Learning estimation system, learning estimation method, and learning estimation program

Abstract

To provide a learning estimation system that estimates future sales even for a store with little information. [Solution] In a communication system including a learning estimation system 100 and store terminals 200a, 200b, the learning estimation system includes an acquisition unit that acquires store location data indicating the location of the store, store business type data indicating the business type of the store, and sales-related data related to the store's sales for each day, a conversion unit that converts the data acquired by the acquisition unit into teacher data, and a learning unit that trains the teacher data into a machine learning model that estimates the store's sales-related data based on weather or people flow data.

Description

[Technical field]

[0001] The present invention relates to a learning estimation system, a learning estimation method, and a learning estimation program capable of estimating sales-related data of a store. [Background technology]

[0002] It is known that sales at a store are affected by the weather and the flow of people on that day. Therefore, Patent Document 1 discloses a technology for predicting demand at each location based on the weather at that location. In addition, Patent Document 2 discloses a technology for predicting the number of customers at a store such as a retail store or a restaurant using a Bayesian network. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-087027 A [Patent Document 2] JP 2005-228014 A Summary of the Invention [Means for solving the problem]

[0004] A learning estimation system according to one embodiment of the present invention includes an acquisition unit that acquires store location data indicating the location of a store, store business type data indicating the business type of the store, and sales-related data related to the store's daily sales, a conversion unit that converts the data acquired by the acquisition unit into teacher data, and a learning unit that trains a machine learning model that estimates the store's sales-related data based on at least the store location and the store business type using the teacher data.

[0005] In addition, in the above learning and estimation system, the acquisition unit may further acquire weather data for a date, and the machine learning model may further be a model that estimates sales-related data of a store based on the weather data.

[0006] In addition, in the above learning estimation system, the acquisition unit may further acquire people flow data at the store location on a day, and the machine learning model may further be a model that estimates sales-related data for the store based on the people flow data.

[0007] In addition, in the above learning and estimation system, the acquisition unit may further acquire calendar information indicating whether the date is a holiday or a weekday on the calendar and whether a specific event occurred on the date, and the machine learning model may further be a model that estimates store sales-related data based on the calendar information.

[0008] In addition, in the above learning and estimation system, the acquisition unit may further acquire promotional data regarding sales promotions of the store on a date, and the machine learning model may further be a model that estimates sales-related data of the store based on the promotional data.

[0009] In the learning and estimation system, the conversion unit may convert the sales-related data acquired by the acquisition unit into the training data by performing a logarithmic transformation.

[0010] In addition, in the above learning and estimation system, the conversion unit may logarithmically transform the sales-related data acquired by the acquisition unit, logarithmically transform the sales-related data for different days, and convert the difference between the two sales-related data after logarithmic transformation into teacher data.

[0011] In addition, in the above learning estimation system, when the sales-related data estimated using the machine learning model is sales-related data after a specified number of days, the conversion unit may perform a logarithmic transformation on the sales-related data acquired by the acquisition unit, and also perform a logarithmic transformation on the sales-related data acquired by the acquisition unit before the specified number of days ago, and convert the difference between the two sales-related data after logarithmic transformation into teacher data.

[0012] In the learning and estimation system, the conversion unit may convert the difference between both sets of sales-related data into the training data by performing an inverse logarithmic transformation.

[0013] In addition, in the above-mentioned learning estimation system, when the sales-related data estimated using the machine learning model is sales-related data after a specified day, the conversion unit may convert information based on the difference between the sales-related data acquired by the acquisition unit and sales-related data acquired a specified day before the sales-related data into teacher data.

[0014] The above learning estimation system may also include a reception unit that receives input of date data relating to the location of the store for which sales-related data is to be estimated, the type of store, and the date, an estimation unit that inputs the date data received by the reception unit into a machine learning model to estimate sales-related data for the date indicated by the date data at the store, and an output unit that outputs the sales-related data estimated by the estimation unit.

[0015] In addition, the above learning estimation system may include a data acquisition unit that acquires weather data for the store location based on the store location, and the machine learning model is a model that estimates sales-related data for the store based on the weather data, and the estimation unit may further input the weather data acquired by the data acquisition unit into the machine learning model to estimate the sales-related data.

[0016] In addition, the above learning estimation system may include a data acquisition unit that acquires weather data for the store location based on the store location, and the machine learning model is a model that estimates sales-related data for the store based on the weather data, and the estimation unit may further input the weather data acquired by the data acquisition unit into the machine learning model to estimate the sales-related data.

[0017] In addition, the above learning estimation system may include a data acquisition unit that acquires calendar information indicating whether the date indicated by the date data is a holiday or a weekday on a calendar and whether a specific event occurred on the calendar, and the machine learning model is a model that estimates sales-related data of a store based on the calendar information, and the estimation unit may further input the calendar information acquired by the data acquisition unit into the machine learning model to estimate the sales-related data.

[0018] In addition, the above learning estimation system may include a data acquisition unit that acquires promotional data related to the store's promotions on the date indicated by the date data, and the machine learning model may further be a model that estimates sales-related data of the store based on the promotional data, and the estimation unit may further input the promotional data acquired by the data acquisition unit into the machine learning model to estimate the sales-related data.

[0019] In addition, in the above learning estimation system, the acquisition unit may be provided with a compensation unit that compensates for sales-related data for a specific date when the acquisition unit is unable to acquire sales-related data for the specific date, and the estimation unit may estimate sales-related data for the date indicated by the date data based on the sales-related data compensated for by the compensation unit.

[0020] In the above learning and estimation system, the compensation unit may compensate for the sales-related data of the specific date by using sales-related data of the same day of the week as the specific date, which is past sales data of the store.

[0021] In the learning and estimation system, the learning unit may learn the sales-related data compensated for by the compensation unit.

