Apparatus, method and recording medium for predicting product demand

Abstract

An apparatus according to one embodiment of the present disclosure comprises one or more processors and one or more memories in which instructions to be executed by the one or more processors are stored. When the instructions are executed, the one or more processors may be configured to identify one or more promotion occurrence intervals in which a promotion was applied to the product in a first time interval in which the product was sold, determine a promotion weight value of the product based on the demand amount of the product in the one or more promotion occurrence intervals, obtain information indicating a first predicted demand amount of the product in a second time interval following the first time interval, wherein the first predicted demand amount is calculated using a learning model trained to predict the demand amount of the product, and generate information regarding a second predicted demand amount of the product based on the promotion weight value of the product and the information indicating the first predicted demand amount of the product.

Description

Technology Field

[0001] The present disclosure relates to a technology for processing information regarding product demand quantities. Background Technology

[0002] Due to recent advancements in communication technology, e-commerce services that enable users to purchase goods online are expanding extensively. A wide variety of products are sold on these services, allowing buyers to select and purchase them. Sales volumes for goods sold on e-commerce services can vary significantly depending on the type of product or the availability of promotions, and may also fluctuate due to the influence of rapidly changing consumer preferences.

[0003] When various promotions are applied to a product, it is common for the product to be sold at a discounted price. Consequently, buyers may be more likely to purchase the product on days when promotions are in effect compared to days without promotions. Therefore, predicting the demand for a product during a promotion period can improve the efficiency of inventory management and demand processing. The problem to be solved

[0004] At least one embodiment of the present disclosure trains a learning model that predicts the demand for a product and can predict the demand for a product using the learning model.

[0005] At least one embodiment of the present disclosure can determine the weighting value for each type of promotion applied to the product in the first time interval.

[0006] At least one embodiment of the present disclosure can generate information regarding the predicted demand quantity of each type of promotion to be applied to a product in a second time interval based on the weight value of each type of promotion.

[0007] The technical problems of the present disclosure are not limited to those mentioned above, and other unmentioned technical problems will be clearly understood by a person skilled in the art of the present disclosure from the description below. means of solving the problem

[0008] The above processor may be configured to include one or more memories in which instructions to be executed by the above processor are stored, and when the instructions are executed, the one or more processors identify one or more promotion occurrence intervals in which a promotion was applied to the product in a first time interval in which the product was sold, determine a promotion weight value of the product based on the demand amount of the product in the one or more promotion occurrence intervals, obtain information indicating a first predicted demand amount of the product in a second time interval following the first time interval - the first predicted demand amount is calculated using a learning model learned to predict the demand amount of the product - and generate information regarding a second predicted demand amount of the product based on the promotion weight value of the product and the information indicating the first predicted demand amount of the product.

[0009] In one embodiment, each of the one or more promotion occurrence intervals includes a promotion application interval where a promotion was applied to the product and one or more general sales intervals that are continuous with the promotion application interval and where no promotion was applied to the product, and the one or more processors may be configured to determine a promotion weight value of the product based on information indicating the demand amount of the product in the promotion application interval and information indicating the demand amount of the product in the one or more general sales intervals.

[0010] In one embodiment, among the one or more general sales periods, the number of days belonging to the first general sales period preceding the promotion application period and the number of days belonging to the second general sales period following the promotion application period may be set to be equal to each other.

[0011] In one embodiment, the one or more processors may be configured to calculate the ratio of the sum of the demand quantities of the product in the one or more general sales periods to the sum of the demand quantities of the product in the promotion application period, and to determine the calculated ratio as the promotion weight value.

[0012] In one embodiment, the one or more processors may be configured to calculate the ratio of the sum of the demand quantities of the product in the one or more general sales periods to the sum of the demand quantities of the product in the promotion application period by determining different time coefficients corresponding to each day belonging to the promotion occurrence period and calculating the ratio by summing the values ​​obtained by multiplying the demand quantities corresponding to each day belonging to the promotion occurrence period by the different time coefficients.

[0013] In one embodiment, the value of the time factor may be determined based on the length between the day belonging to the promotion occurrence interval and the reference point distinguishing the first time interval and the second time interval.

[0014] In one embodiment, the one or more processors may be configured to determine the promotion weight value of the product according to the type of promotion applied to the product in the first time interval.

[0015] In one embodiment, the information regarding the second predicted demand quantity may include information indicating the predicted demand quantity for each type of promotion to be applied to the product in the second time interval.

[0016] In one embodiment, the one or more processors may be configured to identify the type of promotion to be applied to the product and the date on which the promotion for each type of promotion is applied in the second time interval, and to generate information indicating the predicted demand for each type of promotion by multiplying the predicted demand for the date on which the promotion is applied by the promotion weight value corresponding to the type of promotion to be applied to the product.

[0017] In one embodiment, the one or more processors may be configured to determine a promotion weight value to be multiplied by a predicted demand quantity corresponding to a date on which multiple promotions are applied in duplicate, based on a predetermined criterion, upon determining that there exists a date on which multiple promotions are applied in duplicate among the dates on which the promotion is applied.

[0018] A method performed in a device comprising one or more processors and one or more memories for storing instructions to be executed by said one or more processors according to one embodiment of the present disclosure, wherein, at the execution of said instructions, said method, said method may include: an operation in which, said one or more processors identify one or more promotion occurrence intervals in which a promotion was applied to said product in a first time interval in which said product was sold; an operation in which a promotion weight value of said product is determined based on the sales volume of said product in said one or more promotion occurrence intervals; an operation in which information indicating a first predicted demand amount of said product in a second time interval following said first time interval - said first predicted demand amount is calculated using a learning model trained to predict the demand amount of said product -; and an operation in which information regarding a second predicted demand amount of said product according to the promotion weight value of said product and the information indicating the first predicted demand amount of said product is generated based on said promotion occurrence.

