Method, recording medium, and apparatus of predicting change in demand for goods during holiday

Abstract

This invention provides a technique for predicting changes in demand for goods on rest days. One embodiment of the method is implemented using an electronic device and may include the following steps: based on changes in demand for the target commodity, determining whether there exists a base date within a base interval from a past rest day date where an anomaly in demand occurred, the past rest day date being determined based on rest day information; based on the determination that the base date exists, determining an influence interval including the base date, the influence interval being the interval affected by the rest day effect of the change in demand for the target commodity; and generating arrangement data representing changes in demand within the influence interval.

Description

[Technical Field]

[0001] This invention relates to a technique for predicting changes in demand for goods on rest days. [Previous Technology]

[0002] With the development of communication technology, e-commerce services for online transactions are becoming increasingly popular. The scope of goods that can be the subject of e-commerce services is expanding, no longer limited to goods but also including services. Therefore, since a wide variety of goods are traded through e-commerce services, it is necessary for e-commerce service operators to develop a strategy to effectively sell a wide variety of goods to users.

[0003] As an example of the above strategy, the operator of an e-commerce service needs to predict the demand for goods in order to maintain the inventory of goods at an appropriate level. [Summary of the Invention]

[0004] [Problem to be solved by the invention] The technical problem to be solved by the present invention is to provide a technology that can predict the future demand for goods published in e-commerce services.

[0005] Another technical problem that the present invention aims to solve is to provide a technology for exploring the impact of rest days on the demand for goods published in e-commerce services.

[0006] Another technical problem that the present invention aims to solve is to provide a technology that can quantify the impact of rest days on the demand for goods published in e-commerce services.

[0007] Another technical problem that the present invention aims to solve is to provide a technology that can more accurately predict the future demand for a commodity by modifying the demand for the commodity under the circumstances of rest days.

[0008] The technical issues addressed by this invention are not limited to those mentioned above. Those skilled in the art can readily understand other unmentioned technical issues based on the description in this specification. [Technical Means for Solving the Problem]

[0009] One embodiment of the method of the present invention is implemented by an electronic device and may include the following steps: based on changes in demand for the target product, determining whether there is a base date in the base interval from a past rest day date where an anomaly in demand occurs, the past rest day date being determined based on rest day information; based on the determination that the base date exists, determining an influence interval including the base date, the influence interval being the interval affected by the rest day effect of the change in demand for the target product; and generating arrangement data representing the changes in demand within the influence interval.

[0010] In one embodiment, the above method may further include the following steps: based on the number of dates included in the above influence interval, including future rest days corresponding to the above past rest day dates, determine the prediction interval in which the rest day effect is predicted to occur.

[0011] In one embodiment, the above method may further include the following steps: predicting the future demand for the target product by applying the above arrangement data to the above prediction interval.

[0012] In one embodiment, the above method may further include the following steps: performing pre-processing to eliminate changes in demand caused by the promotion of the target product.

[0013] In one embodiment, the step of determining whether the above-mentioned base date exists may include the following steps: determining whether there is a periodicity of the occurrence of the above-mentioned outlier in the past N years (N is a natural number of 2 or more); and determining whether the above-mentioned base date exists based on the determination that the above-mentioned periodicity exists.

[0014] In one embodiment, the step of determining whether the above-mentioned benchmark date exists may include the following steps: determining whether the above-mentioned benchmark date exists based on the difference between the demand for the above-mentioned target product within the above-mentioned benchmark interval and the average demand for the above-mentioned target product.

[0015] In one embodiment, the step of determining whether the reference date exists based on the difference may include the following steps: determining the reference date based on the determination that there is a date where the difference is above the threshold value.

[0016] In one embodiment, the above-mentioned threshold value may be the same for a group of goods including the above-mentioned object goods.

[0017] In one embodiment, the step of determining the above-mentioned influence interval may include the following steps: identifying the first intersection point and the second intersection point of the first curve of the demand for the above-mentioned object product and the second curve of the demand for the above-mentioned object product; and determining the first intersection point as the starting point of the above-mentioned influence interval and determining the second intersection point as the ending point of the above-mentioned influence interval; and the first curve and the second curve may be curves that use different parameters to smooth the demand for the above-mentioned object product.