[0022] Furthermore, a learning and estimation method according to one aspect of the present invention executes an acquisition step in which a computer acquires store location data indicating the location of a store, store business type data indicating the business type of the store, and sales-related data related to the store's daily sales, a conversion step in which the computer converts the data acquired in the acquisition step into teacher data, and a learning step in which the teacher data is trained by a machine learning model that estimates the store's sales-related data based on at least the store location and the store business type.

[0023] In addition, a learning estimation program according to one embodiment of the present invention provides a computer with an acquisition function for acquiring store location data indicating the location of a store, store business type data indicating the business type of the store, and sales-related data related to the store's daily sales, a conversion function for converting the data acquired by the acquisition function into teacher data, and a learning function for training the teacher data into a machine learning model that estimates the store's sales-related data based on at least the store location and the store's business type. [Brief description of the drawings]

[0024] [Figure 1] FIG. 1 is a schematic diagram showing an outline of the invention. [Diagram 2] 1 is a block diagram showing an example of the configuration of an information processing device; [Diagram 3] FIG. 2 is a block diagram showing a configuration example of a store terminal. [Figure 4] 11 is a flowchart showing an example of the operation of the information processing device during learning. [Diagram 5] 13 is a flowchart showing an example of the operation of the store terminal during learning. [Figure 6] 13 is a flowchart illustrating an example of an operation of the information processing device during estimation. [Figure 7] 13 is a flowchart illustrating an example of the operation of a store terminal during estimation. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] Hereinafter, a learning estimation system according to the present invention will be described with reference to the drawings.

[0026] <Summary> It is important to predict sales-related data for a store. The sales-related data may be, but is not limited to, the number of customers who visit the store in a day or the total sales for a day. Any other information may be used as long as it is related to the sales of the store, and the period is not limited to one day. If the number of customers can be predicted as sales-related data, in the case of a store such as a restaurant, it is possible to estimate the necessary purchases and reduce unnecessary purchases. Alternatively, if the total sales can be predicted as sales-related data, it becomes easier to plan renovations and the installation of new equipment in the store. Such predictions can be made using a learning model.

[0027] However, such a learning model requires a large amount of training data for accurate prediction. By using actual past sales-related data of a store as the training data, the prediction for that store can be made accurate. However, a small store or the like has a problem in that it is not possible to prepare training data sufficient to make an accurate prediction. This is also true for new stores. In addition, it may be difficult for the store to prepare a learning model in the first place.

[0028] Therefore, in this embodiment, the objective is to provide a learning estimation system, a learning estimation method, and a learning estimation program that can provide a machine learning model that can make predictions with a certain degree of accuracy even if it is difficult for a store to prepare a learning model.

[0029] FIG. 1 is a schematic diagram showing an overview of the processing of the learning estimation system 100 according to the present embodiment. FIG. 1 shows an example of the configuration of a communication system including the learning estimation system 100, and shows the learning estimation system 100 and store terminals 200a and 200b (hereinafter collectively referred to as store terminals 200. Note that in FIG. 1, stores are shown as substitutes for terminals rather than terminals). The learning estimation system 100 stores a machine learning model that estimates sales-related data of a store based on at least the location of the store and the business type of the store, and this machine learning model may estimate sales-related data based on any one or more of the weather at the store location, the flow of people at the store location, calendar information on whether it is a holiday or weekday, or whether some event is scheduled on that day, and sales promotion data at the store, in addition to the location of the store and the business type of the store. In other words, the machine learning model is a model that learns the correspondence between various store-related information such as the store location and the store type and the sales-related data, and may be a model that learns the correspondence between the store location, the store type, the weather near the store, the flow of people near the store, calendar information, and promotional data, and the sales-related data. It can be said that the learning estimation system 100 learns the sales-related data of stores of various types in various locations. For example, in the case of a restaurant, the type of the store may be the genre of services provided, such as Japanese food, soba, ramen, Italian food, French food, etc., or the business form of the store, such as a kitchen car, a fixed store, a food court, a standing restaurant, or a reservation-only restaurant. In addition, the store is not limited to a restaurant, but may be a clothing store or a furniture store, and the type of the store may change depending on the contents of the store.

[0030] Here, by learning the sales-related data of the store terminal 200, it is possible to provide a learning estimation system 100 that can make more accurate estimations of the sales-related data of the store terminal 200 while compensating for the paucity of training data on the store terminal 200. This will be described in detail below.

[0031] <Configuration> <Configuration of learning estimation system 100> FIG. 2 is a block diagram showing a configuration example of the learning estimation system 100. The learning estimation system 100 is a computer system (information processing device) that converts sales-related data of a store terminal 200 into teacher data for learning, and estimates the sales-related data of the store terminal 200. The learning estimation system 100 may be realized by a PC, a server device, etc., but is not limited to these. In addition, the functions of the learning estimation system 100 may be realized by parallel processing by multiple computers, or may be realized by a cloud server. The learning of sales-related data and its estimation by the learning estimation system 100 may be provided as a web service.

[0032] 2, the learning estimation system 100 includes a communication unit 110, a control unit 130, and a storage unit 140. The learning estimation system 100 may also include an input unit 120 and an output unit 150.

[0033] The communication unit 110 is a communication interface having a function of communicating with an external device of the learning estimation system 100. The communication unit 110 has a function of communicating with the store terminal 200 as an external device. The communication unit 110 receives sales-related data from the store terminal 200 and transmits it to the control unit 130. The communication unit 110 also receives date data indicating the date on which the sales-related data is to be estimated from the store terminal 200 and transmits it to the control unit 130. Here, the date indicated by the date data may be a specific day, or a predetermined period including multiple days (for example, the weekdays of the next week, the next month, or the three months one year from now, but is not limited to these, and may be arbitrary by the store clerk). The communication unit 110 also transmits the estimated sales-related data to the store terminal 200 that requested the estimation according to an instruction from the control unit 130.