[0019] In one embodiment, each of the one or more promotion occurrence sections includes a promotion application section where a promotion was applied to the product and one or more general sales sections that are continuous with the promotion application section and where a promotion was not applied to the product, and the operation of determining the promotion weight value of the product may include: an operation of obtaining information indicating the demand amount of the product in the promotion application section and information indicating the demand amount of the product in the one or more general sales sections; an operation of calculating the ratio of the sum of the demand amounts of the product in the one or more general sales sections to the sum of the demand amounts of the product in the promotion application section; and an operation of determining the calculated ratio as the promotion weight value.

[0020] In one embodiment, the operation of calculating the ratio of the sum of the demand quantities of the product in one or more general sales periods to the sum of the demand quantities of the product may include: the operation of determining different time coefficients corresponding to each day belonging to the promotion occurrence period; and the operation of calculating the ratio by summing the values ​​obtained by multiplying the demand quantities corresponding to each day belonging to the promotion occurrence period by the different time coefficients.

[0021] In one embodiment, the value of the time factor may be determined based on the length between the day belonging to the promotion occurrence interval and the reference point distinguishing the first time interval and the second time interval.

[0022] In a non-transient computer-readable recording medium having instructions to be executed by one or more processors according to one embodiment of the present disclosure, the instructions may be configured such that, at the time of execution of the instructions, the one or more processors identify one or more promotion occurrence intervals in which a promotion was applied to the product in a first time interval in which the product was sold, determine a promotion weight value of the product based on the sales volume of the product in the one or more promotion occurrence intervals, obtain information indicating a first predicted demand amount of the product in a second time interval following the first time interval - the first predicted demand amount is calculated using a learning model learned to predict the demand amount of the product - and generate information regarding a second predicted demand amount of the product based on the promotion weight value of the product and the information indicating the first predicted demand amount of the product. Effects of the invention

[0023] According to at least one embodiment of the present disclosure, a learning model for predicting the demand for a product is trained, and the demand for a product can be predicted using the learning model.

[0024] According to at least one embodiment of the present disclosure, a weight value for each type of promotion applied to a product in a first time interval can be determined.

[0025] According to at least one embodiment of the present disclosure, information regarding the predicted demand quantity of each type of promotion to be applied to a product in a second time interval can be generated based on the weight value of each type of promotion.

[0026] The effects according to the technical concept of the present disclosure are not limited to the effects mentioned above, and other unmentioned effects can be clearly understood by a person skilled in the art from the description of the present disclosure. Brief explanation of the drawing

[0027] FIG. 1 illustrates an environment in which a device according to one embodiment of the present disclosure can be applied. FIG. 2 is a block diagram of a device according to one embodiment of the present disclosure. FIG. 3 is a drawing for explaining information indicating a first predicted demand amount according to one embodiment of the present disclosure. FIG. 4 is a drawing for explaining the process of determining the promotion weight of a product according to one embodiment of the present disclosure. FIG. 5 is a diagram illustrating the process of calculating a time coefficient according to one embodiment of the present disclosure. FIG. 6 is a diagram illustrating the process of generating information regarding a second predicted demand amount according to one embodiment of the present disclosure. FIG. 7 is an operation flowchart according to one embodiment of the present disclosure. Specific details for implementing the invention

[0028] The various embodiments described in this disclosure are illustrative for the purpose of clearly explaining the technical concept of this disclosure and are not intended to limit it to specific embodiments. The technical concept of this disclosure includes various modifications, equivalents, alternatives, and embodiments selectively combined from all or part of each embodiment described in this disclosure. Furthermore, the scope of the technical concept of this disclosure is not limited to the various embodiments presented below or the specific descriptions thereof.

[0029] Terms used in this disclosure, including technical or scientific terms, may have the meaning generally understood by those skilled in the art to which this disclosure pertains, unless otherwise defined.

[0030] Expressions used in this disclosure, such as “comprising,” “may compose,” “possessing,” “possessing,” “having,” and “possessing,” mean that the subject feature (e.g., function, operation, or component, etc.) exists and do not exclude the existence of other additional features. That is, such expressions should be understood as open-ended terms implying the possibility of including other embodiments.

[0031] Singular expressions used in this disclosure may include the meaning of the plural form unless otherwise indicated by the context, and this applies likewise to singular expressions described in the claims.

[0032] Expressions such as "first," "second," or "first," "second," etc., used in this disclosure are used to distinguish one object from another when referring to a plurality of objects of the same kind, unless otherwise indicated by the context, and do not limit the order or importance of said objects.

[0033] Expressions used in the present disclosure, such as “A, B, and C”, “A, B, or C”, “A, B, and / or C”, “at least one of A, B, and C”, “at least one of A, B, or C”, “at least one of A, B, and / or C”, etc., may mean each of the listed items or all possible combinations of the listed items. For example, “at least one of A or B” may refer to (1) at least one A, (2) at least one B, and (3) at least one A and at least one B.

[0034] The expression “based on” as used in this disclosure is used to describe one or more factors affecting an act or action of a decision or judgment described in the phrase or sentence containing this expression, and this expression does not exclude additional factors affecting said act or action of a decision or judgment.

[0035] As used in this disclosure, the expression "configured to" may have meanings such as "set to," "capable of," "modified to," "made to," or "capable of." This expression is not limited to the meaning of "specifically designed in hardware." For example, a processor configured to perform a specific operation may mean a generic-purpose processor capable of performing that specific operation by executing software, or a special-purpose computer structured through programming to perform that specific operation.

[0036] Hereinafter, various embodiments described in this disclosure will be explained with reference to the attached drawings. In the attached drawings and the description thereof, identical or substantially equivalent components may be given the same reference numerals. Furthermore, in the description of the various embodiments below, the description of identical or corresponding components may be omitted, but this does not mean that such components are not included in the embodiments.