[0018] In one embodiment, the first parameter of the first curve may be as follows: the first curve represents the annual demand pattern of the object commodity; the second parameter of the second curve may be as follows: the second curve represents the monthly demand pattern of the object commodity.

[0019] In one embodiment, the step of generating the above arrangement data may include the following steps: determining the baseline demand of the above influence interval by interpolating the first demand corresponding to the start point of the above influence interval and the second demand corresponding to the end point of the above influence interval; and generating the above arrangement data based on the comparison between the demand of the above influence interval and the baseline demand.

[0020] In one embodiment, the step of generating the above arrangement data based on the above comparison may include the following steps: generating the above arrangement data based on the comparison between the above demand and the above benchmark demand according to the dates of each of the above influence intervals.

[0021] In one embodiment, the step of generating the above-mentioned arrangement data may include the following steps: based on the average of multiple arrangement data of the above-mentioned object goods generated for the past N years (N is a natural number of 2 or more), generate representative arrangement data of rest day dates corresponding to the above-mentioned past rest day dates of the above-mentioned object goods.

[0022] In another embodiment of the present invention, a non-transitory computer-readable recording medium records a computer program that can be executed by a processor, wherein the computer program can be configured to cause the processor to execute any of the above methods.

[0023] In another embodiment of the present invention, the electronic device may include: a communication interface configured to communicate with a network; a processor configured to execute a computer program including one or more instructions; and a memory storing the computer program; and the electronic device is configured such that, when the computer program is executed by the processor, the processor executes any of the methods described above. [Effects of the Invention]

[0024] According to the present invention, the future demand for goods published on e-commerce services can be predicted.

[0025] According to the present invention, the impact of rest days on the demand for goods published on e-commerce services can be explored.

[0026] According to the present invention, the impact of rest days on the demand for goods published in e-commerce services can be quantified.

[0027] According to the present invention, the future demand for a commodity can be more accurately predicted by adjusting the demand for the commodity under the circumstances of rest days.

[0028] The effects of the technical concept of the present invention are not limited to the effects mentioned above. Those who are skilled in the art can clearly understand other effects not mentioned based on the description in the specification.

Implementation Method

[0030] The various embodiments described in this invention are illustrated for the purpose of clearly explaining the technical concept of the invention and are not intended to limit them to specific implementation methods. The technical concept of the invention includes various modifications, equivalents, alternatives, and embodiments obtained by selectively combining all or part of the embodiments described in this invention. Furthermore, the scope of the invention's technical concept is not limited to the various embodiments presented below or their specific descriptions.

[0031] The terms used in this invention, including technical or scientific terms, shall have the meanings commonly understood by those with common knowledge in the technical field to which this invention pertains, unless otherwise defined.

[0032] The expressions such as "comprising," "may include," "possibly possess," "may possess," "have," and "may have" used in this invention mean that there are object features (e.g., functions, actions, or constituent elements), and do not exclude the existence of other additional features. That is, the expressions described above should be understood as open-ended terms that have the possibility of including other embodiments.

[0033] The singular expressions used in this invention may include the meaning of the plural form unless otherwise stated in the context, and this also applies to the singular expressions described in the claims of this invention.

[0034] The terms “first,” “second,” “first,” “second,” etc. used in this invention are used to distinguish one object from other objects when referring to a plurality of objects of the same kind, unless otherwise specified in the context, and are not used to limit the order or importance of such objects.

[0035] The expressions “A, B and C”, “A, B or C”, “at least one of A, B and C” or “at least one of A, B or C” used in this invention may refer to 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, or (3) at least one A and at least one B.

[0036] The term "based on" used in this invention is used to describe one or more factors described in the statement or article that affect the determination or judgment of an action or behavior. The term does not exclude other factors that affect the determination or judgment of an action or behavior.

[0037] The expression that a certain constituent element (e.g., the first constituent element) is "connected" or "linked" to another constituent element (e.g., the second constituent element) in this invention may refer not only to the fact that the aforementioned constituent element is directly connected to or linked to the aforementioned other constituent element, but also to the fact that it is connected to or linked to the aforementioned other constituent element through a new other constituent element (e.g., the third constituent element).