[0034] The input unit 120 has a function of accepting input from an operator of the learning estimation system 100, etc., and transmitting the contents of the input to the control unit 130. The input unit 120 may be realized by an input device such as a mouse, a keyboard, or a touch panel, and in the case of voice input, may be realized by a microphone. As an example, the input unit 120 accepts input of sales-related data information from a clerk of the store terminal 200, etc., of the operator of the learning estimation system 100, and transmits the information to the control unit 130.

[0035] The control unit 130 is a processor having a function of controlling each unit of the learning estimation system 100. The control unit 130 may be realized by a single core or a multi-core. The control unit 130 executes various programs stored in the storage unit 140 and uses various data to realize the functions of the learning estimation system 100.

[0036] The control unit 130 converts the sales-related data received from the store terminal 200 into teacher data and causes the machine learning model 141 in the storage unit 140 to learn the data. The control unit 130 then estimates the sales-related data for the date indicated by the date data received from the store terminal 200 and notifies the store terminal 200 of the estimated sales-related data.

[0037] The control unit 130 includes, as functions realized by the control unit 130, an acquisition unit 131, a conversion unit 132, a learning unit 133, a reception unit 134, a data acquisition unit 135, and an estimation unit 136.

[0038] The acquisition unit 131 acquires sales-related data of the store terminal 200. The acquisition unit 131 acquires sales-related data transmitted from the store terminal 200 and received by the communication unit 110. The acquisition unit 131 transmits the acquired sales-related data to the conversion unit 132. The sales-related data acquired by the acquisition unit 131 includes information on the number of customers or sales (which may be total sales or sales per SKU (Stock Keeping Unit)) at the store of the store terminal 200 for one day or each day during a predetermined period, and may also include information on the weather at the location of the store on that day, the flow of people around the store, and whether or not sales promotion was conducted on that day. In addition, when the sales-related data does not include information on the weather or the flow of people, the acquisition unit 131 may acquire the sales-related data from a weather information service that provides weather information at that time or a people flow service that provides people flow information based on the location of the store of the store terminal 200.

[0039] The conversion unit 132 converts the sales-related data transmitted from the acquisition unit 131 into teacher data. Here, converting into teacher data may mean extracting information necessary as teacher data from the sales-related data transmitted from the acquisition unit 131 and the weather or people flow data transmitted from the data acquisition unit 135, and converting the information into data in a format that allows the machine learning model 141 in the storage unit 140 to learn. The conversion unit 132 may convert the sales-related data into teacher data by logarithmic transformation. Since the absolute value of the sales-related data can vary greatly depending on the location and size of the store, the learning estimation system 100 can normalize the numerical value by logarithmic transformation of the sales-related data and perform learning corresponding to various stores, and as a result, can provide more accurate estimated sales-related data for various stores. The conversion unit 132 transmits the generated teacher data to the learning unit 133.

[0040] The learning unit 133 trains the machine learning model 141 on the teacher data transmitted from the conversion unit 132. This allows the machine learning model 141 to perform estimation based on actual sales at the store of the store terminal 200, and therefore allows more accurate estimation at the store than when learning is not performed. The learning unit 133 may perform learning using, for example, a gradient boosting decision tree, a deep neural network, or the like, but is not limited to these.

[0041] The reception unit 134 receives date data transmitted from the store terminal 200 and received by the communication unit 110. The date data received by the reception unit 134 is information indicating the date on which sales-related data in the store terminal 200 is estimated (predicted). The reception unit 134 transmits the received date data to the estimation unit 136. The reception unit 134 also transmits information related to the store of the store terminal 200 that transmitted the date data to the estimation unit 136. The information related to the store may be read out from the storage unit 140, or the information related to the store may be received from the store terminal 200 when the date data is received. The reception unit 134 also transmits information indicating the location of the store, out of the information related to the store, to the data acquisition unit 135.

[0042] The data acquisition unit 135 acquires weather or people flow data for the date indicated by the date data of the store terminal 200 based on the location of the store terminal 200. The data acquisition unit 135 may acquire one or both of the weather and people flow data. For example, the data acquisition unit 135 may access a server that stores weather data for each day of the year and provides the information to acquire the data. Also, for example, the data acquisition unit 135 may access a server that stores people flow data for each location of the year and provides the information to acquire the data. Here, when estimating the sales-related data, the data acquisition unit 135 may acquire the average value of weather or people flow data for the location for the past several years on the date to be estimated. The average value of the weather data may be, for example, the rainfall rate based on information on whether it is rainy or sunny on that day. Also, the average value of the people flow data may be the average value of the number of people present on that day in the past in the regional mesh at the location of the store. In addition, the weather or people flow data on the date indicated by the date data may be data estimated by a trained model that has learned the weather or people flow on each day in the past. The data acquisition unit 135 transmits the acquired weather or people flow data to the conversion unit 132 and the estimation unit 136.

[0043] The estimation unit 136 estimates the sales-related data of the store terminal 200 based on the information transmitted from the reception unit 134 and the data acquisition unit 135. The estimation unit 136 estimates the sales-related data of the store terminal 200 on the (future) date transmitted from the reception unit 134, and estimates the sales-related data of the store of the store terminal 200 on the transmitted date by inputting the date data received by the reception unit 134, information indicating the location of the store terminal 200, information indicating the business type of the store terminal 200, and predicted weather or people flow data on the date indicated by the date data in the store terminal 200, into the machine learning model 141. The date on which this estimation is performed is specified by the date data, and may be one day or multiple days. At this time, the machine learning model 141 may predict the face value of the sales itself, the number of customers, or both, as the sales-related data. The estimation unit 136 transmits the estimated sales-related data to the store terminal 200 via the communication unit 110.

[0044] The acquisition unit 131, the conversion unit 132, the learning unit 133, and the data acquisition unit 135 are functional units that learn sales-related data for the store of the store terminal 200. In addition, the reception unit 134, the data acquisition unit 135, and the estimation unit 136 are functional units that execute estimation processing using the machine learning model 141.