[0037] FIG. 1 illustrates an environment in which a device according to one embodiment of the present disclosure may be applied. The device (100) and the server device (200) may be connected via a network and communicate with each other. The server device (200) may be a server device that manages information about products sold in an e-commerce service or a server device that manages inventory regarding products sold in an e-commerce service. Here, the e-commerce service may be any service that responds to a consumer's purchase request, sells products corresponding to the purchase request, and manages or processes information related thereto.

[0038] The device (100) may be a device for managing information regarding the demand quantity of goods sold in an e-commerce service. For example, the device (100) may be a device for managing information indicating the demand quantity of goods in the past or information indicating the demand quantity of goods in the future. Additionally, the device (100) may be a device for generating and managing information indicating the predicted demand quantity of goods in the future. The device (100) may be implemented as a terminal capable of transmitting and receiving information to and from a server device (200) via a network. For example, the device (100) may be implemented as a computer, a laptop, a portable communication terminal (such as a smartphone), a portable multimedia device, etc. However, the type of device (100) is not limited thereto, and the device (100) may be any device capable of including an input / output interface capable of outputting information and communicating with the server device (200) or other devices or modules via a network. The device (100) may obtain various information regarding the demand quantity of goods from the server device (200). For example, the device (100) can obtain information indicating the actual demand quantity for each product and can obtain information indicating the actual demand quantity for each product in a specific time interval. In addition, the device (100) can obtain information indicating the type of promotion or the duration of the promotion that was applied to each product in a specific time interval.

[0039] The device (100) can provide various information to the user indicating the demand for a product and can receive various inputs from the user to predict the demand for a product. For example, the device (100) can obtain a command from the user through an input / output interface to generate information regarding the predicted demand for each product in the future, and in response to the obtained command, generate information regarding the predicted demand for a product following the occurrence of a promotion, or transmit the generated information regarding the predicted demand to another device or module. Here, the information regarding the predicted demand may include, but is not limited to, information regarding the predicted demand for each product following the occurrence of a promotion when a promotion is applied to the product in the future. The input obtained from the user may include various forms of input, such as a mouse click, a touch using a touchpad or touchscreen, voice recognition, or other electronic inputs.

[0040] The device (100) can be connected to a server device (200) or other devices or modules via a network to communicate with each other. For example, the device (100) and the server device (200), or the device (100) and other devices or modules, can be connected to each other via a network that provides a connection path to transmit and receive packet data. The network can be implemented as any type of wired or wireless network, such as a Local Area Network (LAN), a Wide Area Network (WAN), a Mobile Radio Communication Network, or a Wibro (Wireless Broadband Internet).

[0041] The server device (200) can store various information to generate information regarding the predicted demand for each product sold in the e-commerce service. For example, the server device (200) can store information regarding the actual demand for each product sold in the e-commerce service. Specifically, the server device (200) can store information regarding the actual demand for each product sold in the e-commerce service by time of day, day, week, month, quarter, and period. Additionally, the server device (200) can store information regarding the demand for each product during periods when a promotion was applied to each product and during general sales periods when no promotion was applied to each product.

[0042] FIG. 2 is a block diagram of a device (100) according to one embodiment of the present disclosure. The device (100) may include one or more processors (110) and / or one or more memories (120) as components. Additionally, in one embodiment, at least one of the components of the device (100) may be omitted, or other components may be added to the device (100). In one embodiment, additionally or alternatively, some components may be implemented as a single or multiple entities. In the present disclosure, one or more processors (110) may be referred to as processors (110). Unless the context clearly indicates otherwise, the expression processors (110) may mean a set of one or more processors. In the present disclosure, one or more memories (120) may be referred to as memories (120). Unless the context clearly indicates otherwise, the expression memories (120) may mean a set of one or more memories. In one embodiment, at least some of the internal / external components of the device (100) may be connected to each other via a bus, GPIO (General Purpose Input / Output), SPI (Serial Peripheral Interface), or MIPI (Mobile Industry Processor Interface), etc., to exchange information (data, signals, etc.).

[0043] In one embodiment, the device (100) may further include a communication circuit (130). The device (100) may communicate with a terminal (200) through the communication circuit (130). That is, the communication circuit (130) may be configured to perform wireless or wired communication between the device (100) and the terminal (200). For example, the communication circuit (130) can perform wireless communication according to methods such as eMBB (enhanced Mobile Broadband), URLLC (Ultra Reliable Low-Latency Communications), MMTC (Massive Machine Type Communications), LTE (Long-Term Evolution), LTE-A (LTE Advance), NR (New Radio), UMTS (Universal Mobile Telecommunications System), GSM (Global System for Mobile communications), CDMA (Code Division Multiple Access), WCDMA (Wideband CDMA), WiBro (Wireless Broadband), WiFi (Wireless Fidelity), Bluetooth, NFC (Near Field Communication), GPS (Global Positioning System), or GNSS (Global Navigation Satellite System). For example, the communication circuit (130) can perform wired communication according to methods such as USB (Universal Serial Bus), HDMI (High Definition Multimedia Interface), RS-232 (Recommended Standard-232), or POTS (Plain Old Telephone Service). In one embodiment, the device (100) may be implemented by being integrated with another device.In this case, the communication circuit (130) can function as a connection circuit or interface connecting the device (100) and the other device.

[0044] The processor (110) can control at least one component of the device (100) connected to the processor (110) by running software (e.g., instructions, programs, etc.). Additionally, the processor (110) can perform various operations related to the present disclosure, such as computation, processing, data generation, and processing. Furthermore, the processor (110) can load data, etc. from memory (120) or store it in memory (120). Moreover, the processor (110) can transmit and receive various information to and from a terminal (200) through a communication circuit (130). In one embodiment, the processor (110) can control the communication circuit (130) to transmit various information to the terminal (200) in various forms.