[0038] Depending on the context, the expression "configured to" used in this invention may mean "configured in a specific manner," "possessing the capability to perform specific actions," "modified in a specific manner," "manufactured in a specific manner," or "capable of performing specific actions." This expression is not limited to the meaning of "specially designed in hardware." For example, a processor configured to perform a specific action may refer to a general-purpose processor that can perform that specific action by executing software, or a special-purpose computer that is programmed to perform that specific action.

[0039] Hereinafter, various embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings and descriptions thereof, the same or substantially equivalent constituent elements are given the same reference numerals. Furthermore, in the following descriptions of various embodiments, repeated descriptions of the same or corresponding constituent elements may be omitted, but this does not mean that the constituent element is not included in the embodiment.

[0040] FIG1 illustrates an environment 100 in which the apparatus 110, 120 of one embodiment of the present invention can be applied. The environment 100 may include the management device 110 and the user terminal 120.

[0041] On the other hand, FIG1 illustrates one example of a user terminal 120 communicating with a management device 110 via a network, but this is provided only for ease of understanding, and the number of user terminals 120 that can communicate with the management device 110 via a network can of course be changed. That is, one or more user terminals 120 can each connect to the e-commerce services provided by the management device 110. Furthermore, FIG1 only illustrates a preferred embodiment for achieving the purpose of the present invention, and some constituent elements can be added as needed.

[0042] Hereinafter, the constituent elements shown in Figure 1 will be explained in more detail.

[0043] The management device 110 may be a server device that provides e-commerce services. That is, the management device 110 may be a server device that operates under the management of the entity that operates the e-commerce service.

[0044] The management device 110 can manage transactions of goods (e.g., products or services) published in e-commerce services.

[0045] For example, the management device 110 can classify and manage multiple product categories. These product categories can be determined by various criteria; if different types of products are included within the same category, then those different types of products can share the attributes of that category. Each of the multiple products classified in this way can be published in e-commerce services, in other words, on e-commerce-related pages. As another example, the management device 110 can transmit information to the user terminal 120 so that product pages can be displayed on the user terminal 120. As yet another example, the management device 110 can also transmit various responses corresponding to different product-related requests received from the user terminal 120.

[0046] Furthermore, the management device 110 can analyze user logs related to e-commerce services and manage the analysis results. For example, the management device 110 can record and store product demand based on user logs. Based on this product demand, the management device 110 can predict future demand for the product. When predicting this demand, the management device 110 can refer to the impact of rest days on product demand. To avoid repetition, the specific method by which the management device 110 refers to the impact of rest days on product demand will be described below with reference to FIG3.

[0047] In addition to the examples described above, the management device 110 may perform operations related to known technologies that can be performed by server devices providing e-commerce services. Therefore, applying the technical concept of the present invention with reference to operations related to known technologies is not excluded from the scope of the present invention.

[0048] The aforementioned management device 110 may be implemented by one or more computing devices. For example, all functions of the management device 110 may be implemented in a single computing device. As another example, the first function of the management device 110 may be implemented in a first computing device, and the second function may be implemented in a second computing device. For example, the computing device may be a desktop computer, a laptop computer, an application server, a proxy server, or a cloud server, but is not limited to these; any type of device with computing capabilities may be a computing device.

[0049] User terminal 120 may be a terminal for a user utilizing e-commerce services. User terminal 120 may display pages related to e-commerce services provided by management device 110 on its display. Such pages may utilize user interfaces defined as functions associated with e-commerce services. The application of such user interfaces may be managed by the aforementioned management device 110. Furthermore, user terminal 120 may obtain user input on the page from the user and transmit the user input to management device 110. Here, user terminal 120 may obtain a response corresponding to the user input from management device 110, thereby performing an action defined as corresponding to the user input. As described above, in order for a user to utilize e-commerce services through user terminal 120, a web browser or application may be installed in user terminal 120.