[0045] The storage unit 140 has a function of storing various programs and data required for the operation of the learning estimation system 100. The storage unit 140 can be realized by, for example, a hard disk drive (HDD), a solid state drive (SSD), a flash memory, etc., but is not limited to these. The storage unit 140 may be a cloud storage accessible by the learning estimation system 100. The storage unit 140 may store various programs and various data for realizing the functions to be performed by the learning estimation system 100. The storage unit 140 may store a program that converts the sales-related data of the store terminal 200 into teacher data and makes the machine learning model 141 learn the data, or a program that estimates and transmits sales-related data of the store of the store terminal 200 at the date and time transmitted from the store terminal 200. The storage unit 140 also stores a machine learning model 141 that estimates sales-related data of a store based on weather or human flow. The machine learning model 141 is a learning model that is trained using the date and time when sales (total sales amount, number of customers, or number of items sold per SKU, or a combination of these) occurred, the store location, and the store type as explanatory variables, and sales-related data (total sales amount, number of customers, or both) as objective variables. The machine learning model 141 may further train weather information, people flow information, calendar information, and store promotion information for the store location as explanatory variables. The store location may be managed in units of a predetermined range when a map is divided into units of a predetermined range (for example, a range of 500 m square, but is not limited to this, and may be 1 km square, 5 km square, etc.). In this case, the weather data may be weather information corresponding to the day of the sales-related data in the predetermined range, and this weather information may be weather information for each time period (for example, a time period of 1 hour, but is not limited to this, and may be 30 minutes, 2 hours, morning, afternoon, night, etc.). Furthermore, the people flow data may be the number of people present at each time within the specified range (which may be, for example, one-hour units, but is not limited to this, and may also be in 10-minute units, 30-minute units, or even one day units).The weather information may be obtained from, for example, a server of the Japan Meteorological Agency, but is not limited thereto. The people flow data may be the number of people estimated from the number of terminals connected to a base station in a communication carrier, and may be obtained from the communication carrier. The calendar information may be information indicating whether the date of the sales-related data to be learned is a weekday, a holiday, or a day on a specific event (for example, Valentine's Day or Christmas, but is not limited thereto, and may be an event set by the organizer, such as a fireworks display, a sports match, or a chorus contest), and the information may be obtained from normal calendar data. The sales promotion information may be information on whether or not a sales promotion is being carried out in the store, and may further include the details of the sales promotion. Here, the sales promotion may be a measure taken by the store with the aim of increasing sales at the store of the store terminal 200, and may be, for example, a sale only during a specific time period, issuing a special offer such as a coupon, or receiving a free gift with the purchase of a certain product, but is not limited thereto. The storage unit 140 may also store information about the store of the store terminal 200 that uses the learning estimation system 100. The information about the store may be information indicating the name, location, and type of business of the store, and may also include information such as the store area and the number of customers that can be attracted.

[0046] The output unit 150 has a function of outputting designated information in accordance with an instruction from the control unit 130. The output unit 150 may output character information or image information, for example, and in that case, the output unit 150 is realized by a monitor provided in or connected to the learning estimation system 100. The output unit 150 may output audio information, for example, and in that case, the output unit 150 is realized by a speaker provided in or connected to the learning estimation system. The output unit 150 may output estimated sales-related data, for example.

[0047] The above is an example of the configuration of the learning estimation system 100.

[0048] <Configuration of the store terminal 200> 3 is a block diagram showing a configuration example of the store terminal 200. The store terminal 200 is a terminal related to a store, and may be a computer system that manages information such as sales at the store. The store terminal 200 may be realized by, but is not limited to, a PC, a tablet terminal, a server device, a smartphone, etc. As long as the store terminal 200 is related to the store, it does not have to be installed in the store, or it may be installed in the store.

[0049] As shown in FIG. 3, the store terminal 200 includes a communication unit 210, an input unit 220, a control unit 230, a storage unit 240, and an output unit 250.

[0050] The communication unit 210 is a communication interface having a function of communicating with an external device of the store terminal 200. The communication unit 210 has a function of communicating with the learning estimation system 100 as an external device. The communication unit 210 transmits sales-related data of the store to the learning estimation system 100 according to an instruction from the control unit 230.

[0051] The input unit 220 has a function of accepting input from an operator of the store terminal 200 and transmitting the input contents to the control unit 230. The input unit 220 may be realized by an input device such as a mouse, a keyboard, or a touch panel, and may be realized by a microphone in the case of voice input. As an example, the input unit 220 accepts input of date data indicating a date for which a store clerk wants to predict sales and the number of customers, and transmits it to the control unit 230. As described above, the date indicated by the date data may be a specific one day, or a predetermined period including multiple days (for example, the weekdays of the next week, the next month, or the three months one year from now, but is not limited to these, and may be arbitrary by the store clerk). In addition, the input unit 220 may accept input of information on the sales amount and the number of customers from the store clerk and transmit it to the control unit 230.

[0052] The control unit 230 is a processor having a function of controlling each unit of the store terminal 200. The control unit 230 may be realized by a single core or a multi-core. The control unit 230 executes various programs stored in the storage unit 240 and realizes the functions of the store terminal 200 by using various data.

[0053] The control unit 230 transmits the sales-related data to the learning estimation system 100 and outputs the sales-related data for a date specified by a store clerk or the like. When the control unit 230 receives date data for which sales-related data is to be predicted from the input unit 220, the control unit 230 may transmit the date data to the learning estimation system 100 via the communication unit 210. At this time, if a sales promotion is scheduled to be carried out on the date indicated by the date data, the sales promotion information may be included in the transmission.

[0054] The control unit 230 includes a sales-related data transmission unit 231 and a display control unit 232 as functions to be performed by the control unit 230.