[0045] The memory (120) can store various information (data). The information stored in the memory (120) is information acquired, processed, or used by at least one component of the device (100), and may include software (e.g., instructions, programs, etc.). The memory (120) may include volatile and / or non-volatile memory. In the present disclosure, instructions or programs are software stored in the memory (120) and may include an operating system for controlling the resources of the device (100), an application, and / or middleware that provides various functions to the application so that the application can utilize the resources of the device (100). In one embodiment, the memory (120) may store instructions that cause the processor (110) to perform calculations when executed by the processor (110). The memory (120) may store at least a portion of information received from the terminal (200) through the communication circuit (130) and / or information transmitted to the terminal (200) through the communication circuit (130). For example, the memory (120) may include information for predicting the predicted demand for each product, such as the type of promotion to be applied to the product, the period of the promotion to be applied to the product, and information indicating the predicted demand in the period when no promotion is applied to the product, but is not limited thereto.

[0046] FIG. 3 is a diagram illustrating information indicating a first predicted demand amount according to an embodiment of the present disclosure. Information indicating a first predicted demand amount can be calculated using a learning model (10) trained to predict the demand amount of a product. Here, the learning model (10) is a learning model trained based on data having various attributes and may include various artificial neural network models or machine learning models. In one embodiment, the learning model (10) may include a Gradient Boosting Model (GBM) model, which is an ensemble model of boosting techniques, but is not limited to such examples.

[0047] In one embodiment, the processor (110) may generate training data to train a learning model (10). The processor (110) may generate training data having various features. The training data may include various features necessary for predicting the demand for a product. For example, the processor (110) may generate data representing the attributes of a product as training data that is necessary for predicting the demand for each product. The processor (110) may generate product data (SKU_inf) for identifying a product as part of the training data. In one embodiment, the product data (SKU_inf) may include data regarding a product ID, etc., for identifying each product. The processor (110) may generate date data (Date_inf) representing the sales date of a product as part of the training data that is necessary for predicting the demand for each product. In one embodiment, the date data (Date_inf) may include index information for identifying the month, week, and day of the date the product was sold, and may include index information for identifying the season to which the date the product was sold belongs. The processor (110) may generate statistical data representing statistical indicators of the predicted demand for a product as part of the training data, which is necessary for predicting the demand for each product. In one embodiment, the statistical data may include mean data (Mean_inf), standard deviation data (STD_inf), quartile data (Quantile_inf), etc., between the predicted demand for a product in a specific time interval. The processor (110) may generate promotion data representing promotions applied to a product as part of the training data, which is necessary for predicting the demand for each product. In one embodiment, the promotion data may include data representing the date on which a promotion was applied to a product (Date info_promo), etc.The processor (110) may generate data related to product stock shortages as part of the training data, which is necessary for predicting the demand for each product. In one embodiment, the data related to product stock shortages may include data (Date info_OOS) indicating the date on which a product stock shortage occurred.

[0048] In one embodiment, the processor (110) may be configured to train a learning model (10) based on generated learning data. The processor (110) may be configured to generate a learning data set, which is a set of various learning data, and to train a learning model (10) based on the various learning data set. The output produced as a result of training the learning model (10) may include information indicating a first predicted demand amount for a product in the future. For example, the information indicating the first predicted demand amount may include information indicating the predicted demand amount for each date of the product belonging to the second time interval (T2) described later. The information indicating the first predicted demand amount may include information indicating the predicted demand amount (a1, a2, a3, a4, a6) for the product for each date (D1, D2, D3, D4, Dm) belonging to the second time interval (T2). Here, the information indicating the first predicted demand quantity may be information indicating the predicted demand quantity when assuming that no promotion is applied to the products for each date belonging to the second time interval (T2). The processor (110) can obtain information indicating the first predicted demand quantity calculated using the learning model (10).

[0049] FIG. 4 is a diagram illustrating the process of determining a promotion weight of a product according to one embodiment of the present disclosure. Referring to FIG. 4, a one-dimensional graph is shown in which the horizontal axis represents time. For example, the graph shown in FIG. 4 may represent a promotion occurrence period for a specific product during a first time interval (T1) in which a specific product was sold. A processor (110) may determine a promotion weight value of a product based on the demand for the product during the first time interval (T1). The processor (110) may obtain information indicating the demand for the product during the first time interval (T1) and information regarding the promotion applied to the product during the first time interval (T1) from a server device (200).

[0050] The processor (110) can identify a first time interval (T1) based on a reference point (d0). In one embodiment, the reference point (d0) may be a point in time that distinguishes the first time interval (T1) and the second time interval (T2). That is, the first time interval (T1) may be a past time interval distinguished based on the reference point (d0), and the second time interval (T2) may be a future time interval distinguished based on the reference point (d0). In one embodiment, the first time interval (T1) may be set to have various lengths, for example, the first time interval (T1) may be set to a time interval up to 3 months prior to the reference point (d0), a time interval up to 6 months prior, a time interval up to 1 year prior, etc.

[0051] In one embodiment, the processor (110) can identify one or more promotion occurrence periods in which a promotion was applied to a product during a first time interval (T1). Here, a promotion occurrence period may be a period in which the demand for the product is higher than usual due to the application of an event such as a discount on the product. The processor (110) can identify promotion occurrence periods (1, 2, 3, 4, 5) in which a promotion was applied to a product during the first time interval (T1). For example, the processor (110) can identify the start date and end date of each promotion occurrence period (1, 2, 3, 4, 5), and the type of promotion applied to the product during each promotion occurrence period (1, 2, 3, 4, 5). Promotions may be classified into different types based on the discount rate applied to the product price, or based on a method for discounting the product price, such as an immediate discount promotion that discounts the product price or a coupon promotion that issues a coupon to discount the product price.

