer or order. After the distribution result creation unit 126 allocates the maximum quantity of "42," the distribution result editing unit 128 accepts an editing operation to correct the allocated amount to "44."

[0181] In this case, the learned result of the deviation tolerance is (corrected allocation amount - number of orders) / number of orders = (44 - 40) / 40 = 10%.

[0182] The accuracy of deviation tolerance obtained as a learning result is calculated using, for example, the weighted sum q_1×f+q_2×r+q_3×c of the appearance frequency in the most recent fixed period f, the number of days since the last appearance r, and the appearance frequency c in the same period of the previous year. The pattern with the largest weight, or a pattern that exceeds the threshold, is presented as the learning result. [q_1, q_2, q_3] are weights, and if [q_1, q_2, q_3] = [1, -1, 1], the greater the appearance frequency in the most recent fixed period f or the appearance frequency c in the same period of the previous year, or the smaller the number of days since the last appearance r, the greater the contribution.

[0183] As described above, the loading support device 100 of this embodiment accepts user editing operations, including operations to confirm or change some of the allocation quantities included in the created predetermined loading result, changes the first coefficient according to the editing operation, and recreates the predetermined loading result, excluding the allocation quantities instructed to be confirmed or changed based on the editing operation. This creates high-quality loading results and reduces the workload associated with loading. Furthermore, loading results that take various circumstances into account are created, reducing the workload associated with loading.

[0184] Furthermore, by presenting the user with the results of the distribution that reflect the distribution conditions specified by the user, the workload associated with the distribution can be reduced.

[0185] Furthermore, by repeatedly creating predetermined loading results and changing parameters according to the situation, the quality of the loading results can be improved and the burden of loading work can be reduced. Furthermore, since parameters can be adjusted after the system has started operation, the system implementation period can be shortened.

[0186] In addition, the created distribution results can be saved and saved distribution results can be loaded.

[0187] In addition, when supply information or order information is newly added or updated, a proposal to overwrite the already created distribution results can be presented.

[0188] Furthermore, by using the operation history in which the user has confirmed or corrected an order, important attributes for the orderer or order can be learned and presented to the user.

[0189] In addition, the order information can be overwritten or supplemented with information specified by the user. Furthermore, the history of past overwriting or supplementation by the user can be applied to new order information.

[0190] In addition, when entering order information, frequently used or most recently used attributes, including designated origin, designated grade, designated class, and designated packaging style, are presented as templates, which can be used or modified to enter the order information.

[0191] In addition, the allowable deviation of the allocation amount from the order quantity can be learned from the history of the allocation amount changed by the user in the distribution result editing section, and can be presented to the user.

[0192] [Second Embodiment] The second embodiment is an aspect in which the first coefficient is calculated using history information. The calculation unit 124 may calculate the first coefficient by further using history information. In calculating the weighted sum related to the first coefficient, if history information exists for the ith supply in the supply information and the jth order in the order information, the history information is taken into consideration as shown in the following formula (3). The history information is expressed as a frequency and an interval (frequency, recency) as (f ｉ，ｊ , r ｉ，ｊ )

[0193] c ｉ，ｊ '=c ｉ，ｊ +q １ ×f ｉ，ｊ +q ２ ×r ｉ，ｊ ...(3)

[0194] Here, [q １ , q ２ ] is the weight and [q １ , q ２]=[1, -1], the greater the frequency and the smaller the recency, the greater the contribution. ｉ，ｊ ) by adding history information (frequency and interval weighted to adjust the contribution rate) to ｉ，ｊ As described above, when there is history information for a supply-order pair, the calculation unit 124 of the second embodiment calculates the first coefficient by adding values ​​related to the frequency and interval in the history information and calculating a weighted sum.

[0195] [Other Modifications] Although the weights of the attributes are "4, 2, 1" in the example shown, they are not limited to this and may be set to any ratio, such as [100, 10, 1].

[0196] Regarding the distribution result creation unit 126, the allocation amount q from the i-th supply to the j-th order ｉ，ｊ is calculated using a mathematical optimization technique, but any other technique that optimizes (maximizes or minimizes) the objective function may be applied.

[0197] In the above-described embodiment, the importance of an attribute pair is set based on the degree of matching between the attribute pair, i.e., "1.0 if the supply and order are the same, and 0.5 if they are different." However, this is not limiting. For example, the importance of an attribute pair may be assigned to each attribute according to the desired conditions. For example, as shown in the example of FIG. 27 , the importance may be set lower as the difference in rank increases. FIG. 27 illustrates an example in which, for an order rank "M," when the supply rank is "S," which is lower than rank "M," and "L," which is higher than rank "M," the importance of the attribute pair is the same, i.e., "0.5." Furthermore, as shown in the example of FIG. 28 , the importance of an attribute pair for a lower rank may be set lower than that for a higher rank. FIG. 28 illustrates an example in which the importance of an attribute pair decreases as the difference in rank increases. Also, Figure 28 shows an example in which the importance of an attribute pair is lower when the supply is of grade "C" below grade "B" compared to when the supply is of grade "A" above grade "B" for an order of grade "B."