[0055] The sales related data transmission unit 231 transmits the sales related data stored in the memory unit 240 of the store terminal 200 to the learning estimation system 100 via the communication unit 210. The sales related data transmission unit 231 may transmit the sales related data to the learning estimation system 100 every time new sales related data is stored in the memory unit 240, or may transmit the sales related data collectively every predetermined period (for example, every week, every month, etc., but is not limited to these). The sales related data transmission unit 231 may transmit the sales related data in response to an instruction from a clerk at the store where the store terminal 200 is located. The sales-related data transmission unit 231 may transmit at least one day's sales and the number of customers as sales-related data together with identification information that can identify the store to the learning estimation system 100. In addition, as information related to the store, information indicating the store's location (longitude and latitude information, address, etc.), information indicating the store's business type, store area, number of parking spaces, number of cash registers, number of seats in the store, opening time, closing time, business hours, and promotional data if any, may also be transmitted. When store information is registered in the learning estimation system 100, it is not necessary to transmit such information. The sales-related data transmission unit 231 may transmit sales-related data to the learning estimation system 100 at the instruction of a clerk at the store terminal 200.

[0056] The display control unit 232 causes the output unit 250 to output information based on the sales forecast data transmitted from the learning estimation system 100.

[0057] The storage unit 240 has a function of storing various programs and data required for the operation of the store terminal 200. The storage unit 240 can be realized, for example, by a hard disk drive (HDD), a solid state drive (SSD), a flash memory, etc., but is not limited to these. The storage unit 240 may store various programs and various data for realizing the functions to be performed by the store terminal 200. The storage unit 240 may also store sales-related data 241. The sales-related data 241 is a database including information related to sales at the store corresponding to the store terminal 200. The sales-related information includes information such as daily sales amount and number of visitors. The storage unit 240 may also store information indicating the location of the store (such as longitude and latitude information and address) and information indicating the type of business of the store, as well as information regarding the store area, number of parking spaces, number of cash registers, number of seats in the store, opening hours, closing hours, business hours, and information regarding sales promotion.

[0058] The output unit 250 has a function of outputting designated information according to an instruction from the control unit 230. The output unit 250 may output character information or image information, and in that case, the output unit 250 is realized by a monitor provided in or connected to the store terminal 200. The output unit 250 may output audio information, and in that case, the output unit 250 is realized by a speaker provided in or connected to the learning estimation system. The output unit 250 displays information indicating the estimated sales-related data, that is, information on the sales amount or number of customers of the store on a certain date in the future.

[0059] The above is an example of the configuration of the store terminal 200.

[0060] <Operation> Now, the operations of the learning estimation system 100 and the shop terminal 200 will be described.

[0061] FIG. 4 is a flowchart showing an example of the operation of the learning estimation system 100 when learning sales-related data of the store terminal 200.

[0062] As shown in FIG. 4, the acquisition unit 131 of the learning estimation system 100 acquires sales-related data transmitted from the store terminal 200 via the communication unit 110. The sales-related data includes information on the number of customers and sales at the store, and may include sales promotion information if a sales promotion is being conducted. The acquisition unit 131 also acquires store location data indicating the location of the store and store business type data indicating the business type of the store (step S401). The acquisition unit 131 may acquire the store location data indicating the location of the store and the store business type data indicating the business type of the store from the store terminal 200 via the communication unit 110, or may acquire the store location data indicating the location of the store and the store business type data indicating the business type of the store from the storage unit 140 if store information is stored in advance in the storage unit 140. The acquisition unit 131 transmits the acquired sales-related data to the conversion unit 132.

[0063] The conversion unit 132 converts each acquired data into teacher data (step S402). The conversion unit 132 transmits the generated teacher data to the learning unit 133. The teacher data includes information indicating the location of the store, information indicating the type of business of the store, and sales-related data of the store, and may include weather information at the store location, information on the flow of people in the vicinity of the store, calendar information indicating whether the date of the sales-related data is a weekday, holiday, or day with a specific event, and promotional information indicating whether a promotional event was held on the date of the sales-related data.

[0064] The learning unit 133 causes the machine learning model 141 to learn the teacher data transmitted from the conversion unit 132 (step S403). As a result, the machine learning model 141 also learns the sales related data of the store terminal 200, and when predicting the sales related data of the store terminal 200, it becomes possible to realize a more accurate prediction.

[0065] FIG. 5 is a flowchart showing an example of an operation performed when uploading sales-related data from the store terminal 200 to the learning estimation system 100.

[0066] As shown in Fig. 5, the sales related data transmission unit 231 of the store terminal 200 judges whether the timing for transmitting sales related data has arrived (step S501). The transmission timing may be set in advance by a store clerk at the store terminal 200, and may be, for example, the arrival of a specific time such as midnight on that day, but is not limited to this. The store terminal 200 waits until the timing for transmitting sales related data arrives (NO in step S501).

[0067] When the timing for transmitting sales-related data arrives (YES in step S501), the sales-related data transmitting unit 231 acquires untransmitted sales-related data from the storage unit 240 (step S502). Then, the sales-related data transmitting unit 231 transmits the acquired sales-related data to the learning estimation system 100 via the communication unit 210 (step S503). At this time, the sales-related data transmitting unit 231 may also transmit at least identification information for identifying the store as information related to the store, and may also transmit information indicating the store's location and information indicating the store's business type, but if the learning estimation system 100 has already stored this information, it is not necessary to transmit it. Furthermore, the sales-related data to be transmitted may include sales promotion information if a sales promotion was conducted on that day.

[0068] Since the store terminal 200 transmits sales-related data to the learning estimation system 100 as needed, the learning estimation system 100 learns the information and can realize accurate estimation when estimating the sales-related data.

[0069] FIG. 6 is a flowchart showing an example of the operation of the learning estimation system 100 when estimating sales-related data of a store.

