Shelf allocation inference device, shelf allocation inference method, and recording medium

Abstract

This shelf allocation inference device includes an acquisition unit, an inference unit, and an output unit. The acquisition unit acquires products to be displayed on a shelf and the classification of each product. The inference unit infers, on the basis of an index related to a position on the shelf corresponding to the classification of each product, an index related to products that have different classifications and are sold in parallel, and an index related to the position on the shelf in which products having the same classification are to be displayed, a shelf allocation with which an evaluation value related to sales of the products is maximized when the products are displayed on the shelf by classification. The output unit outputs the inferred shelf allocation.

Description

Shelf allocation estimation device, shelf allocation estimation method, and recording medium

[0001] The present disclosure relates to a shelf allocation estimation device and the like.

[0002] The display position of products in a store can have a significant impact on sales. For this reason, planogram creators create shelf layouts based on, for example, sales performance data. Furthermore, shelf layout creation may involve the use of systems that support product layout creation.

[0003] The automatic shelving planogram creation device of Patent Document 1 generates a learning model based on shelving planogram map data that includes product information and information about the shelves on which the products are displayed.The automatic shelving planogram creation device of Patent Document 1 then uses the learning model to create shelving planogram map data from the product information.

[0004] Japanese Patent Application Laid-Open No. 2021-121904

[0005] The technology described in Patent Document 1 may not be able to estimate shelf layouts with sufficient accuracy.

[0006] In order to solve the above-mentioned problems, the present disclosure aims to provide a shelf layout estimation device and the like that can improve the accuracy of shelf layout estimation.

[0007] In order to solve the above problems, the shelf allocation estimation device of the present disclosure includes an acquisition means for acquiring the products to be displayed on shelves and the classification of each product, an estimation means for estimating the shelf allocation that maximizes the evaluation value related to product sales when the products are displayed on shelves by classification, based on an index related to the shelf position according to the product classification, an index related to products that are sold side by side in different classifications, and an index related to the shelf position where products of the same classification are displayed, and an output means for outputting the estimated shelf allocation.

[0008] The shelf allocation estimation method disclosed herein obtains the products to be displayed on shelves and their respective classifications, and estimates the shelf allocation that maximizes the evaluation value related to product sales when the products are displayed on shelves by classification based on an index related to the shelf position according to the product classification, an index related to products that are sold side by side but in different classifications, and an index related to the shelf position where products of the same classification are displayed, and outputs the estimated shelf allocation.

[0009] The recording medium of the present disclosure non-temporarily records a shelf allocation estimation program that causes a computer to execute the following processes: a process of acquiring the products to be displayed on shelves and the classification of each product; a process of estimating the shelf allocation that maximizes the evaluation value related to product sales when the products are displayed on shelves by classification, based on an index related to the shelf position according to the classification of the products, an index related to products that are sold side by side in different classifications, and an index related to the shelf position where products of the same classification are displayed; and a process of outputting the estimated shelf allocation.

[0010] According to the present disclosure, it is possible to improve the accuracy of estimating shelf layout.

[0011] FIG. 1 is a diagram illustrating an example of the configuration of a shelf allocation estimation system according to the present disclosure. FIG. 2 is a diagram illustrating an example of a map of shelves in a store according to the present disclosure. FIG. 3 is a diagram illustrating an example of the configuration of a shelf allocation estimation device according to the present disclosure. FIG. 4 is a diagram illustrating an example of product classification according to the present disclosure. FIG. 5 is a diagram illustrating an example of sales priority data according to the present disclosure. FIG. 6 is a diagram illustrating an example of position weight data according to the present disclosure. FIG. 7 is a diagram illustrating an example of data in which classifications are assigned to shelves according to the present disclosure. FIG. 8 is a diagram illustrating an example of classification assignment according to the present disclosure. FIG. 9 is a diagram illustrating an example of a shelf allocation estimation result according to the present disclosure. FIG. 10 is a diagram illustrating an example of a sales prediction result according to the present disclosure. FIG. 11 is a diagram illustrating an example of the operation flow of a shelf allocation estimation device according to the present disclosure. FIG. 12 is a diagram illustrating an example of the hardware configuration of a shelf allocation estimation device according to the present disclosure.

[0012] An embodiment of the present disclosure will be described in detail with reference to the drawings. FIG. 1 is a diagram illustrating an example of the configuration of a shelf allocation estimation system. The shelf allocation estimation system includes, for example, a shelf allocation estimation device 10, a management device 20, and a terminal device 30. The shelf allocation estimation device 10 is connected to the management device 20, for example, via a network. The shelf allocation estimation device 10 is also connected to the terminal device 30, for example, via a network. There may be a plurality of management devices 20 and a plurality of terminal devices 30. The number of management devices 20 and the number of terminal devices 30 may be set as appropriate.

[0013] A shelf allocation estimation system is, for example, a system that estimates shelf allocation within a store. For example, the shelf allocation estimation system estimates a shelf allocation that optimizes the evaluation value related to sales of products displayed on shelves in the store. The optimal value may include values ​​close to the optimal value. Furthermore, the evaluation value related to product sales is, for example, an index indicating how well a product has been sold. In this case, the shelf allocation estimation system estimates, for example, a shelf allocation that maximizes the evaluation value related to sales of products displayed on shelves in the store. Product sales are, for example, the total price of products sold in the store or the total number of products sold. In other words, the evaluation value related to product sales is, for example, an index indicating the total price of products sold or the total number of products sold. The evaluation value related to product sales may be, for example, the sales amount or number of products sold. Furthermore, the evaluation value related to product sales may include an index indicating the number of customers visiting the store.

[0014] Furthermore, the shelf allocation estimation system estimates shelf allocations based on, for example, the classification of products. That is, the shelf allocation estimation system estimates a shelf allocation that maximizes the evaluation value when, for example, one classification is assigned to one shelf. Products belonging to the assigned classification are displayed on each shelf. The same classification may also be assigned to multiple shelves. For example, if there are three shelves, one shelf may be assigned to classification A and the remaining two shelves may be assigned to classification B.

[0015] Fig. 2 is a diagram showing an example of a map showing the layout positions of shelves within a store. In the example of Fig. 2, a counter for product checkout is installed at the entrance to the store. Sixteen shelves are arranged with shelf IDs 1 to 16. In the example of Fig. 2, the shelf allocation estimation device 10 estimates, as the shelf allocation, the product categories to be assigned to each of the shelves with shelf IDs 1 to 16, for example, so as to maximize the evaluation value related to product sales.

[0016] Here, an example of the configuration of the shelf allocation estimation device 10 will be described. Fig. 3 is a diagram showing an example of the configuration of the shelf allocation estimation device 10. The shelf allocation estimation device 10 basically includes an acquisition unit 11, an estimation unit 13, and an output unit 15. The shelf allocation estimation device 10 may further include, for example, a classification generation unit 12, a sales forecasting unit 14, and a storage unit 16.