[0198] Furthermore, as shown in FIG. 29 , for items that are equivalent but have different notations, the importance of the attribute pair may be assigned a slightly lower value to compensate for the notational differences and prioritize items with the same notation. FIG. 29 illustrates an example in which the importance of an attribute pair is slightly lower when the order grades are "A," "B," and "C" and the supply grades are "Excellent," "Excellent," and "Good" compared to when the order grades are "A," "B," and "C." This does not need to be achieved with a single table; as shown in the examples of FIGS. 30 and 31 , a conversion table for converting attributes (in this example, grades) to ranks and a table using ranks may be prepared in advance. Furthermore, assigning a slightly lower value to items that are equivalent but have different notations means that the value is reduced when the order and supply are in different groups. For example, if the order grade is "A" and the supply grade is "Excellent," the order grade "A" would have a rank of "1" and a group of "1" (see the dashed line on the left side of FIG. 30 ). Also, the supply grade "Excellent" has a rank of "2" and a group of "2" (see the dashed line on the right side of Figure 30). Based on the table in Figure 31, the importance of the attribute pair is "0.5" (see the dashed line in Figure 31), but since the groups are different, 0.1 is subtracted to make it "0.4".

[0199] Furthermore, the editing screen used when editing the distribution results in the fourth input unit 116 may be in a table format as shown in FIG. 32. On this editing screen, orders are arranged on the vertical axis and supplies on the horizontal axis, and the allocation quantity is entered in the column indicated by the dashed line in (1) of FIG. 32, allowing for simultaneous linking and quantity entry. A check box for confirmation may be provided in column (1) of FIG. 32. Since the table is expected to become large, a scroll bar may be provided. Functions such as quantity allocation, automatic distribution, reset, save / load, filter, and sort, as shown in the above embodiment, may also be provided.

[0200] Also, the number of items displayed in the table on the editing screen may be increased or decreased as shown in Fig. 33. In Fig. 33, the important attributes and tolerances are displayed in the dashed line portion of (1), and the first coefficient is displayed in the dashed line portion of (2), so that the items can be used as reference when manually dividing the load.

[0201] In addition, in the example of Fig. 32, orders are arranged on the vertical axis and supplies on the horizontal axis, but supplies may be arranged on the vertical axis and orders on the horizontal axis as in Fig. 34. The data may be grouped and displayed by shipper or orderer, or the table may be divided.

[0202] In the above embodiments, the support processing executed by the CPU after reading the software (program) may be executed by various processors other than the CPU. Examples of such processors include programmable logic devices (PLDs) whose circuit configuration can be changed after manufacture, such as field-programmable gate arrays (FPGAs), graphics processing units (GPUs), and dedicated electrical circuits, such as application-specific integrated circuits (ASICs), which are processors having a circuit configuration specifically designed to execute specific processing. The support processing may be executed by one of these various processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, a combination of a CPU and an FPGA, etc.). Furthermore, the hardware structure of these various processors is, more specifically, an electric circuit that combines circuit elements such as semiconductor elements.

[0203] In the above embodiment, the program is pre-stored (installed) in the storage 14, but the present invention is not limited to this. The program may be provided in a form stored on a non-transitory storage medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. The program may also be downloaded from an external device via a network.

[0204] The following additional notes are provided regarding the above-described embodiments.

[0205] (Supplementary Item 1) A system comprising: a memory; and at least one processor connected to the memory, wherein the processor: acquires, as information for allocating and distributing products to be supplied to orders, supply information including a plurality of product attributes and quantities for the i-th supply from the supplier, order information including a plurality of product attributes and quantities for the j-th order from the orderer, and distribution conditions related to the importance of each of the attributes and the order, based on the acquired supply information, order information, and distribution conditions, an index corresponding to the distribution conditions from a database storing predetermined indexes related to attributes, and calculates, using the acquired index, a first coefficient related to a combination of supply and order taking into account the importance of the attribute and a second coefficient related to the importance of the order, creates a predetermined distribution result including an allocation quantity of the product in the supply information for the order information based on the supply information, order information, first coefficient, and second coefficient, accepts a user's editing operation including an operation to instruct to confirm a portion of the allocation quantity included in the created predetermined distribution result, or an operation to instruct to change a portion of the allocation quantity, and changes the first coefficient in accordance with the editing operation, A distribution support device configured to recreate the specified distribution result based on the editing operation, excluding the allocation quantity that has been instructed to be confirmed or the allocation quantity that has been instructed to be changed.