[0070] The reception unit 134 of the learning estimation system 100 receives store information and date data for which sales or the number of customers, or both, are to be predicted (step S601). The reception unit 134 receives the date data from the store terminal 200 via the communication unit 110. The reception unit 134 may receive store information from the store terminal 200 or may acquire the store information from the storage unit 140. The store information is information indicating at least the store's location and the store's business type. The reception unit 134 transmits the acquired store location information and date data to the data acquisition unit 135. The reception unit 134 transmits the acquired store information and date data to the estimation unit 136. If a sales promotion is scheduled to be held at the store on the date indicated by the date data, the reception unit 134 may also receive the sales promotion information from the store terminal 200 and transmit it to the estimation unit 136.

[0071] The data acquisition unit 135 may acquire weather information or people flow data or calendar information predicted for the date indicated by the transmitted date data at the location of the store based on the transmitted store location (step S602). The data acquisition unit 135 transmits the acquired weather information or people flow data to the estimation unit 136. If weather information or people flow data is not used to estimate the sales-related data, step S602 can be omitted. The estimation unit 136 inputs the date and time indicated by the transmitted date and time information, the location of the store and the type of the store as store information, and weather information, people flow data, calendar information, and sales promotion information as necessary, into the machine learning model 141 to estimate the sales-related data at the date and time indicated by the date and time information (step S603).

[0072] Then, the estimation unit 136 outputs the estimated sales-related data (step S604). The estimation unit 136 outputs the estimated sales-related data by transmitting it to the store terminal 200 via the communication unit 110, and ends the process.

[0073] FIG. 7 is a flowchart showing an example of the operation of the store terminal 200 when presenting predicted sales-related data to a store clerk.

[0074] 7, the store terminal 200 receives data on a date for which sales are to be predicted from a store clerk via the input unit 220 (step S701). The input unit 220 transmits the received date data to the control unit 230.

[0075] When the control unit 230 receives the date data, it transmits the received date data and information indicating the store to the learning estimation system 100 via the communication unit 210 (step S702). The information indicating the store includes the identification information of the store, and may include information indicating the store's location and the store's business type. If the learning estimation system 100 stores information indicating the store's location and business type, only the identification information of the store is sufficient.

[0076] The communication unit 210 of the store terminal 200 receives the sales-related data estimated by the learning estimation system 100 (step S703). The communication unit 210 transmits the received sales-related data to the control unit 230.

[0077] When the display control unit 232 of the control unit 230 receives the sales-related data, it outputs the information to the output unit 250 (monitor) (step S704), and ends the process.

[0078] This allows the store clerk at the store terminal 200 to recognize the store's sales and the number of customers on a specified date, which can be useful for, for example, purchasing goods.

[0079] The above is an example of the operation of the learning estimation system 100 and the shop terminal 200 according to the embodiment.

[0080] <Summary> As described above, the learning estimation system 100 originally stores the machine learning model 141 that has learned the sales-related data of each of various stores by associating it with the date of the sales, the location of the store, the type of the store, and the weather or the flow of people at that time. By preparing a machine learning model that has learned the sales-related data of stores in advance in correspondence with various types of stores in various locations, it is possible to estimate the sales-related data with a certain degree of accuracy even for small stores or new stores where it is difficult to prepare a parameter of teacher data. The learning estimation system 100 learns the sales-related data sent from the store terminal 200 by associating it with the location, type of store, the weather or the flow of people at that time, calendar information, promotional information, etc., of the store, so that it can estimate the sales-related data of the store. The more the sales-related data of the store is learned, the more the estimation accuracy by the learning estimation system 100 can be improved. In addition, the store terminal 200 simply sends the sales-related data to the learning estimation system 100, which automatically converts the data into training data in a format suitable for the machine learning model 141 and trains the data. This allows the store clerk or other staff member at the store terminal 200 to easily obtain sales-related data for the date and time specified by the store from the learning estimation system 100, which can be used to improve the store's operations.

[0081] <Modification> The learning estimation system 100 and the store terminal 200 according to the above embodiment are not limited to the above embodiment, and may be realized by other methods. Various modified examples will be described below.

[0082] (1) In the above embodiment, the learning and estimation system 100 has been described as a computer system that executes both learning and estimation. However, the learning and estimation may be executed by separate devices.

[0083] (2) In the above embodiment, the learning estimation system 100 has been described as estimating sales-related data for a specified date based on the location of the store, the type of the store, weather, or foot traffic. However, the learning estimation system 100 may estimate sales-related data using other information related to the store of the store terminal 200, i.e., information such as the store area, the number of parking spaces, the number of cash registers, the number of seats in the store, opening hours, closing hours, and business hours. In this case, the machine learning model 141 must further learn in advance information used to estimate the store area, the number of parking spaces, the number of cash registers, the number of seats in the store, opening hours, closing hours, and business hours. By further adding this information, more accurate estimation is expected.

[0084] (3) In the above embodiment, the learning estimation system 100 may be provided with a compensation unit that receives sales-related data for a certain period from the store terminal 200 and, if there is a day within the certain period for which there is no sales-related data even though it is a business day, compensates for the sales-related data for that day.

[0085] For example, the compensation unit may calculate an average value of sales related data for a day for which there is no sales related data for the store for the number of years for which there is data for the same day in the past or for a predetermined number of years, and compensate the calculated average value as the sales related data. The compensation unit may compensate for the sales related data estimated by the estimation unit 136, or may compensate for the sales related data for a day for which there is no sales related data when sales related data for multiple consecutive days is received from the store terminal 200.

[0086] The learning unit 133 may or may not learn the compensated data as training data for the shop terminal 200.

[0087] (4) In the above embodiment, the learning estimation system 100 acquires sales-related data of a store and learns from it to improve the estimation accuracy of the sales-related data of the store. Here, the learning estimation system 100 may estimate the sales-related data without sales-related data of the store, i.e., without learning sales at the store. In this case, only the location and business type of the store are input to the learning estimation system 100 to estimate the sales-related data. As an example of such a usage method, when a new store is to be opened in a specific location, sales-related data for that location can be acquired, and thus information for determining whether the location is suitable for the new store can be obtained.