[0017] The acquisition unit 11 acquires the products to be displayed on the shelves and the classifications of the products. For example, the acquisition unit 11 acquires the product names of the products to be displayed on the shelves and the classification names of the classifications of the products.

[0018] Product classifications may be hierarchical. "Hierarchical" refers to a hierarchy in which a higher-level classification includes a lower-level classification. For example, "hierarchical" refers to a hierarchy in which a higher-level classification is a superordinate concept of a lower-level classification. For example, a product A that is beer may be hierarchically classified as "beverages," "alcoholic beverages," and "beer." In this case, "beverages" includes "alcoholic beverages," so "beverages" is a classification in a higher level than "alcoholic beverages." Furthermore, "alcoholic beverages" includes "beer," so "alcoholic beverages" is a classification in a higher level than "beer." Furthermore, "hierarchical" may refer to a hierarchy in which a higher-level classification is a classification that integrates multiple lower-level classifications. For example, a product A that is beer may be classified as "beer_cocktail" in a higher level and as "beer" in a lower level. In this case, since "beer_cocktail" is a classification that combines "beer" and "cocktail," "beer_cocktail" is a classification in a higher hierarchy than "beer" and "cocktail." The acquisition unit 11 may acquire only some of the hierarchy of the hierarchy of the hierarchy. For example, the acquisition unit 11 may acquire only "alcoholic beverage" from the hierarchy of the classification of product A, such as "beverage," "alcoholic beverage," and "beer." When the product classification is hierarchical, the number of hierarchy levels can be set as appropriate. The acquisition unit 11 acquires, for example, from the terminal device 30, the products to be displayed on the shelves and the classifications of each product.

[0019] The acquisition unit 11 may acquire, as designation information, information that designates the classification of products to be displayed on at least one shelf out of the multiple shelves that are the subject of shelving planogram calculation. The information that designates the classification of products to be displayed is information that designates a shelf to be treated as a fixed shelf in the shelving planogram estimation process by the estimation unit 13. In other words, the designation information is, for example, information that indicates a fixed shelf. A fixed shelf is a shelf for which classification allocation does not change in, for example, the mathematical optimization process performed to estimate the shelving planogram. The mathematical optimization process is, for example, a process that estimates the shelving planogram that maximizes the evaluation value related to product sales. Furthermore, a shelf for which classification allocation changes in the mathematical optimization process performed to estimate the shelving planogram is called, for example, a variable shelf.

[0020] The designation information includes, for example, a shelf identifier that designates the category of the product to be displayed and the category of the product to be displayed on the shelf. For example, in the example of Figure 2, if there is a product in the category of "beverages" that needs to be displayed on a shelf with a refrigeration function, and the shelf with "shelf ID 4" has a refrigeration function, the designation information is information indicating that "beverages" should be assigned to "shelf ID 4." The acquisition unit 11 acquires the designation information from, for example, the terminal device 30.

[0021] The acquisition unit 11 may further acquire information related to product sales. The information related to product sales is, for example, the sales amount or sales quantity of each product. Furthermore, the information related to product sales may further include information related to products purchased at the same time. Information related to products purchased at the same time is also referred to as cross-sale information. Information related to products purchased at the same time is, for example, information indicating products purchased at the same time. "Purchased at the same time" refers to, for example, multiple products being purchased in a single transaction. In this case, for example, products purchased at the same time are products listed on a single receipt. "Purchased at the same time" may also refer to multiple products being purchased during a single visit to the store. Furthermore, the information related to product sales is not limited to the above. The acquisition unit 11 acquires information related to product sales from, for example, the management device 20.

[0022] The acquisition unit 11 may acquire, as change information, information indicating a shelf to which an assigned classification is to be changed in the shelving planogram estimated by the estimation unit 13. The change information is, for example, information for changing the classification of at least one shelf from the shelving planogram estimated by the estimation unit 13. For example, the change information includes information specifying a shelf to which a classification is to be changed and the changed classification of the shelf. The change information may be information indicating the details of the classification swap between two shelves. In this case, the change information includes, for example, information specifying a shelf to which a classification is to be swapped. The change information may be information indicating the details of the classification swap between three or more shelves. In this case, the change information includes, for example, information specifying a shelf to which a classification is to be changed and the changed classification of the shelf. The acquisition unit 11 acquires the change information from, for example, the terminal device 30.

[0023] The category generation unit 12, for example, combines at least two or more product categories to generate a new category. By generating a new category, for example, each of the products that belonged to two or more categories before the combination becomes a product that belongs to the combined category. In the process of combining categories, the category generation unit 12 combines the categories so that the sum of the number of combined categories and the number of uncombined categories is the same as the number of shelves. Furthermore, when one category is assigned to two or more shelves, the sum of the number of uncombined categories may be smaller than the number of shelves. For example, when products in category S are displayed on three shelves, the category generation unit 12 considers category S to be three categories and combines categories other than category S so that the sum of the number of categories is the same as the number of shelves.

[0024] For example, when the number of product shelves is smaller than the number of product categories, the category generation unit 12 combines two or more categories to generate a new category. In the process of generating a new category, the category generation unit 12 generates the new category based on, for example, at least one of the similarity of the product categories and the sales performance of the products. When generating a new category based on the similarity of the product categories, the category generation unit 12 generates, for example, a category using elements common to the categories to be combined as the new category. For example, the category generation unit 12 generates, as the new category, a category that is a superordinate concept of the categories to be combined. For example, assume that there is a product A that is categorized as "beer" and a product B that is categorized as "cocktail." In this case, the category generation unit 12 combines, for example, "beer" and "cocktail" to generate a category called "alcoholic beverages." When the category "alcoholic beverages" is generated, product A and product B become products classified as "alcoholic beverages."

[0025] When generating a new category based on the sales performance of a product, the category generation unit 12 generates a new category, for example, by integrating categories with lower sales than other categories. For example, assume that there is a product C categorized as "dry foods" and a product D categorized as "canned foods," and that these products have lower sales performance than the products in the other categories. In this case, the category generation unit 12 integrates, for example, "dry foods" and "canned foods" to generate a category called "preserved foods."

[0026] The category generation unit 12 may also generate a new category by integrating categories with a small number of products. For example, if there is a category with fewer products than one shelf because the store handles a small number of products, the category with a small number of products is integrated to generate a new category. The relationship between the categories before and after integration is set, for example, using a list that associates the categories before integration with the categories after integration. The category generation unit 12 generates a new category by referring to the list, for example, based on the product categories acquired by the acquisition unit 11. The category generation unit 12 may also generate a new category by having some or all of the categories to be integrated into the new category set by the person in charge of creating the shelf planogram.

[0027] When many identical products are displayed on one shelf, the category generation unit 12 may generate more detailed categories. That is, the category generation unit 12 may, for example, divide one category into two or more categories to generate new categories. Furthermore, the same product may include similar products. For example, when many products that are the subject of a campaign are displayed on one shelf, the category generation unit 12 divides the category to which the products that are the subject of the campaign belong and generates a lower-level category to which the products that are the subject of the campaign belong.