[0052] In one embodiment, the processor (110) may determine a product promotion weight value based on the demand for the product in promotion occurrence intervals (1, 2, 3, 4, 5). The processor (110) may determine a product promotion weight value based on the type of promotion applied to the product. For example, the processor (110) may determine a product first promotion weight value based on the demand for the product in the first promotion occurrence interval (1, 3, 5) where a first promotion applied to the product occurred. Additionally, the processor (110) may determine a product second promotion weight value based on the demand for the product in the second promotion occurrence interval (2) where the product was applied. Additionally, the processor (110) may determine a product third promotion weight value based on the demand for the product in the third promotion occurrence interval (4) where the product was applied. Each promotion weight value can be used to generate information regarding the second predicted demand amount resulting from the occurrence of a promotion to be applied to the product in the second time interval (T2), along with information indicating the first predicted demand amount calculated by the learning model (10).

[0053] In one embodiment, each of the promotion occurrence intervals (1, 2, 3, 4, 5) may include a promotion application interval where a promotion was applied to a product, and one or more general sales intervals that are continuous with the promotion application interval and where no promotion was applied to the product. For example, each of the first promotion occurrence intervals (1, 3, 5) where a first promotion was applied to a product in the first time interval (T1) may include a promotion application interval (1a, 3a, 5a) where the first promotion was actually applied to a product, and a general sales interval (1b, 1c, 3b, 3c, 5a, 5c) that is continuous with each of the promotion application intervals (1a, 3a, 5a) where the first promotion was applied. Additionally, the second promotion application section (2), in which the second promotion was applied to the product during the first time interval (T1), may include a promotion application section (2a) in which the second promotion was actually applied to the product and a general sales section (2b, 2c) that is continuous with the promotion application section (2a) in which the second promotion was applied. Additionally, the third promotion application section (4), in which the third promotion was applied to the product during the first time interval (T1), may include a promotion application section (4a) in which the third promotion was actually applied to the product and a general sales section (4b, 4c) that is continuous with the promotion application section (2a) in which the second promotion was applied.

[0054] In one embodiment, the number of days belonging to one or more general sales periods consecutive to each promotion application period in which a promotion was applied to a product may be set to be equal to each other. For example, the number of days belonging to the general sales period (1b) preceding the first promotion application period (1a) and the number of days belonging to the general sales period (1c) following the first promotion application period (1a) may be set to be equal to each other. Additionally, the number of days belonging to the general sales period (2b) preceding the second promotion application period (2a) and the number of days belonging to the general sales period (2c) following the second promotion application period (2b) may be set to be equal to each other. Similarly, the number of days belonging to the general sales period (4b) preceding the third promotion application period (4a) and the number of days belonging to the general sales period (4c) following the third promotion application period (4a) may be set to be equal to each other. In one embodiment, the number of days belonging to each of the general sales intervals (1b, 1c, 2b, 2c, 3b, 3c, 4b, 4c, 5b, 5c) may be, for example, 7 days, but is not limited to this example.

[0055] If there is a day when a shortage of product stock occurs in one or more general sales sections consecutive to a promotion application section, the processor (110) may exclude the day when a shortage of product stock occurs in the general sales section and add general sales sections equal to the number of excluded days. In one embodiment, among the general sales sections (4b, 4c) consecutive to the third promotion application section (4a) to which the third promotion is applied, a shortage of product stock may occur in the general sales section (4c) following the third promotion application section (4a). The processor (110) may exclude the day (4d) when a shortage of product stock occurred from the general sales section (4c) and add general sales sections (4e) equal to the number of excluded days. In this case, the number of days belonging to the general sales section (4b) preceding the third promotion application section (4a) and the number of days belonging to the general sales section following the third promotion application section (4a) may be set to be the same.

[0056] In one embodiment, the processor (110) may determine a promotion weight value of a product based on information indicating the demand for the product in a promotion application period and information indicating the demand for the product in a general sales period. For example, the processor (110) may determine a first promotion weight value based on information indicating the demand for the product in a promotion application period (1a, 3a, 5a) where a first promotion was applied to the product in a first time period (T1) and information indicating the demand for the product in a general sales period (1b, 1c, 3b, 3c, 5b, 5c) where the first promotion was not applied to the product in the first time period (T1). Additionally, the processor (110) may determine a second promotion weight value based on information indicating the demand for the product in a promotion application period (2a) where a second promotion was applied to the product in the first time period (T1) and information indicating the demand for the product in a general sales period (2b, 2c) where the second promotion was not applied to the product in the first time period (T1). Additionally, the processor (110) can determine a third promotion weight value based on information indicating the demand for a product in a promotion application period (4a) where a third promotion was applied to the product in a first time period (T1) and information indicating the demand for a product in a general sales period (4b, 4c) where a third promotion was not applied to the product in the first time period (T1). The processor (110) can calculate the ratio of the sum of the demand for a product in one or more general sales periods to the sum of the demand for a product in each promotion application period, and determine the calculated ratio as a promotion weight value.For example, the processor (110) can calculate the ratio of the sum of the demand for products in general sales sections (1b, 1c, 3b, 3c, 5b, 5c) where the first promotion was not applied to the products to the sum of the demand for products in promotion sections (1a, 3a, 5a) where the first promotion was applied to the products, and determine the calculated ratio as the first promotion weight value. Additionally, the processor (110) can calculate the ratio of the sum of the demand for products in general sales sections (2b, 2c) where the second promotion was not applied to the products to the sum of the demand for products in promotion sections (2a) where the second promotion was applied to the products, and determine the calculated ratio as the second promotion weight value. Additionally, the processor (110) can calculate the ratio of the sum of the demand for the product in the general sales period (4b, 4c, 4e) where the third promotion was not applied to the product to the demand for the product in the promotion application period (4a) where the third promotion was applied to the product, and determine the calculated ratio as the third promotion weight value.

[0057] FIG. 5 is a diagram illustrating the process of calculating a time factor according to an embodiment of the present disclosure. A processor (110) may calculate the ratio of the sum of the demand quantities of a product in one or more general sales sections to the sum of the demand quantities of a product in each promotion application section, and determine the calculated ratio as a promotion weight value. In one embodiment, the processor (110) may be configured to calculate the ratio by further considering a time factor in the process of calculating the ratio of the sum of the demand quantities of a product in one or more general sales sections to the sum of the demand quantities of a product in each promotion application section. For example, since consumer consumption tendencies or consumption trends may change over time, even if the same type of promotion is applied to the same product, the trend of increase in demand due to the occurrence of the promotion may differ. Based on the time factor, the processor (110) may determine the promotion weight value by giving a relatively weight to a promotion that occurred at a time close to the reference point (d0) and giving a relatively small weight to a promotion that occurred at a time far from the reference point (d0).