[0088] (5) In the above embodiment, the learning estimation system 100 has been described as collecting, learning, and estimating sales-related data on a daily basis, but this is not limited thereto. The learning estimation system 100 may learn and estimate the sales-related data in a desired unit. For example, the sales-related data may be collected, learned, and estimated in units of one hour, a predetermined time unit such as morning and afternoon, or monthly, depending on the estimation unit required for the learning estimation system 100. This allows the learning estimation system 100 to perform estimation in various time unit widths.

[0089] (6) In the above embodiment, the learning estimation system 100 has been described as performing logarithmic transformation of the sales-related data acquired by the acquisition unit 131 to convert it into teacher data, and causing the learning unit 133 to learn the data. However, this is not limited to the above. The conversion unit 132 may perform logarithmic transformation of the sales-related data, calculate the difference between the logarithmic transformed value and the logarithmic transformed value of the sales-related data of a different day, and use the calculated difference as teacher data. Here, the sales-related data of a different day may be the sales-related data of the acquired sales-related data a predetermined day before, when the sales-related data estimated by the machine learning model 141 is the sales-related data after a predetermined day. Furthermore, the teacher data may be a value obtained by further performing an inverse logarithmic transformation of the calculated difference. In this case, the output of the machine learning model 141 may be a value obtained by performing an inverse logarithmic transformation of the difference between the value obtained by logarithmic transformation of the acquired sales-related data and the value obtained by logarithmic transformation of the estimated sales-related data after a predetermined day.

[0090] The learning unit 133 may also learn the difference between sales-related data that has not been logarithmically transformed as teacher data. That is, when the machine learning model 141 estimates sales-related data after a predetermined number of days, the learning unit 133 may learn the difference between the sales-related data acquired by the acquisition unit 131 and the sales-related data acquired a predetermined number of days before the sales-related data was acquired as teacher data. This difference may also be a second-order difference. In the case of a first-order difference, the amount of change in the sales-related data is learned, and in the case of a second-order difference, the change in the amount of change in the sales-related data is learned. The learning unit 133 may also learn both differences.

[0091] The learning unit 133 may learn teacher data indicating this difference. By learning the difference, the learning unit 133 can make the time series data (sales-related data along a time series) stationary. The time series data having stationarity means that the statistical characteristics of the time series data are less likely to change. Therefore, the machine learning model 141 generated by the learning unit 133 through learning can make estimations that correspond to various conditions (follow trends) on the day on which the sales-related data is estimated.

[0092] (7) In the above embodiment, the estimation unit 136 of the learning estimation system 100 estimates sales-related data for a date indicated by the date data received by the reception unit 134, but this is not limited to the above example. The estimation unit 136 may estimate sales-related data for a predetermined date and transmit it to the store terminal 200. For example, when the store terminal 200 transmits sales-related data for a certain day, the estimation unit 136 may estimate sales-related data for a predetermined period (for example, two weeks, but not limited thereto) from that day.

[0093] (8) The program for the learning estimation system 100 of the present disclosure to convert sales-related data of a store into teacher data, learn, and predict sales-related data of the store may be provided in a state stored in a computer-readable storage medium. The storage medium is capable of storing the program in a "non-transient tangible medium." The storage medium may include any suitable storage medium such as an HDD or SSD, or a suitable combination of two or more of these. The storage medium may be volatile, non-volatile, or a combination of volatile and non-volatile. Note that the storage medium is not limited to these examples, and may be any device or medium capable of storing the program.

[0094] The learning estimation system 100 can realize the functions of the multiple functional units shown in each embodiment by, for example, reading a program stored in a storage medium and executing the read program. The program may be provided to the learning estimation system 100 via any transmission medium (such as a communication network or broadcast waves). The learning estimation system 100 realizes the functions of the multiple functional units shown in each embodiment by, for example, executing a program downloaded via the Internet or the like. This program may be executed by the learning estimation system 100 or the like.

[0095] The program can be implemented using, for example, a scripting language such as ActionScript or JavaScript (registered trademark), an object-oriented programming language such as Objective-C, Java (registered trademark), or Python (registered trademark), or a markup language such as HTML5, but is not limited to these.

[0096] At least a part of the processing in the learning estimation system 100 may be realized by cloud computing configured by one or more computers. In addition, each functional unit of the learning estimation system 100 may be realized by one or more circuits that realize the functions shown in the above embodiments, and the functions of multiple functional units may be realized by one circuit.

[0097] (9) The above embodiment and each modification may be appropriately combined within the scope of the purpose of generating and learning teacher data for sales-related data of a store and predicting the sales-related data of the store.

[0098] (10) According to each aspect of the present disclosure described above, it is possible to estimate store sales-related data and optimize store operations, thereby contributing to the achievement of Goal 9 of the Sustainable Development Goals (SDGs), which is to “build resilient infrastructure, promote inclusive and sustainable industrialization, and promote innovation and infrastructure.” [Explanation of symbols]

[0099] 100 Learning Estimation System 110 Communications Department 120 Input section 130 Control section 131 Acquisition Department 132 Conversion unit 133 Learning Department 134 Reception Department 135 Data Acquisition Section 136 Estimation Department 140 Storage section 150 Output section 200 store terminals 210 Communications Department 220 Input section 230 Control Unit 231 Sales-related data transmission department 232 Display control unit 240 Storage section 250 Output section

Claims

1. an acquisition unit that acquires store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; A conversion unit that converts the data acquired by the acquisition unit into teacher data; A learning unit that causes a machine learning model that estimates sales-related data of a store based on at least the location of the store and the type of the store to learn the training data; Equipped with The conversion unit performs logarithmic transformation on the sales-related data acquired by the acquisition unit, performs logarithmic transformation on the sales-related data for different days, and converts the difference between the two sales-related data after logarithmic transformation into the training data.

2. The acquisition unit further acquires weather data for the date, The machine learning model is further a model that estimates sales-related data of a store based on weather data. The learning estimation system according to claim 1 .