[0028] FIG. 4 shows an example of a classification generated by merging two or more classifications. In the example of FIG. 4, a "classification name before aggregation" and a "classification name after aggregation" are associated with each other. The "classification name before aggregation" is, for example, the classification name before the integration. The "classification name after aggregation" is, for example, the classification name after the integration. In the example of FIG. 4, for example, "Ice Cream / Ice Cream_Ice Cream" in the "classification name before aggregation" indicates that the classifications before the integration are "Ice Cream / Ice Cream" and "Ice Cream." Furthermore, "Ice Cream" associated with "Ice Cream / Ice Cream_Ice Cream" indicates that the classification name after the integration of "Ice Cream / Ice Cream_Ice Cream" and "Ice Cream" is "Ice Cream." In this case, "Ice Cream" is a concept common to "Ice Cream / Ice Cream" and "Ice Cream." Furthermore, in the example of FIG. 4, for example, "Soft Drinks_Tea" in the "classification name before aggregation" indicates that the classifications before the integration are "Soft Drinks" and "Tea." Furthermore, "Beverages" associated with "Soft Drinks_Tea" indicates that the classification name after the integration of "Soft Drinks" and "Tea" is "Beverages." In this case, "drink" is a generic concept common to "soft drink" and "tea."

[0029] The estimation unit 13 estimates a shelf allocation that maximizes the evaluation value related to product sales when products are displayed on shelves by category. In the process of estimating the shelf allocation, the estimation unit 13 estimates a shelf allocation that maximizes the evaluation value related to product sales based on an index related to the shelf position according to the product category, an index related to products sold in parallel across different categories, and an index related to the shelf position where products of the same category are displayed. The evaluation value is calculated, for example, using the following formula: Evaluation value = (index related to shelf position) + (index related to products sold in parallel across different categories) + (index related to shelf position where products of the same category are displayed). In the above formula for calculating the evaluation value, the term related to the index related to shelf position indicates the sales potential of a single product when the product is located in an arbitrary position, for example, taking into account sales priority. In other words, the term related to the index related to shelf position indicates, for example, the sales effect due to the position where the product is displayed.

[0030] The index relating to shelf position is calculated, for example, using the formula (reciprocal of sales priority) x (position weight) x (single-item sales). "Sales priority" is an index whose numerical value becomes smaller, for example, for a product that is desired to be sold with higher priority. Products that are desired to be sold with higher priority are, for example, products in a category where increased sales volume would be beneficial to the store operator. "Sales priority" may be an index whose numerical value becomes higher, for example, for a product that is desired to be sold with higher priority. In this case, the index relating to shelf position is calculated, for example, using the formula (sales priority) x (position weight) x (single-item sales).

[0031] FIG. 5 shows an example of sales priority. In the example of FIG. 5, for example, "products" and "sales priority" are associated with each other. In the example of FIG. 5, "products" are, for example, product categories. Also, in the example of FIG. 5, "sales priority" is, for example, an index indicating the sales priority. That is, in the example of FIG. 5, the smaller the "sales priority" value, the higher the sales priority. For example, in the example of FIG. 5, products categorized as "lunch boxes" and "bread" have higher sales priority than products categorized as "confectionery."

[0032] Furthermore, "position weight" is an index whose numerical value increases, for example, the more likely a product is purchased from a shelf. A product is likely to be purchased from a shelf that is easily visible to customers. For example, areas near the entrance and exit of a store and near the cash register are on the path of customer traffic, so customers are likely to see the products displayed there. For this reason, the "position weight" of shelves on the path of customer traffic is set to a higher value than other shelves, for example.

[0033] FIG. 6 shows an example of position weighting. In the example of FIG. 6, for example, a "shelf ID," a "weight," and "information" are associated with each other. In the example of FIG. 6, the "shelf ID" is, for example, a shelf identifier. In the example of FIG. 6, the "weight" is, for example, a weight according to the shelf position. In the example of FIG. 6, for example, the "information" is information indicating the shelf type used in the shelving planogram estimation. The type is information indicating whether the shelf is treated as a variable shelf or a fixed shelf in the shelving planogram estimation. A variable shelf is, for example, a shelf whose classification is changed when the shelving planogram is estimated. A fixed shelf is, for example, a shelf whose classification is fixed when the shelving planogram is estimated. For example, a classification that needs to be displayed on a shelf with refrigeration is fixed to the shelf with refrigeration. In this case, a shelf with a fixed classification is set to, for example, a fixed shelf. In addition, for example, if there are multiple shelves with refrigeration, a classification that needs to be displayed on a shelf with refrigeration may be assigned to one of the fixed shelves.

[0034] Furthermore, "single-item sales" is, for example, an index that indicates the sales performance of products included in each category. For "single-item sales", an index calculated from sales data for each product is used, for example. The index for "single-item sales" is calculated, for example, using a function that uses sales data for each product as an explanatory variable. "Single-item sales" may be, for example, the sales of each product included in a category. Furthermore, the parameters included in the index term related to shelf position are not limited to the inverse of sales priority, position weight, and single-item sales. The types of parameters included in the index term related to shelf position can be set as appropriate.

[0035] In the above formula for calculating the evaluation value, the index term for products sold in different categories indicates, for example, the sales potential for cross-selling due to targeted purchases and the sales potential for impulse buying that occurs when the products are located in an arbitrary position. The index term for products sold in different categories is calculated, for example, using the formula (cross-selling sales) + (weighting between shelf IDs) × (cross-selling effect value). "Cross-selling sales" is an index indicating, for example, the actual number of products purchased simultaneously from two categories. "Weighting between shelf IDs" is an index indicating the contribution of shelf location to the cross-selling effect. For example, the "weighting between shelf IDs" is set so that the closer the shelf locations are, the larger the value. "Cross-selling effect value" is an index indicating, for example, the likelihood of products being purchased simultaneously across product categories. In other words, the "cross-selling effect value" is an index indicating the effect of cross-selling due to the nature of products that are likely to be purchased simultaneously. For example, the "cross-selling effect value" is set so that the more likely products are purchased simultaneously, the larger the value. The "co-sales effect value" is calculated, for example, based on information about products purchased simultaneously in a single transaction. Furthermore, the parameters included in the index term for products sold in different categories are not limited to the co-sale sales, the weight between shelf IDs, and the co-sales effect value. The types of parameters included in the index term for products sold in different categories can be set as appropriate.