[0058] The processor (110) can determine different time coefficients corresponding to each day belonging to the promotion occurrence interval. For convenience of explanation, the promotion occurrence interval (3) among the promotion occurrence intervals (1, 3, 5) where the first promotion occurred on the product in the first time interval (T1) is described as an example, but is not limited thereto. In one embodiment, the processor (110) can determine different time coefficients corresponding to each day belonging to the promotion occurrence interval (3). For example, the processor (110) can determine different time coefficients corresponding to each day belonging to the general sales interval (3c) following the promotion application interval (3a) where the first promotion was applied. The processor (110) can determine a time factor (TW1) corresponding to the day (da) closest to the reference point (d0) among the days belonging to the general sales period (3c), determine a time factor (TW2) corresponding to the day (db) adjacent to the day (da) closest to the reference point (d0) among the days belonging to the general sales period (3c), and determine a time factor corresponding to the previous day (dc) adjacent to the day (db). Likewise, the processor (110) can determine time factors (TW1, TW2, TW3, TW k) corresponding to each of the days belonging to the promotion occurrence period (3).

[0059] The value of the time factor can be determined based on the length between the day belonging to the promotion occurrence interval and the reference point (d0). For example, the value of the time factor (TW1) corresponding to the day (da) closest to the reference point (d0) among the days belonging to the promotion occurrence interval (3) may be the largest, and the value of the time factor (TWk) corresponding to the day (dk) furthest from the reference point (d0) may be the smallest. In one embodiment, the function determining the time factor may be expressed in the form of an exponential function in which the corresponding value decreases as the input value increases, but is not limited thereto. In determining the first promotion weight value, the processor (110) may determine the time factor for all days belonging to each of the promotion occurrence intervals (1, 5).

[0060] In the process of calculating the ratio of the sum of the demand for products in one or more general sales sections to the sum of the demand for products in each promotion application section and determining the calculated ratio as a promotion weight value, the processor (110) may calculate the ratio based on the value obtained by multiplying the demand for each day belonging to the promotion occurrence section by a different time factor. For example, in the process of determining the sum of the demand for products in the promotion application sections (1a, 3a, 5a), the processor (110) may sum the values ​​obtained by multiplying the demand for each day belonging to the promotion application sections (1a, 3a, 5a) by the determined different time factors. Additionally, in the process of determining the sum of the demand quantities of products in the general sales interval (1b, 1c, 3b, 3c, 5b, 5c), the processor (110) may sum the values ​​obtained by multiplying the demand quantities corresponding to each day in the general sales interval (1b, 1c, 3b, 3c, 5b, 5c) by different determined time factors. The processor (110) may determine the ratio of the value obtained by multiplying the demand quantities corresponding to each day in the general sales interval (1b, 1c, 3b, 3c, 5b, 5c) by different determined time factors to the value obtained by multiplying the demand quantities corresponding to each day in the promotion application interval (1a, 3a, 5a), calculated by further considering the time factors, by different determined time factors, as the first promotion weighting value.

[0061] FIG. 6 is a diagram illustrating a process for generating information regarding a second predicted demand amount according to an embodiment of the present disclosure. A processor (110) may be configured to generate information regarding a second predicted demand amount based on information indicating a determined promotion weight value and a first predicted demand amount, in accordance with the occurrence of a promotion in a second time interval (T2) following a first time interval (T1).

[0062] In one embodiment, information regarding the second predicted demand quantity may include information indicating the predicted demand quantity for each type of promotion to be applied to the product in the second time interval (T2). Based on the information indicating the first predicted demand quantity, the processor (110) may obtain the predicted demand quantity for each date included in the second time interval (T2), identify the date on which the promotion is to be applied in the second time interval (T2), and generate information regarding the second predicted demand quantity by multiplying the predicted demand quantity corresponding to the date on which the promotion is to be applied by a promotion weight value corresponding to the type of promotion to be applied. For example, it may be assumed that during the second time interval (T2), the interval (6, 7) for which the first promotion for the product is to be applied is scheduled, and the interval (8) for which the second promotion for the product is to be applied is scheduled. In this case, the processor (110) may obtain information regarding the predicted demand quantity for each day belonging to the interval (6, 7) for which the first promotion is to be applied, based on the information indicating the first predicted demand quantity. The processor (110) can determine the predicted demand for a product corresponding to each day in the interval (6, 7) to which the first promotion is applied by multiplying the predicted demand for each day in the interval (6, 7) to which the first promotion is applied by a first promotion weighting value determined based on the demand for the product in the first time interval (T1). Additionally, the processor (110) can obtain information regarding the predicted demand for each day in the interval (8) to which the second promotion is applied based on information indicating the first predicted demand. The processor (110) can determine the predicted demand for a product corresponding to each day in the interval (8) to which the second promotion is applied by multiplying the predicted demand for each day in the interval (8) to which the second promotion is applied by a second promotion weighting value determined based on the demand for the product in the first time interval (T1).