[0050] The user terminal 120 described above may be any of the following devices: desktop computer, laptop computer, tablet computer, wearable device or smartphone, but is not limited thereto. All types of devices with computing functions may be user terminal 120.

[0051] The management device 110 and user terminal 120 shown in Figure 1 can communicate via a network. This network can be implemented by any type of wired or wireless network, such as a Local Area Network (LAN), a Wide Area Network (WAN), a Mobile Radio Communication Network (MRCN), or WiBro (Wireless Broadband).

[0052] FIG2 illustrates a computing device 200 that can implement the apparatus 110, 120 of one embodiment of the present invention. That is, the management device 110 or user terminal 120 shown in FIG1 can be implemented by the computing device 200 shown in FIG2. For reference, in the present invention, computing device 200 can be used interchangeably with electronic device.

[0053] The computing device 200 may include one or more processors 210, one or more memory units 220, or communication interfaces 230. In one embodiment, some components of the computing device 200 may be deleted, or other components (e.g., a display or input device) may be added to the computing device 200. Furthermore, some components may be additionally or alternatively integrated, or implemented as a single or multiple entities. In this invention, "one or more processors 210" may refer to processor 210. Unless otherwise explicitly stated herein, the term "processor 210" may refer to a collection of one or more processors. Similarly, in this invention, "one or more memory units 220" may refer to memory units 220. Unless otherwise explicitly stated herein, the term "memory units 220" may refer to a collection of one or more memory units.

[0054] The constituent elements shown in Figure 2 will be explained in more detail below.

[0055] The processor 210 can perform control of the various components of the computing device 200 or computational or information processing related to communication. Specifically, the processor 210 can drive software (or computer programs) received from other components to control at least one component of the computing device 200 connected to the processor 210. As an example, the processor 210 can load commands (e.g., instructions, codes, or code segments) or information into the memory 220, process the commands or information stored in the memory 220, and store the result information obtained from the processing in the memory 220. Furthermore, the processor 210 can be operatively connected to the components of the computing device 200 to perform various computational, processing, generation, or manipulation actions related to the present invention.

[0056] Memory 220 can store various types of information. The information stored in memory 220, as information obtained, processed, or used by at least one component of computing device 200, may include software. The software may include one or more commands that, when loaded into memory 220, cause processor 210 to perform actions according to various embodiments of the present invention. That is, processor 210 can perform actions according to various embodiments of the present invention by executing one or more of the aforementioned commands. Memory 220 may include, for example, volatile or non-volatile memory. In one embodiment, the program, as software stored in memory 220, may include an operating system, application program, or middleware for controlling the resources of computing device 200, the middleware providing various functions to the application program so that the application program can utilize the resources of computing device 200.

[0057] The communication interface 230 can establish wired or wireless communication channels with other devices to send and receive various information. In one embodiment, the communication interface 230 may include at least one port for connecting to other devices via a wired cable in order to conduct wired communication with other devices. In this case, the communication interface 230 can communicate with other wired connected devices via at least one port. In one embodiment, the communication interface 230 may be configured to include a cellular communication module and connect to a cellular network (e.g., 3G, LTE, 5G, Wibro, or WiMAX) to send and receive information with other devices. In one embodiment, the communication interface 230 may include a short-range communication module to use short-range communication (e.g., Wi-Fi, Bluetooth, Bluetooth Low Energy (BLE), UWB) to send and receive information with external devices. In one embodiment, the communication interface 230 may include a contactless communication module for performing contactless communication. Contactless communication may include, for example, at least one of the following short-range contactless communication technologies: NFC (Near Field Communication), RFID (Radio Frequency Identification), or MST (Magnetic Secure Transmission). In addition to the various examples described above, the computing device 200 can also be implemented in various known ways for communicating with other devices; the scope of the invention is not limited to the examples described above.

[0058] In one embodiment, the computing device 200 may include a display. The display may display various screens (e.g., more than one page) based on the control of the processor 210. To display screens with various interfaces on the display, a web browser or dedicated application may be installed in the computing device 200, for example. Furthermore, the display, as a component capable of interacting with the user, may receive user input from the user. Such a display may be implemented in the form of a touch sensor panel (TSP) that can recognize the contact or proximity of various external objects (e.g., the user's finger or a pen).