[0036] In the above formula for calculating the evaluation value, the index related to the shelf positions where products of the same category are displayed is, for example, an index indicating the negative effect on sales that occurs when products of the same category are not adjacent to each other. The index related to the shelf positions where products of the same category are displayed is calculated, for example, using the formula (-1) × (shelf ID adjacency coefficient) × (single-item sales). The "shelf ID adjacency coefficient" is, for example, an index indicating the distance between shelves. The "shelf ID adjacency coefficient" is set, for example, so that the greater the distance between shelves, the larger the value. In this case, the index related to the shelf positions where products of the same category are displayed is calculated to be a negative value, so the greater the distance between shelves, the smaller the evaluation value. Furthermore, "single-item sales" is, for example, an index indicating the sales performance of products included in each category. "Single-item sales" is, for example, the same index as the term related to the shelf position index. Furthermore, the parameters included in the term related to the shelf position index where products of the same category are displayed are not limited to the shelf ID adjacency coefficient and single-item sales. The types of parameters included in the index terms relating to the shelf positions where products of the same category are displayed can be set as appropriate. The formula for calculating the evaluation value may also include terms other than the above-mentioned indexes. The formula for calculating the evaluation value may also include a constant term.

[0037] The estimation unit 13 estimates the shelf planogram that maximizes the evaluation value related to product sales, for example, by using a well-known mathematical optimization algorithm that optimizes combinations. The estimation unit 13 estimates the shelf planogram that maximizes the evaluation value related to product sales, for example, by using designation information that specifies part of the shelf classification as a constraint condition. For example, the estimation unit 13 estimates the shelf planogram that maximizes the evaluation value related to product sales by solving an optimization problem using an objective function that uses the evaluation value related to product sales as an objective variable. In this case, the objective function includes, for example, the shelf planogram as an explanatory variable.

[0038] The variable indicating the shelf allocation is expressed, for example, using a matrix with shelf IDs as rows and classifications as columns. Constraints are set so that the sum of each row and each column is 1. Of these, the condition for the sum of each row to be 1 is the condition that multiple products are not assigned to any one shelf ID. In this case, if the variable indicating the shelf allocation is represented as x(i,j), where i = (shelf ID) and j = (product classification), the condition for the sum of each row to be 1 is set to satisfy x(1,1) + x(1,2) + x(1,3) + ... = 1. Furthermore, the condition for any one product classification not to be located on multiple shelf IDs is set to satisfy x(1,1) + x(2,1) + x(3,1) + ... = 1. For example, if "ice cream" is assigned to the shelf with shelf ID "5," the variable indicating the shelf allocation is x(5, ice cream) = 1. In addition, all variables other than "Ice Cream" on the shelf with shelf ID "5" will always be 0, and all columns and variables in the "Ice Cream" column with shelf IDs other than "5" will always be 0. In other words, the equation x(5, other than ice cream) = x(other than 5, ice cream) = 0 is set.

[0039] The constraints may also be set so that shelves for products of the same category are arranged side by side. For example, if it is desired to arrange shelves for "bread" next to each other, the constraints are set so that x(i,bread) = x(i+1,bread) = 1 is satisfied. By setting the constraints in this way, a shelf planogram is generated in which, for example, multiple shelves to which "bread" is assigned are arranged side by side.

[0040] The specification information that specifies part of the shelf classification is a condition that must be met. For this reason, it is also called a hard constraint specified by the specification information. Furthermore, parameters other than the specification information are also called soft constraints because they do not necessarily need to be met. Soft constraints are, for example, the setting values ​​of each item in the formula that calculates the evaluation value.

[0041] FIG. 7 shows an example of data used to set variables indicating classifications assigned to each shelf. In the example data of FIG. 7 , for example, for one shelf ID, only one classification is set to "1," and the other classifications are set to "0." Also, in the example data of FIG. 7 , when products in the same classification are assigned to multiple shelves, they are considered to be independent classifications, even if they belong to the same classification. For example, in the example data of FIG. 7 , "bread" is assigned to two shelves, but the two "breads" are considered to be separate classifications in terms of data handling. In the example data of FIG. 7 , the classification assigned to each shelf ID is set to "1." Therefore, the sum of all the data for each shelf ID is always 1. Also, in the example data of FIG. 7 , for example, for each classification, only one shelf ID is set to "1," and the other shelf IDs are set to "0." In the example data of FIG. 7 , the shelf ID assigned to each classification is set to "1." This indicates the shelf to which the classification is assigned. Therefore, the sum of all the data for each classification is always "1."

[0042] The estimation unit 13 estimates the shelf planogram that maximizes the evaluation value, for example, using the above-mentioned formula for calculating the evaluation value, in which a matrix based on the data in FIG. 7 is used as a variable indicating the classification assigned to each shelf. In the formula for calculating the evaluation value, the "index related to shelf position" is the product of a matrix indicating "sales priority," "position weight," and "single-item sales" and a matrix based on the data in FIG. 7. In this way, by estimating the shelf planogram that maximizes the evaluation value using both fixed shelves and variable shelves, the relationship between products of a classification displayed on fixed shelves and products of a classification displayed on variable shelves can be reflected in the estimated shelf planogram, for example, in the index related to products of different classifications that are sold together and the index related to the position of shelves where products of the same classification are displayed.

[0043] FIG. 8 also shows an example of classifications assigned to shelves in each cycle of the optimization process. In the example of FIG. 8, the classifications assigned to "shelf ID 1," "shelf ID 2," and "shelf ID 6" change with each cycle. That is, "shelf ID 1," "shelf ID 2," and "shelf ID 6" are each set as variable shelves. On the other hand, in the example of FIG. 8, the classifications assigned to "shelf ID 3," "shelf ID 4," and "shelf ID 5" are fixed. That is, "shelf ID 3," "shelf ID 4," and "shelf ID 5" are each fixed shelves. The estimation unit 13 estimates the shelf allocation that maximizes the evaluation value related to sales by solving the optimization problem using the constraints that, for example, "shelf ID 3" is "beverages," "shelf ID 4" is "beverages," and "shelf ID 5" is "ice cream."

[0044] The estimation unit 13 may estimate the shelf planogram using multiple optimization algorithms. For example, the estimation unit 13 estimates the shelf planogram with the highest evaluation value among the estimation results of the multiple optimization algorithms as the optimal shelf planogram. Furthermore, the estimation unit 13 may select the shelf planogram that has been estimated by the largest number of optimization algorithms among the estimation results of the multiple optimization algorithms as the estimated result of the shelf planogram that will increase sales. Furthermore, the estimation unit 13 may select the shelf planogram that is closest to a set selection condition among the estimation results of the multiple optimization algorithms as the estimated result of the shelf planogram that will increase sales. For example, if the selection condition is set to be that the shelf positions between specified categories are closest to the specified categories, the estimation unit 13 selects the shelf planogram that is closest to the shelf positions between the specified categories among the estimation results of the multiple optimization algorithms as the estimated result of the shelf planogram that will increase sales. Furthermore, the estimation unit 13 may score each condition included in the selection condition and select the shelf planogram with the highest score as the estimated result of the shelf planogram that will increase sales. How a shelf planogram that will increase sales is selected from the estimation results of the multiple optimization algorithms is not limited to the above. Furthermore, the estimation unit 13 may use each of the estimation results of the multiple optimization algorithms as an estimation result of a shelf layout that will increase sales.