[0063] If there are dates on which different promotions for a product are applied during the second time interval (T2), the processor (110) may determine a promotion weight value to be multiplied to the predicted demand corresponding to the dates on which different promotions are applied, based on predetermined criteria. For example, there may be an overlapping date (di) between the interval (7) on which the first promotion is applied and the interval (8) on which the second promotion is applied during the second time interval (T2). The processor (110) may determine a promotion weight value to be multiplied to the predicted demand corresponding to the overlapping date (di) based on predetermined criteria. For example, if the processor (110) determines the promotion weight value to be multiplied to the predicted demand corresponding to the overlapping date (di) as the first promotion weight value, the processor (110) may determine the predicted demand resulting from the promotion on the overlapping date (di) by multiplying the predicted demand on the overlapping date (di) by the first promotion weight value. Additionally, when the processor (110) determines a second promotion weight value to be multiplied to the predicted demand amount corresponding to the overlapping date (di), the processor (110) can determine the predicted demand amount due to the occurrence of a promotion on the overlapping date (di) by multiplying the second promotion weight value to the predicted demand amount on the overlapping date (di). In one embodiment, a predetermined criterion may be determined based on the ratio of increase in demand amount by type of promotion. For example, if the ratio of increase in the demand amount of a product when a first promotion is applied to the product is greater than the ratio of increase in the demand amount of a product when a second promotion is applied, the processor (110) may select a first promotion weight value when the first promotion and the second promotion overlap. The predetermined criterion may include information indicating the priority by type of promotion based on the ratio of increase in the demand amount of a product due to the occurrence of a promotion.

[0064] An apparatus (100) according to one embodiment of the present disclosure can determine a promotion weight value based on the demand for a product in a promotion occurrence period that was applied to the product in a first time interval (T1), and thus can efficiently predict the demand for each type of promotion to be applied in a second time interval (T2) based on this. In addition, since the apparatus (100) can predict the demand for each product and each type of promotion in the second time interval (T2), it can perform more efficient inventory management and product order management.

[0065] FIG. 7 is a flowchart of operations according to one embodiment of the present disclosure. A method according to one embodiment of the present disclosure may include operations S710 to S740. The operations S710 to S740 described below may be understood as being performed through a device (100) comprising one or more processors (110) and one or more memories (120) storing instructions to be executed by one or more processors (110).

[0066] In operation S710, the processor (110) of the device (100) can identify a promotion occurrence period of a product in a first time interval (T1). In one embodiment, the promotion occurrence period may include a promotion application period in which a promotion was applied to the product and one or more general sales periods that are continuous with the promotion application period and in which no promotion was applied to the product.

[0067] In operation S720, the processor (110) of the device (100) can determine a promotion weight value for the product. In one embodiment, the processor (110) of the device (100) can obtain information indicating the demand for the product in the promotion application period and information indicating the demand for the product in the general sales period. The processor (110) of the device (100) can calculate the ratio of the sum of the demand for the product in one or more general sales periods to the sum of the demand for the product in the promotion application period, and determine the calculated ratio as a promotion weight value. In one embodiment, the processor (110) of the device (100) can determine different time coefficients corresponding to each day belonging to the promotion occurrence period, calculate the ratio by summing the values ​​obtained by multiplying the demand for each day belonging to the promotion occurrence period by the different time coefficients, and determine the calculated ratio as a promotion weight value. Here, the value of the time coefficient can be determined based on the length between the day belonging to the promotion occurrence period and the reference point distinguishing the first time period and the second time period.

[0068] In operation S730, the processor (110) of the device (100) may obtain information indicating a first predicted demand amount of a product in a second time interval (T2). In one embodiment, the information indicating the first predicted demand amount may be information calculated using a learning model trained to predict the demand amount of a product.

[0069] In operation S740, the processor (110) of the device (100) may generate information regarding a second predicted demand amount based on information indicating a promotion weight value and a first predicted demand amount. In one embodiment, the processor (110) of the device (100) may identify a promotion date to be applied to a product in a second time interval (T2), and generate information indicating a predicted demand amount due to the occurrence of a promotion by multiplying the demand amount corresponding to the date to which the promotion is applied by the promotion weight value.

[0070] In the flowchart according to the present disclosure, each step of the method or algorithm is described in a sequential order, but the steps may be performed in any order that can be arbitrarily combined according to the present disclosure, in addition to being performed sequentially. The description of the flowchart of the present disclosure does not exclude changes or modifications to the method or algorithm and does not imply that any step is essential or preferred. In one embodiment, at least some of the steps may be performed in parallel, iteratively, or heuristically. In another embodiment, at least some of the steps may be omitted or other steps may be added.

[0071] Various embodiments according to the present disclosure may be implemented as software on a machine-readable storage medium. The software may be software for implementing the various embodiments described in the present disclosure. The software may be inferred from the various embodiments described in the present disclosure by programmers skilled in the art to which the present disclosure pertains. For example, the software may be a program containing machine-readable instructions (e.g., code or code segments). The machine may be a device capable of operating according to instructions called from the storage medium, for example, a computer. In one embodiment, the machine may be a device (100, see FIG. 2) according to the various embodiments described in the present disclosure. In one embodiment, the processor of the machine may execute the called instruction to cause the components of the machine to perform a function corresponding to this instruction. In one embodiment, the processor may be a processor (110, see FIG. 2) according to the embodiments described in the present disclosure. The storage medium may mean any type of recording medium in which data is stored that can be read by the machine. The storage medium may include, for example, ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. In one embodiment, the storage medium may be a memory (120, see FIG. 2). In one embodiment, the storage medium may be implemented in a distributed form in a computer system connected by a network, etc. In this case, software may be stored and executed in a distributed manner in the computer system, etc. In another embodiment, the storage medium may be a non-transitory storage medium. A non-transitory storage medium refers to a tangible medium that exists regardless of whether data is stored semi-permanently or temporarily, and does not include a signal that propagates transitorily.

[0072] Although the technical concept of the present disclosure has been described by various embodiments above, the technical concept of the present disclosure includes various substitutions, modifications, and alterations that can be made within the scope of understanding of those skilled in the art to which the present disclosure pertains. Furthermore, it should be understood that such substitutions, modifications, and alterations may be included within the scope of the appended claims.