[0059] In one embodiment, the computing device 200 may include an input device (e.g., a mouse or keyboard). The input device may receive information from outside the computing device 200 (e.g., a user) that will be used in the components of the computing device 200.

[0060] The processor 210, memory 220 and communication interface 230 shown in Figure 2 can be connected to each other via bus, GPIO (General Purpose Input / Output), SPI (Serial Peripheral Interface) or MIPI (Mobile Industry Processor Interface) to send and receive information or signals.

[0061] Hereinafter, the methods of various embodiments of the present invention will be described in detail. It should be noted that the actions are shown in a specific order in the following figures, but it is not necessary to perform the actions only in the specific order shown or sequentially, or to perform all the shown actions in order to obtain the desired result.

[0062] Furthermore, the actions of the method described with reference to the following figures can be performed by a computing device. In other words, the actions of the method can be implemented by one or more instructions executed by the processor of the computing device. All actions included in this method can be performed by a single physical computing device, but the first action of the method can be performed by a first computing device, and the second action of the method can be performed by a second computing device.

[0063] Hereinafter, it will be assumed that the operation of the above method is performed by the management device 110 shown in FIG1 and the description will continue. For ease of explanation, the main body of the operation included in the method may be omitted, but unless otherwise stated in the context, it can be interpreted as the operation being performed by the management device 110.

[0064] Figure 3 is a sequence diagram illustrating method S300 of one embodiment of the present invention. Method S300 shown in Figure 3 includes a series of actions: exploring past intervals affected by the rest day effect that causes changes in demand for goods, and extracting data representing changes in demand during those intervals. The detailed actions in Figure 3 will be described below.

[0065] Based on the changes in demand for the target product, it can be determined whether there is a base date in the base interval from the date of the past rest day where an outlier in demand occurred, and the date of the past rest day is determined based on rest day information (S310).

[0066] The object of the assessment is the product that becomes the object of the determination of demand changes, and may be any of the products published in the e-commerce service. Depending on the attributes of the product, the object of the assessment may be included in any of a plurality of categories. For example, if the object of the assessment is "vegetables", it may be included in the "fresh food category", and if the object of the assessment is "shirts", it may be included in the "clothing category". A collection of more than one product included in a category that includes the object of the assessment may be referred to as a product group.

[0067] The demand for the target product can be the number of times that multiple users purchase the target product using e-commerce services.

[0068] Changes in demand for a commodity can refer to an increase or decrease in demand when compared with the average demand for the commodity. The average demand for the commodity can be, for example, the average demand on a weekly basis, the average demand on a monthly basis, or the average demand on a yearly basis.

[0069] Rest day information may be information on rest days that can affect changes in demand for the target product. Rest day information may include, for example, statutory public holidays or temporary public holidays based on the Gregorian calendar, or statutory public holidays or temporary public holidays based on the lunar calendar. Furthermore, rest day information may also include dates specified by the operator of the e-commerce service.

[0070] Past rest days can be past rest days included in rest day information that ensures demand for the target product (e.g., the number of times the target product is purchased). For example, using 2024 as a base year, "December 25, 2023 (Christmas Day)" can be a past rest day date. In other words, any rest day in the year preceding the base year can be a past rest day date. However, the scope of this invention application is not limited to this example; if it is a past rest day that ensures demand for the target product, then any rest day can be a past rest day date.

[0071] The reference interval may be an interval determined based on past rest day dates. More specifically, the reference interval may be an interval with a specific date width before and after the past rest day date. For example, an interval starting from "X days (X is 0 or a natural number) before" the past rest day date and ending from "Y days (Y is 0 or a natural number) after" may be a reference interval. The width of the reference interval determined based on the past rest day date may be adjusted according to the service operation of the e-commerce service operator. If it is a width that can capture changes in demand for the target product from the past rest day date, then any width may be included within the scope of the invention application patent of this invention.

[0072] An outlier in the demand for the target product may be one where there is a significant change in demand compared to the average demand for the target product. The determination of specific outliers will be described below with reference to various embodiments related to this action S310.