[0045] The sales forecasting unit 14, for example, predicts sales when the classification of at least one shelf is changed from the shelving planogram estimated by the estimation unit 13. That is, the sales forecasting unit 14 predicts sales when the classification of some shelves is changed from the shelving planogram estimated by the estimation unit 13. The sales forecasting unit 14 may, for example, predict changes in sales when the classification of at least one shelf is changed from the shelving planogram estimated by the estimation unit 13. The sales forecasting unit 14 may, for example, predict sales for the changed shelving planogram based on change information acquired by the acquisition unit 11. The sales forecasting unit 14 may, for example, predict sales when the shelving planogram is randomly changed from the estimated shelving planogram. When the shelving planogram is changed randomly, the number of shelves to be changed may be set by the person in charge of creating the shelving planogram.

[0046] The sales forecasting unit 14 predicts sales when some shelves are reclassified from the estimated shelf layout, for example, using the same function as the function used by the estimation unit 13 to calculate the evaluation value. In this case, the sales forecasting unit 14 may convert the evaluation value into sales. The sales is, for example, a predicted value of sales amount or sales volume when products are displayed in the shelf layout that is the subject of the prediction. For example, the sales forecasting unit 14 converts the evaluation value into sales using a function that defines the relationship between the evaluation value and sales.

[0047] The sales forecasting unit 14 may use a sales forecasting model to predict sales when the classification of some shelves is changed from the shelf planogram estimated by the estimation unit 13. The sales forecasting model is, for example, a machine learning model that predicts sales from data related to shelf planograms. The data related to shelf planograms is, for example, the classification of products assigned to each shelf. The sales model is generated by learning the relationship between shelf planograms and sales. The sales forecasting unit 14 may also predict sales based on the shelf planogram estimated by the estimation unit 13. The sales forecasting unit 14 may also predict sales for the current shelf planogram or an arbitrary shelf planogram. For example, the sales forecasting unit 14 predicts sales for the current shelf planogram using the current shelf planogram as input for the sales forecasting model.

[0048] The output unit 15 outputs, for example, the shelf planogram estimated by the estimation unit 13 as the estimation result. The output unit 15 may output the shelf planogram estimated by the estimation unit 13 and an evaluation value of the shelf planogram. The output unit 15 may also output multiple shelf planograms in descending order of evaluation value. Furthermore, when a part of the shelf planogram is changed from the shelf planogram estimated by the estimation unit 13, the output unit 15 outputs, for example, sales after the change. When a part of the shelf planogram is changed from the shelf planogram estimated by the estimation unit 13, the output unit 15 may output the amount of change in the evaluation value due to the change in the shelf planogram. The output unit 15 outputs the shelf planogram estimation result to, for example, the terminal device 30.

[0049] When the sales forecasting unit 14 forecasts sales, the output unit 15 outputs, for example, the forecasted sales results. When the shelf plan is partially changed from the shelf plan estimated by the estimation unit 13, the output unit 15 may output, for example, the amount of change in sales due to the change in the shelf plan.

[0050] FIG. 9 shows an example of the estimated results of shelving allocation output by the output unit 15. In the example of the estimated results in FIG. 9, a "shelf ID" and a "post-aggregation category name" are displayed. In the example of the estimated results in FIG. 9, the "shelf ID" indicates the shelf identifier. In addition, in the example of the estimated results in FIG. 9, the "post-aggregation category name" indicates the category assigned to each shelf. For example, if the categories have been integrated in the category generation unit 12, the "post-aggregation category name" is the category name after integration. In addition, if the categories have not been integrated in the category generation unit 12, the "post-aggregation category name" is the category associated with the product. For example, in the example of the estimated results in FIG. 9, "bread," "lunch box," "drinks," and "drinks" are assigned to "shelf ID 1," "shelf ID 2," "shelf ID 3," and "shelf ID 4."

[0051] Fig. 10 shows an example of displaying the estimated results of a shelf plan on a map showing the arrangement of shelves in a store. In the example of the estimated results in Fig. 10, the classification assigned to each shelf is displayed at the shelf's position on the map. In addition, in the example of the estimated results in Fig. 10, the shelf plan plans with the second highest evaluation values ​​in the estimated results are displayed as "shelf plan plan 1" and "shelf plan plan 2."

[0052] FIG. 11 shows an example of a display of sales forecast results when the shelf layout is changed from the estimated results on a map showing the shelf arrangement in the store. In the example of the prediction results in FIG. 11 , the "Before Change" in the upper row indicates the shelf layout estimated by the estimation unit 13. In addition, in the example of the prediction results in FIG. 11 , the "After Change" in the lower row indicates the shelf layout changed from the shelf layout estimated by the estimation unit 13. In addition, in the example of the prediction results in FIG. 11 , the shelves for "sweets" and "beverages" have been swapped from the shelf layout estimated by the estimation unit 13. The example of the prediction results in FIG. 11 shows that the change from the shelf layout estimated by the estimation unit 13 results in a decrease of 40,000 yen, from 350,000 yen to 310,000 yen.

[0053] The memory unit 16 stores information related to the process of estimating shelf layout. The memory unit 16 stores, for example, the products to be displayed on the shelves and the classification of each product acquired by the acquisition unit 11. The memory unit 16 stores, for example, the results of shelf layout estimation. The memory unit 16 also stores, for example, information indicating the classification hierarchy of the products. When a sales forecast is made using a prediction model, the memory unit 16 also stores, for example, the prediction model. The prediction model may be stored in a storage means other than the memory unit 16.

[0054] The management device 20 stores data related to sales at the store, for example, in a data storage unit (not shown). Product sales are, for example, the sales amount of the product or the number of products sold. The sales data is, for example, data for each transaction of purchased products. In other words, the sales data is, for example, data on products purchased by a customer in a single shopping trip. Therefore, the sales data includes information on products purchased by the customer at the same time. The sales data may also be sales data for each product. The sales data may also be sales data for each product category.

[0055] The management device 20 stores data related to product classifications in, for example, a data storage unit. The data related to product classifications is, for example, data associating product names with the classifications of the respective products. The management device 20 may also store store planogram data for the store in the data storage unit. The management device 20 outputs data related to product sales to, for example, the acquisition unit 11 of the planogram estimation device 10.