3. The acquisition unit further acquires people flow data at the location of the store on the date, The machine learning model is further a model that estimates sales-related data of a store based on people flow data. The learning estimation system according to claim 1 .

4. The acquisition unit further acquires calendar information indicating whether the date is a holiday or a weekday on a calendar and whether a specific event occurred on the date; The machine learning model further estimates store sales-related data based on calendar information. The learning estimation system according to claim 1 .

5. The acquisition unit further acquires sales promotion data related to sales promotion of the store on the date, The machine learning model is further a model that estimates sales-related data of a store based on sales promotion data. The learning estimation system according to claim 1 .

6. When the sales-related data estimated using the machine learning model is sales-related data after a predetermined number of days, the conversion unit performs a logarithmic transformation on the sales-related data acquired by the acquisition unit, and also performs a logarithmic transformation on the sales-related data acquired by the acquisition unit before the predetermined number of days, and converts the difference between the two sales-related data after the logarithmic transformation into the training data. The learning estimation system according to claim 1 .

7. The conversion unit converts the difference between the two pieces of sales-related data into the training data by performing an inverse logarithmic conversion.

7. The learning estimation system according to claim 6.

8. When the sales-related data estimated using the machine learning model is sales-related data after a predetermined date, the conversion unit converts, as the teacher data, information based on a difference between the sales-related data acquired by the acquisition unit and sales-related data acquired before the predetermined date of the sales-related data. The learning estimation system according to any one of claims 1 to 7.

9. The computer An acquisition step of acquiring store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; a conversion step of converting the data acquired in the acquisition step into teacher data; A learning step of learning the training data into a machine learning model that estimates sales-related data of a store based on at least the location of the store and the type of the store; Run The conversion step is a learning estimation method in which the sales-related data acquired in the acquisition step is logarithmically transformed, sales-related data for different days is logarithmically transformed, and the difference between the two sales-related data after logarithmic transformation is converted into the training data.

10. On the computer, An acquisition function for acquiring store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; A conversion function for converting the data acquired by the acquisition function into teacher data; A learning function for learning the training data into a machine learning model that estimates sales-related data of a store based on at least the location of the store and the type of the store; Realize this, The conversion function is a learning estimation program that logarithmically transforms the sales-related data acquired by the acquisition function, logarithmically transforms the sales-related data for different days, and converts the difference between the two sales-related data after logarithmic transformation into the training data.

11. a reception unit that receives input of date data relating to the location of the store, the type of the store, and the date for which sales-related data is to be estimated; an acquisition unit that acquires store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; a compensation unit that compensates for the sales-related data for a specific date when the acquisition unit is unable to acquire the sales-related data for the specific date; an estimation unit that inputs the date data received by the reception unit into a machine learning model that estimates sales-related data of a store based on at least the location of the store and the type of business of the store, and estimates sales-related data for the store on the date indicated by the date data; an output unit that outputs the sales-related data estimated by the estimation unit, The estimation unit estimates sales-related data for the date indicated by the date data based on the sales-related data filled in by the filling unit. Learning estimation system.

12. a data acquisition unit that acquires meteorological data for the location of the store based on the location of the store; The machine learning model is further a model that estimates sales-related data of a store based on weather data, The estimation unit further inputs the weather data acquired by the data acquisition unit into the machine learning model to estimate sales-related data. The learning estimation system according to claim 11 .

13. a data acquisition unit that acquires people flow data at the store location based on the store location; The machine learning model is further a model that estimates sales-related data of a store based on people flow data, The estimation unit further inputs the people flow data acquired by the data acquisition unit into the machine learning model to estimate sales-related data. The learning estimation system according to claim 11 .

14. a data acquisition unit that acquires calendar information indicating whether the date indicated by the date data is a holiday or a weekday on a calendar and whether a specific event occurred on the date; The machine learning model is further a model that estimates sales-related data of a store based on calendar information, The estimation unit further inputs the calendar information acquired by the data acquisition unit into the machine learning model to estimate sales-related data. The learning estimation system according to claim 11 .

15. a data acquisition unit that acquires sales promotion data related to sales promotion of the store on the date indicated by the date data; The machine learning model is further a model for estimating sales-related data of a store based on sales promotion data, The estimation unit further inputs the sales promotion data acquired by the data acquisition unit into the machine learning model to estimate sales-related data. The learning estimation system according to claim 11 .

16. The compensation unit compensates for the sales-related data of the specific date by using sales-related data of the store that is past sales data of the store and is for the same day of the week as the specific date. The learning estimation system according to claim 11 .

17. A computer comprising: a receiving step of receiving input of date data relating to the location of the store, the type of the store, and the date for which sales-related data is to be estimated; An acquisition step of acquiring store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; a compensation step of compensating for sales-related data on a specific date when the acquisition step is unable to acquire the sales-related data on the specific date; an estimation step of inputting the date data received in the receiving step into a machine learning model that estimates sales-related data of a store based on at least a store location and a store type, and estimating sales-related data for the store on the date indicated by the date data; an output step of outputting the sales-related data estimated in the estimation step; The estimation step estimates sales-related data for the date indicated by the date data based on the sales-related data compensated for in the compensation step. Learning estimation methods.

18. A computer comprising: a reception function for receiving input of date data relating to the location of the store, the type of the store, and the date for which sales-related data is to be estimated; An acquisition function for acquiring store location data indicating a location of a store, store format data indicating a format of the store, and sales-related data relating to daily sales of the store; a compensation function for compensating for sales-related data on a specific date when the acquisition function is unable to acquire the sales-related data on the specific date; an estimation function that inputs the date data received by the reception function into a machine learning model that estimates sales-related data of a store based on at least the location of the store and the type of business of the store, and estimates sales-related data for the store on the date indicated by the date data; an output function for outputting the sales-related data estimated by the estimation function; The estimation function estimates sales-related data for the date indicated by the date data based on the sales-related data compensated for by the compensation function. Learning estimation program.

Images