[0073] The base date is one or more dates within the base range, and may be the date on which an outlier in the demand occurs.

[0074] Regarding outliers in demand, this action S310 may include the following action: determining whether a base date exists based on the difference between the demand for the target product within the base period and the average demand for the target product (e.g., average demand per year). In other words, the existence of a base date can be determined by determining the difference between the average demand for the target product and the demand for the target product within the base period. This difference determination action can be performed using the date units included in the base period.

[0075] Regarding the determination of the difference, in one embodiment, the demand for the target product within the baseline interval is not the actual demand, and can be replaced by the value obtained by applying a smoothing function (e.g., a Gaussian filter) to the demand curve within the baseline interval. In this case, factors that can temporarily affect the demand for the target product, such as the product being sold out, can be eliminated, in addition to changes in demand caused by rest days. The parameters of the smoothing function that can be applied to the demand curve can be adjusted according to the service operation of the e-commerce service operator. Preferably, parameters can be set that do not distort changes in demand for the target product caused by rest days.

[0076] Regarding the determination of the difference, in another embodiment, the base date can be determined based on the date on which the difference between the demand for the target product and the average demand for the target product within the base interval is determined to be above a critical value. Here, the critical value can be a value uniquely determined for the target product, or a value uniquely determined for a group of products including the target product (i.e., the same critical value is applied to the entire group of products), or a value uniformly determined for all products published in the e-commerce service. For example, when the difference between the demand for the target product and the average demand is "100", it can be determined that there is an outlier in the demand, and the date on which the outlier in the demand occurs can be determined as the base date.

[0077] Regarding the determination of the difference, in another embodiment, when the date in the distribution of the difference between the demand for the target product and the average demand for the target product within a specific period shows that the date of the difference in the upper 95% or more of the interval is included in the benchmark interval, the date can be determined as the benchmark date.

[0078] Regarding outliers in demand, this action S310 may include the following actions: determining whether there is a periodicity in the occurrence of outliers over the past N years (N is a natural number greater than 2, for example, N is 3); and determining whether a base date exists based on the determination that a periodicity exists. In this embodiment, it is determined whether outliers occur periodically near the rest day dates corresponding to past rest day dates, and this is used to determine whether a base date exists. For example, using 2024 as the base date, even if an outlier is found near "December 25, 2023", the date on which the outlier occurred will not be immediately determined as the base date. It can be determined whether outliers have been found near "December 25, 2022" and "December 25, 2021", in other words, whether there is a periodicity in the occurrence of outliers. If a periodicity of outliers is determined, the date on which an outlier occurred near "December 25, 2023" can be used as the base date. If a periodicity of outliers is determined, the date on which an outlier occurred near "December 25, 2023" cannot be used as the base date. When outliers occur without periodicity within a specific year, it may not be suitable to use this as a base date for predicting changes in demand due to rest days. According to this embodiment, a suitable base date for predicting changes in demand due to rest days can be determined by determining the periodicity of outlier occurrences.

[0079] Based on the determination that a base date exists, the influence range including the base date can be determined (S320).

[0080] The affected interval includes the base date and can be the interval affected by the rest day effect of changes in demand for the target product.

[0081] This action S320 may include the following actions: identifying the first intersection point and the second intersection point where the first curve of the demand for the target product intersects with the second curve of the demand for the target product; and determining the first intersection point as the starting point of the influence interval and the second intersection point as the ending point of the influence interval.

[0082] Regarding the actions of determining the start and end points, in one embodiment, the first curve and the second curve may each be a curve for applying a smoothing function (e.g., a Gaussian filter) to the demand curve of the target product. Here, the first curve and the second curve may be curves that use different parameters to smooth the demand for the target product. For example, when using a Gaussian filter as the smoothing function, the degree of smoothing can be adjusted by adjusting the sigma (i.e., parameter) of the Gaussian filter. Specifically, the larger the sigma, the higher the degree of smoothing; the smaller the sigma, the lower the degree of smoothing.