[0056] The terminal device 30 is, for example, a terminal device used for operations in estimating shelf allocation. For example, the terminal device 30 is a terminal device used by a person in charge of creating shelf allocation. The terminal device 30 acquires the shelf allocation estimation result, for example, from the output unit 15 of the shelf allocation estimation device 10. Then, the terminal device 30 outputs the shelf allocation estimation result, for example, to a display device (not shown). Furthermore, when sales are predicted in the shelf allocation estimation device 10, the terminal device 30 acquires the sales prediction result, for example, from the output unit 15 of the shelf allocation estimation device 10. Then, the terminal device 30 outputs the sales prediction result, for example, to a display device (not shown). Furthermore, when sales are predicted for a changed shelf allocation, the terminal device 30 acquires the sales prediction result for the changed shelf allocation, for example, from the output unit 15 of the shelf allocation estimation device 10. Then, the terminal device 30 outputs the sales prediction result for the changed shelf allocation, for example, to a display device (not shown).

[0057] Furthermore, when the shelf planogram is changed from the estimated result, the terminal device 30 acquires change information input by, for example, an operation by a person in charge of creating the shelf planogram, and then outputs the change information.

[0058] The terminal device 30 acquires, for example, from the output unit 15 of the shelf allocation estimation device 10, data for a display screen that displays the estimated shelf allocation results on a map that shows the arrangement of shelves in the store. The terminal device 30 outputs, for example, to a display device (not shown), a display screen that displays the estimated shelf allocation results on a map that shows the arrangement of shelves in the store. The terminal device 30 also acquires, for example, shelf allocation change information that is input by an operation by a person in charge on the display screen. The shelf allocation change information is input, for example, by changing the classification displayed on the map. The terminal device 30 then outputs the shelf allocation change information to the acquisition unit 11 of the shelf allocation estimation device 10.

[0059] An example of the operation of the shelf allocation estimation device 10 to estimate a shelf allocation will be described. Fig. 12 shows an example of the flow of processing to estimate a shelf allocation in the shelf allocation estimation device 10. In the following example, when the number of categories is greater than the number of shelves, the category generation unit 12 performs processing to integrate the categories and estimate a shelf allocation. Also in the following example, when a part of the shelf allocation indicated by the estimation result is changed, processing is performed to predict sales based on the changed shelf allocation.

[0060] The acquisition unit 11 acquires the products to be displayed on the shelves and the classifications of the products (step S11). The acquisition unit 11 acquires the products to be displayed on the shelves and the classifications of the products from the terminal device 30, for example.

[0061] If the number of acquired product categories is greater than the number of shelves (Yes in step S12), the category generation unit 12, for example, integrates the acquired categories to generate new categories (step S13). The category generation unit 12 integrates the categories to generate new categories, for example, so that the number of categories after integration is equal to or less than the number of shelves (step S13).

[0062] When the integrated classification is generated, the estimation unit 13 estimates the shelf allocation that maximizes the evaluation value related to product sales when the products are displayed on shelves by classification (step S14). The estimation unit 13 estimates the shelf allocation that maximizes the evaluation value related to product sales based on an index related to the shelf position according to the product classification, an index related to products that are sold in different classifications, and an index related to the shelf position where products of the same classification are displayed.

[0063] Also, in step S12, if the acquired product classification is equal to or less than the number of shelves (No in step S12), the estimation unit 13 estimates the shelf allocation that will maximize the evaluation value for product sales when the products are displayed on shelves by classification (step S14).

[0064] When the shelf planogram is estimated in step S14, the output unit 15 outputs the estimated shelf planogram result (step S15). The output unit 15 outputs the estimated shelf planogram result to the terminal device 30, for example.

[0065] When the acquisition unit 11 acquires change information for changing part of the shelf layout (Yes in step S16), the sales forecasting unit 14 predicts sales when the shelf layout is changed (step S17).

[0066] When the sales forecast for the case where the shelf plan is changed is performed, the output unit 15 outputs the sales forecast result for the case where the shelf plan is changed (Step S18). The output unit 15 outputs the sales forecast result for the case where the shelf plan is changed to, for example, the terminal device 30.

[0067] In step S16, if change information that changes part of the shelf planogram has not been acquired (No in step S16), the shelf planogram estimation device 10 ends the processing related to the shelf planogram estimation. For example, if the person in charge of creating the shelf planogram selects to end the processing, the shelf planogram estimation device 10 ends the processing related to the shelf planogram estimation.

[0068] In the above explanation, an example of estimating shelf allocation within a floor of a store has been described, but the shelf allocation estimation device 10 may estimate shelf allocation for some of the shelves installed within the floor. Furthermore, the shelf allocation estimation device 10 may estimate shelf allocation for shelves installed on multiple floors. The shelf allocation estimation device 10 may estimate the display position of products on one shelf. Furthermore, the shelf allocation estimation device 10 may estimate shelf allocation when each product is displayed on a shelf. Furthermore, instead of estimating shelf allocation, the shelf allocation estimation device 10 may be used to estimate the layout of a sales area that will maximize sales.

[0069] The shelf allocation estimation device 10 estimates the shelf allocation that maximizes the evaluation value of product sales based on an index related to shelf positions according to product classification, an index related to products of different classifications that are sold together, and an index related to shelf positions where products of the same classification are displayed. By estimating the shelf allocation based on shelf positions, products that are sold together, and shelf positions where products of the same classification are displayed, the shelf allocation estimation device 10 can improve the accuracy of shelf allocation estimation.

[0070] Furthermore, when there is a category that needs to be displayed on a specific shelf in a store, the shelf allocation estimation device 10 can, for example, estimate a shelf allocation by fixing the category of the specific shelf as a constraint, thereby assigning the category that needs to be assigned to the specific shelf to the specific shelf and estimating a shelf allocation that will increase sales. Therefore, for example, when it is desired to display products of the same category on shelves close to each other, the shelf allocation estimation device 10 can estimate a shelf allocation that will increase sales while assigning products of the same category to shelves close to each other. Furthermore, for example, when it is desired to create a shelf plan by fixing the shelf of a category that is desired to sell, the shelf allocation estimation device 10 can estimate a shelf plan that will increase sales while fixing the shelf of the category of the product that is desired to sell. Furthermore, for example, when it is desired to create a shelf plan by fixing the shelf of a category that needs to be displayed on a specific shelf, the shelf allocation estimation device 10 can estimate a shelf plan that will increase sales while fixing the shelf of the category that needs to be displayed on a specific shelf. In this way, the shelf allocation estimation device 10 can estimate a shelf plan that will increase sales based on, for example, the intention of a person in charge of creating the shelf plan.

[0071] Furthermore, by estimating the shelf planogram that maximizes the evaluation value using both fixed shelves and variable shelves, the relationship between the categories of products displayed on fixed shelves and the categories of products displayed on variable shelves can be reflected in the estimated shelf planogram. As a result, the shelf planogram estimation device 10 can further improve the accuracy of estimating shelf planograms that will increase sales, for example.

[0072] Furthermore, by estimating the shelf allocation using a classification that combines two or more classifications, the shelf allocation estimation device 10 can estimate the shelf allocation even when there are many product classifications, for example.