[0206] (Supplementary Item 2) A non-transitory storage medium storing a program executable by a computer to execute support processing, wherein the support processing comprises: acquiring supply information including a plurality of attributes and quantities of the product for the i-th supply from the supplier, order information including a plurality of attributes and quantities of the product for the j-th order from the ordering party, and distribution conditions related to the importance of each of the attributes and the order, as information for allocating and distributing products to be supplied to orders; acquiring indicators corresponding to the distribution conditions from a database storing predetermined indicators related to attributes based on the acquired supply information, order information, and distribution conditions, and using the acquired indicators, calculating a first coefficient related to a combination of supply and order taking into account the importance of the attribute and a second coefficient related to the importance of the order; creating a predetermined distribution result including an allocation quantity of the product in the supply information for the order information based on the supply information, order information, first coefficient, and second coefficient; accepting editing operations from a user including an operation to instruct to confirm a portion of the allocation quantities included in the created predetermined distribution result, or an operation to instruct to change a portion of the allocation quantities; A non-transitory storage medium that changes the first coefficient in accordance with the editing operation, and recreates the specified distribution result based on the editing operation, excluding the allocation quantity instructed to be confirmed or the allocation quantity instructed to be changed.

[0207] REFERENCE SIGNS LIST 1 Support system 11 CPU 12 ROM 13 RAM 14 Storage 15 Input unit 16 Display interface (I / F) 17 Communication interface (I / F) 19 Bus 100 Load distribution support device 110 User terminal 112 First input unit 114 Second input unit 116 Third input unit 118 Fourth input unit 120 Acquisition unit 122 Database 124 Calculation unit 126 Load distribution result creation unit 128 Load distribution result editing unit 130 Learning unit

Claims

1. An acquisition unit acquires, as information for allocating and loading products to orders, supply information including multiple attributes and quantities of the product for the i-th supply from the supplier, order information including multiple attributes and quantities of the product for the j-th order from the orderer, and loading conditions related to the importance of each of the attributes and the order; a calculation unit acquires, based on the acquired supply information, order information, and loading conditions, indicators corresponding to the loading conditions from a database storing predetermined indicators related to attributes, and calculates, using the acquired indicators, a first coefficient related to a combination of supply and order taking into account the importance of the attribute and a second coefficient related to the importance of the order; a loading result creation unit creates, based on the supply information, order information, the first coefficient, and the second coefficient, a predetermined loading result including the allocation quantities of the product in the supply information for the order information; and a loading result editing unit that accepts editing operations from a user, including an operation to instruct to confirm a portion of the allocation quantities included in the created predetermined loading result, or an operation to instruct to change a portion of the allocation quantities, wherein the calculation unit changes the first coefficient in accordance with the editing operation, The distribution result creation unit is a distribution support device that recreates the specified distribution result based on the editing operation, excluding the allocation quantity that has been instructed to be confirmed or the allocation quantity that has been instructed to be changed.

2. The acquisition unit further acquires the updated contents of the supply information, the added supply information, the updated contents of the order information, or the added order information; the calculation unit further calculates the first coefficient and the second coefficient for the updated or added supply information or order information; and the distribution result editing unit, if there is updated or added supply information or order information after the specified distribution result is created, notifies the user of the supply and order combination to be allocated based on the first coefficient, and accepts from the user an instruction as to whether to adopt the supply and order combination to be allocated, or an instruction as to whether to create the specified distribution result again.

3. The distribution conditions include the most important attribute among the plurality of attributes, and the distribution support device described in claim 1 further includes a learning unit that learns the important attributes for an orderer or an order based on the past editing operations by the user, and the calculation unit calculates the first coefficient and the second coefficient based on the distribution conditions including the learned important attribute.

4. The distribution support device according to claim 1, wherein the acquisition unit acquires the order information input by the orderer by overwriting the attribute and quantity input by the user.

5. The distribution support device according to claim 1, wherein the acquisition unit presents to the orderer patterns of past order information input by the orderer, patterns whose appearance frequency or a value based on the number of days since the last appearance exceeds a threshold, and acquires the order information using the patterns.

6. The distribution conditions include a tolerance for deviation between the quantity of the order information and the allocation quantity assigned to the order information, and the distribution support device described in claim 1 further includes a learning unit that learns the tolerance based on past editing operations by the user, and the acquisition unit presents the learned tolerance to the user to acquire the distribution conditions including the tolerance.

7. Acquire supply information including multiple attributes and quantities of the product for the i-th supply from the supplier, order information including multiple attributes and quantities of the product for the j-th order from the ordering party, and distribution conditions related to the importance of each of the attributes and orders, as information for allocating and distributing the products to be supplied to orders; acquire indicators corresponding to the distribution conditions from a database storing predetermined indicators related to attributes based on the acquired supply information, order information, and distribution conditions; calculate a first coefficient related to the combination of supply and order taking into account the importance of the attributes and a second coefficient related to the importance of the order using the acquired indicators; create a predetermined distribution result including the allocation quantities of the products in the supply information for the order information based on the supply information, order information, first coefficient, and second coefficient; accept user editing operations including an operation to instruct the confirmation of a portion of the allocation quantities included in the created predetermined distribution result, or an operation to instruct the change of a portion of the allocation quantities; change the first coefficient in accordance with the edit operation; A distribution support method in which a computer executes a process to recreate the specified distribution result, excluding the allocation quantity instructed to be confirmed or the allocation quantity instructed to be changed, based on the editing operation.

8. A program for causing a computer to function as each part of the distribution support device according to any one of claims 1 to 6.

Images