Claims

Claim 1 One or more processors; A device comprising one or more memories storing instructions to be executed by one or more processors, wherein, upon execution of the instructions, the one or more processors identify one or more promotion occurrence periods in which a promotion was applied to the product in a first time period in which the product was sold, determine a promotion weight value of the product based on the demand quantity of the product in each of the one or more promotion occurrence periods and a time factor corresponding to each of the days in the one or more promotion occurrence periods, obtain information indicating a first predicted demand quantity of the product in a second time period following the first time period, wherein the first predicted demand quantity is calculated using a learning model trained to predict the demand quantity of the product, and generate information regarding a second predicted demand quantity of the product following the promotion occurrence based on the promotion weight value of the product and the information indicating the first predicted demand quantity of the product, wherein the time factor corresponding to each of the one or more days in the promotion occurrence periods is a value determined based on the length between each day in the promotion occurrence period and a reference point distinguishing the first time period and the second time period. Claim 2 A device according to claim 1, wherein each of the one or more promotion occurrence intervals comprises a promotion application interval in which a promotion was applied to the product and one or more general sales intervals that are continuous with the promotion application interval and in which a promotion was not applied to the product, and wherein the one or more processors are configured to determine a promotion weight value of the product based on information indicating the demand quantity of the product in the promotion application interval and information indicating the demand quantity of the product in the one or more general sales intervals. Claim 3 A device according to paragraph 2, wherein the number of days belonging to a first general sales section preceding the promotion application section among the one or more general sales sections and the number of days belonging to a second general sales section following the promotion application section are set to be equal to each other. Claim 4 A device configured such that, in paragraph 2, one or more processors calculate the ratio of the sum of the demand quantities of the product in one or more general sales sections to the sum of the demand quantities of the product in the promotion application section, and determine the calculated ratio as the promotion weighting value. Claim 5 In paragraph 4, the device is configured such that, in the process of calculating the ratio of the sum of the demand quantities of the product in the one or more general sales periods to the sum of the demand quantities of the product in the promotion application period, the ratio is calculated by summing the values ​​obtained by multiplying the demand quantities corresponding to each day belonging to the promotion occurrence period by different time coefficients. Claim 6 delete Claim 7 A device according to claim 1, wherein one or more processors are configured to determine a promotion weight value of the product according to the type of promotion applied to the product in the first time interval. Claim 8 In claim 7, the information regarding the second predicted demand quantity includes information indicating the predicted demand quantity for each type of promotion to be applied to the product in the second time interval. Claim 9 A device configured such that, in claim 8, one or more processors identify the type of promotion to be applied to the product and the date on which the promotion for each type of promotion is applied in the second time interval, and generate information indicating the predicted demand for each type of promotion by multiplying the predicted demand for the date on which the promotion is applied by a promotion weight value corresponding to the type of promotion to be applied to the product. Claim 10 A device configured such that, in accordance with the determination that there exists a date on which multiple promotions are applied in overlap among the dates on which the promotions are applied, the one or more processors determine a promotion weight value to be multiplied by a predicted demand quantity corresponding to the date on which multiple promotions are applied in overlap, based on a predetermined criterion. Claim 11 A method performed in a device comprising one or more processors and one or more memories storing instructions to be executed by said one or more processors, wherein, upon execution of said instructions, said one or more processors perform the operation of identifying one or more promotion occurrence periods in which a promotion was applied to said product in a first time period in which said product was sold; the operation of determining a promotion weight value of said product based on the sales volume of said product in each of the days belonging to said one or more promotion occurrence periods and a time coefficient corresponding to each of the days belonging to said one or more promotion occurrence periods; and the operation of obtaining information indicating a first predicted demand quantity of said product in a second time period following said first time period, wherein the first predicted demand quantity is calculated using a learning model trained to predict the demand quantity of said product. A method comprising the operation of generating information regarding a second predicted demand quantity following the occurrence of a promotion of the product based on information indicating a promotion weight value of the product and a first predicted demand quantity of the product, wherein the time coefficient corresponding to each of the days belonging to the one or more promotion occurrence intervals is a value determined based on the length between each of the days belonging to the promotion occurrence intervals and the reference point distinguishing the first time interval and the second time interval. Claim 12 In claim 11, each of the above one or more promotion occurrence intervals includes a promotion application interval where a promotion was applied to the product and one or more general sales intervals that are continuous with the promotion application interval and where a promotion was not applied to the product, and the operation of determining the promotion weight value of the product includes: an operation of obtaining information indicating the demand quantity of the product in the promotion application interval and information indicating the demand quantity of the product in the one or more general sales intervals; an operation of calculating the ratio of the sum of the demand quantities of the product in the one or more general sales intervals to the sum of the demand quantities of the product in the promotion application interval; and an operation of determining the calculated ratio as the promotion weight value. Claim 13 In claim 12, the operation of calculating the ratio of the sum of the demand quantities of the product in one or more general sales periods to the sum of the demand quantities of the product comprises the operation of calculating the ratio by summing the values ​​obtained by multiplying the demand quantities corresponding to each day belonging to the promotion occurrence period by different time coefficients. Claim 14 delete Claim 15 In a non-transient computer-readable recording medium recording instructions to be executed by one or more processors, said instructions are configured such that, upon execution of said instructions, said one or more processors identify one or more promotion occurrence periods in which a promotion was applied to said product in a first time period in which said product was sold, determine a promotion weight value of said product based on the sales volume of said product on each day belonging to said one or more promotion occurrence periods and a time factor corresponding to each day belonging to said one or more promotion occurrence periods, obtain information indicating a first predicted demand quantity of said product in a second time period following said first time period - said first predicted demand quantity is calculated using a learning model trained to predict the demand quantity of said product - and generate information regarding a second predicted demand quantity of said product following the promotion occurrence based on the promotion weight value of said product and the information indicating the first predicted demand quantity of said product, and the time factor corresponding to each day belonging to said one or more promotion occurrence periods is the length between each day belonging to said promotion occurrence periods and a reference point distinguishing said first time period and said second time period A non-transient computer-readable recording medium, which is a value determined based on.

Images