[0073] The processes in the shelf allocation estimation device 10 may be distributed and executed in a plurality of information processing devices connected via a network. For example, the processes in the classification generation unit 12 and the estimation unit 13 and the sales forecasting unit 14 may be performed in different information processing devices. Furthermore, for example, the processes in the classification generation unit 12 and the estimation unit 13 may be performed in different information processing devices. It can be set as appropriate which information processing device performs each process in the shelf allocation estimation device 10.

[0074] Each process in the shelf allocation estimation device 10 can be realized by executing a computer program on a computer. Fig. 13 shows an example of the configuration of a computer 100 that executes a computer program that performs each process in the shelf allocation estimation device 10. The computer 100 includes a CPU (Central Processing Unit) 101, a memory 102, a storage device 103, an input / output I / F (Interface) 104, and a communication I / F 105.

[0075] The CPU 101 reads and executes computer programs for performing each process from the storage device 103. The CPU 101 may be configured with a combination of multiple CPUs. The CPU 101 may also be configured with a combination of a CPU and another type of processor. For example, the CPU 101 may be configured with a combination of a CPU and a graphics processing unit (GPU). The memory 102 is configured with a dynamic random access memory (DRAM) or the like, and temporarily stores computer programs executed by the CPU 101 and data being processed. The storage device 103 stores the computer programs executed by the CPU 101. The storage device 103 is configured with, for example, a non-volatile semiconductor storage device. Other storage devices such as a hard disk drive may also be used for the storage device 103. The input / output I / F 104 is an interface that accepts input from an operator and outputs display data, etc. The communication I / F 105 is an interface that transmits and receives data to and from other information processing devices. The management device 20 and the terminal device 30 may also have a configuration similar to that of the computer 100.

[0076] The computer program used to execute each process can also be stored and distributed on a computer-readable recording medium that non-temporarily stores data. Examples of recording media that can be used include magnetic tapes for recording data and magnetic disks such as hard disks. Optical disks such as CD-ROMs (Compact Disc Read Only Memory) can also be used as recording media. Non-volatile semiconductor storage devices can also be used as recording media.

[0077] Some or all of the above-described embodiments can be described as, but are not limited to, the following supplementary notes.

[0078] [Supplementary Note 1] A shelf allocation estimation device comprising: an acquisition means for acquiring products to be displayed on shelves and the classification of each of the products; an estimation means for estimating a shelf allocation that maximizes an evaluation value related to sales of the products when the products are displayed on shelves by classification, based on an index related to shelf position according to the classification of the products, an index related to products that are sold in parallel but in different classifications, and an index related to shelf position where products of the same classification are displayed; and an output means for outputting the estimated shelf allocation.

[0079] [Supplementary Note 2] The shelf allocation estimation device described in Supplementary Note 1, wherein the acquisition means acquires, as designation information, information designating a category of products to be displayed on at least one shelf out of a plurality of shelves that are the subject of the shelf allocation, and the estimation means estimates the shelf allocation using the designation information as a constraint condition.

[0080] [Supplementary Note 3] The shelf planogram estimation device according to Supplementary Note 1, further comprising a classification generation means for integrating at least two or more classifications from among the classifications of the products to generate a new classification, wherein the estimation means estimates the shelf planogram using the new classification instead of the classification before integration.

[0081] [Supplementary Note 4] The shelf-planography estimation device according to Supplementary Note 3, wherein the classification generating means generates the new classification based on at least one of a similarity between classifications of the products and sales performance of the products.

[0082] [Supplementary Note 5] The shelf allocation estimation device according to Supplementary Note 1, further comprising a sales forecasting means for forecasting sales when at least one shelf classification is changed from the shelf allocation estimated by the estimation means, and the output means outputs the forecasted sales when the shelf allocation is changed.

[0083] [Supplementary Note 6] The shelf planogram estimation device according to Supplementary Note 5, wherein the sales forecasting means predicts sales in a shelf planogram in which the shelf classification has been changed based on the designation of a shelf to be changed from the shelf planogram estimated by the estimation means.

[0084] [Supplementary Note 7] The shelf planogram estimation device according to Supplementary Note 1, wherein the estimation means estimates the shelf planogram using an index related to the positions of the shelves on which the products of the same category are displayed, such that the farther apart the shelves on which the products of the same category are displayed, the lower the index related to sales of the products.

[0085] [Supplementary Note 8] The shelf planogram estimation device according to any one of Supplementary Notes 1 to 7, wherein the estimation means estimates the shelf planogram using a plurality of optimization algorithms.

[0086] [Supplementary Note 9] The shelf allocation estimation device according to any one of Supplementary Notes 1 to 7, wherein the output means outputs to the terminal device display screen data that displays the shelf allocation estimation result on a map that shows the arrangement of shelves in the store.

[0087] [Supplementary Note 10] The shelf allocation estimation device according to Supplementary Note 4, wherein the category generation means integrates categories so that the sum of the number of integrated categories and the number of non-integrated categories is equal to the number of shelves.

[0088] [Supplementary Note 11] The shelf allocation estimation device according to Supplementary Note 4, wherein the category generation means generates the new category by integrating categories with sales lower than other categories.

[0089] [Supplementary Note 12] The shelf allocation estimation device according to Supplementary Note 4, wherein the category generation means combines categories with sales lower than other categories, and generates, as the new category, a category that includes the category before the combination as a lower hierarchical level.

[0090] [Supplementary Note 13] The shelf allocation estimation device according to any one of Supplementary Notes 1 to 7, wherein the index related to the shelf position is calculated based on sales priority, weight of the shelf position, and sales performance of individual products included in each category.

[0091] [Supplementary Note 14] The shelf planogram estimation device according to Supplementary Note 13, wherein the weight of the shelf position is an index whose numerical value becomes larger as the shelf is located in a position where a product is more likely to be purchased.

[0092] [Supplementary Note 15] The shelf allocation estimation device according to Supplementary Note 14, wherein a weight of the position of the shelf on a shelf along a customer's flow line is set to a higher value than that of other shelves.

[0093] [Supplementary Note 16] The shelf allocation estimation device described in any one of Supplementary Notes 1 to 7, wherein the index related to the products that are sold in different categories is calculated based on an index indicating the record of simultaneous purchase of products in the two categories and an index indicating the contribution of shelf position to the side-sale effect.

[0094] [Supplementary Note 17] The shelf planogram estimation device according to any one of Supplementary Notes 1 to 7, wherein the estimation means fixes a category that needs to be displayed on a shelf having a refrigeration function to the shelf having a refrigeration function, and estimates a shelf planogram that maximizes an evaluation value related to sales of the product.

[0095] [Supplementary Note 18] The shelf allocation estimation device described in Supplementary Note 5, wherein the output means outputs to the terminal device display screen data that displays the estimated shelf allocation results on a map that shows the arrangement of shelves in the store, the acquisition means acquires from the terminal device change information that is information that changes the shelf allocation classification input by an operation of a staff member on the display screen, and the estimation means estimates the shelf allocation that maximizes the evaluation value related to sales of the product based on the change information.

[0096] [Supplementary Note 19] A shelf allocation estimation method comprising: acquiring products to be displayed on shelves and the classification of each of the products; estimating a shelf allocation that maximizes an evaluation value related to sales of the products when the products are displayed on shelves by classification, based on an index related to shelf position according to the product classification, an index related to products that are sold in parallel but have different classifications, and an index related to shelf position where products of the same classification are displayed; and outputting the estimated shelf allocation.

[0097] [Supplementary Note 20] A recording medium that non-temporarily records a shelf allocation estimation program that causes a computer to execute the following processes: a process of acquiring products to be displayed on shelves and the classification of each of the products; a process of estimating a shelf allocation that maximizes the evaluation value of sales of the products when the products are displayed on shelves by classification, based on an index related to shelf positions according to the classification of the products, an index related to products that are sold in different classifications, and an index related to shelf positions where products of the same classification are displayed; and a process of outputting the estimated shelf allocation.

[0098] Furthermore, some or all of the configurations described in Supplementary Notes 2 to 18, which are dependent on Supplementary Note 1, may also be dependent on Supplementary Notes 19 and 20 in the same dependent relationship as Supplementary Notes 2 to 18. Furthermore, not limited to Supplementary Notes 1, 9, and 10, some or all of the configurations described as Supplements may be made dependent on various hardware, software, various recording means for recording software, or systems, within the scope of each of the above-mentioned embodiments.

[0099] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0100] This application claims priority based on Japanese Patent Application No. 2024-44453, filed March 21, 2024, the disclosure of which is incorporated herein in its entirety by reference.

[0101] REFERENCE SIGNS LIST 10 Shelf allocation estimation device 11 Acquisition unit 12 Classification generation unit 13 Estimation unit 14 Sales forecast unit 15 Output unit 16 Storage unit 20 Management device 30 Terminal device 100 Computer 101 CPU 102 Memory 103 Storage device 104 Input / output I / F 105 Communication I / F

Claims

1. A shelf allocation estimation device comprising: an acquisition means for acquiring products to be displayed on shelves and the classification of each of said products; an estimation means for estimating the shelf allocation that will maximize the evaluation value of sales of said products when said products are displayed on shelves by classification, based on an index relating to the shelf position according to the classification of the product, an index relating to products that are sold in parallel but have different classifications, and an index relating to the shelf position where products of the same classification are displayed; and an output means for outputting the estimated shelf allocation.

2. The shelf allocation estimation device according to claim 1, wherein the acquisition means acquires, as specified information, information specifying the classification of products to be displayed on at least one shelf out of the plurality of shelves that are the subject of the shelf allocation, and the estimation means estimates the shelf allocation using the specified information as a constraint condition.

3. The shelf allocation estimation device according to claim 1, further comprising a classification generation means for integrating at least two or more classifications of the product classifications to generate a new classification, and the estimation means estimates the shelf allocation using the new classification in place of the classification before integration.

4. The shelf allocation estimation device according to claim 3, wherein the classification generating means generates the new classification based on at least one of the similarity of the classification of the products and the sales performance of the products.

5. The shelf allocation estimation device according to claim 1, further comprising a sales forecasting means for forecasting sales when at least one shelf classification is changed from the shelf allocation estimated by said estimation means, and said output means outputs the forecast results of sales when the shelf allocation is changed.

6. The shelf allocation estimation device according to claim 5, wherein the sales forecasting means predicts sales in a shelf allocation in which the shelf classification has been changed based on the designation of the shelf to be changed from the shelf allocation estimated by the estimation means.

7. The shelf allocation estimation device according to claim 1, wherein the estimation means estimates the shelf allocation using an index related to the shelf positions on which the products of the same category are displayed, such that the farther apart the products of the same category are displayed on the shelves, the lower the index related to sales of the products.

8. A shelf allocation estimation device according to any one of claims 1 to 7, wherein the estimation means estimates the shelf allocation using a plurality of optimization algorithms.

9. A shelf allocation estimation device according to any one of claims 1 to 7, wherein the output means outputs to a terminal device display screen data that displays the shelf allocation estimation results on a map showing the shelf arrangement within the store.

10. The shelf allocation estimation device according to claim 4, wherein the classification generation means integrates classifications so that the sum of the number of integrated classifications and the number of non-integrated classifications is the same as the number of shelves.

11. The shelf allocation estimation device according to claim 4, wherein the category generation means generates the new category by integrating categories with sales lower than other categories.

12. The shelf allocation estimation device according to claim 4, wherein the classification generation means combines classifications with sales lower than other classifications, and generates as the new classification a classification that includes the classification before the combination as a lower hierarchical level.

13. A shelf allocation estimation device according to any one of claims 1 to 7, wherein the index relating to shelf position is calculated based on sales priority, weight of shelf position, and sales performance of individual products included in each category.

14. The shelf allocation estimation device according to claim 13, wherein the weight of the shelf position is an index whose numerical value increases as the shelf is located in a position where products are more likely to be purchased.

15. The shelf allocation estimation device according to claim 14, wherein the weight of the position of the shelf on the line of customer movement is set to a higher value than that of other shelves.

16. A shelf allocation estimation device as described in any one of claims 1 to 7, wherein the index relating to the products that are sold in different categories is calculated based on an index indicating the record of simultaneous purchase of products in the two categories and an index indicating the contribution of shelf position to the effect of co-selling.

17. A shelf allocation estimation device as described in any one of claims 1 to 7, wherein the estimation means fixes categories that need to be displayed on shelves with refrigeration functions to shelves with refrigeration functions, and estimates shelf allocations that maximize the evaluation value related to sales of the products.

18. The shelf allocation estimation device according to claim 5, wherein the output means outputs to the terminal device display screen data that displays the estimated shelf allocation results on a map showing the shelf arrangement within the store, the acquisition means acquires from the terminal device change information that is information that changes the shelf allocation classification entered by an operation of a staff member on the display screen, and the estimation means estimates the shelf allocation that maximizes the evaluation value related to sales of the product based on the change information.

19. A shelf allocation estimation method which obtains products to be displayed on shelves and the classification of each of said products, estimates the shelf allocation which will maximize the evaluation value of sales for said products when said products are displayed on shelves by classification, based on an index relating to the shelf position according to the product classification, an index relating to products that are sold in parallel but have different classifications, and an index relating to the shelf position where products of the same classification are displayed, and outputs said estimated shelf allocation.

20. A recording medium that non-temporarily records a shelf allocation estimation program that causes a computer to execute the following processes: a process of acquiring products to be displayed on shelves and the classification of each of the products; a process of estimating the shelf allocation that will maximize the evaluation value of sales for the products when the products are displayed on shelves by classification, based on an index related to the shelf position according to the product classification, an index related to products that are sold in parallel but have different classifications, and an index related to the shelf position where products of the same classification are displayed; and a process of outputting the estimated shelf allocation.

Images