[0083] Here, the first parameter of the first curve can make the first curve represent the annual demand pattern of the target product, and the second parameter of the second curve can make the second curve represent the monthly demand pattern of the target product. That is, the first parameter can be a value greater than the second parameter. However, unlike the example above, the adjustment of the parameters as described above can be adjusted according to the service operation of the e-commerce service operator. As long as the first parameter and the second parameter are set in a way that the influence range determined by the intersection of the first curve and the second curve includes the base date, the setting of any parameter can be included within the scope of the invention application patent of this invention.

[0084] It can generate arrangement data representing changes in demand within the affected interval (S330).

[0085] Permutation data can be data representing changes in demand within an affected period. Such permutation data can represent changes in demand for each date within the affected period. For example, in the case where the affected period is from December 23, 2023 to December 27, 2023, the set of values ​​for the dates December 23, 2023, December 24, 2023, December 25, 2023, December 26, 2023, and December 27, 2023 can be permutation data. Here, permutation data can be the set of values ​​for each date included in the affected period of the previous year.

[0086] This action S330 may include the following actions: determining the baseline demand of the affected interval by interpolating the first demand corresponding to the start point of the affected interval and the second demand corresponding to the end point of the affected interval; and generating arrangement data based on the comparison between the demand of the affected interval and the baseline demand.

[0087] Interpolation can be performed as follows: a linear connection is made between the first demand corresponding to the beginning of the affected interval and the second demand corresponding to the end. The baseline demand for the affected interval can be determined based on the line formed by this interpolation. Here, since the baseline demand is based on a line connecting the demands before and after the affected interval where the rest day effect occurs, it can be a demand that can be compared with the demands of the affected interval where the rest day effect occurs. This baseline demand can be determined by the date units included in the affected interval.

[0088] The comparison between the demand in the affected interval and the baseline demand can be performed as follows: calculate their equal ratio. For example, if the demand in the affected interval for a specific date is "150" and the baseline demand is "100", then the arrangement data for that specific date can be generated as "1.5". As described above, the arrangement data can be generated as a set of date units included in the affected interval.

[0089] Furthermore, the comparison between the demand in the affected interval and the baseline demand can be performed by calculating the slope of their differences. To calculate the slope, the difference in their differences can also be calculated. As described above, the arrangement of data can generate a set of date units included in the affected interval.

[0090] Furthermore, this action S330 may include the following action: generating representative arrangement data of rest day dates corresponding to the past rest day dates of the object product based on the average of multiple arrangement data of the object product generated for the past N years (N is a natural number of 2 or more).

[0091] According to this embodiment, arrangement data for each year of N years corresponding to a rest day (e.g., Christmas) can be generated, and representative arrangement data corresponding to that rest day can be generated based on their average. The generation of arrangement data for each year can be referred to the description of the above-described operations S310 to S330. Here, the representative arrangement data can be generated as a set of date units included in the influence interval.

[0092] According to the method S300 described so far with reference to FIG3, the change in demand for the target product within the influence interval corresponding to the rest day can be quantitatively calculated. The operator of the e-commerce service can predict future demand based on the change in demand on past rest days. This will be explained below with reference to FIG4.

[0093] Figure 4 is a sequence diagram illustrating method S400 of one embodiment of the present invention. Method S400 shown in Figure 4 may include a series of actions such as applying arrangement data calculated based on past demand to future demand forecasting. The detailed actions in Figure 4 will be described below.

[0094] Based on the number of dates included in the influence interval, which may include future rest days corresponding to past rest days, the prediction interval for which the rest day effect is predicted to occur is determined (S410).

[0095] In this action S410, the Gregorian calendar is used as the basis. For example, if the past influence period is determined to be from December 23, 2023 to December 27, 2023 (including Christmas (December 25)), the prediction period can be determined to be from December 23, 2024 to December 27, 2024. Using the lunar calendar as the basis, as another example, if the past influence period is determined to be from September 27, 2023 to September 30, 2023 (including the Mid-Autumn Festival on September 29, 2023), the prediction period including the Mid-Autumn Festival on September 17, 2024 can be determined to be from September 15, 2024 to September 18, 2024. That is, the distance between the start and end points of the influence period is calculated based on past rest days, and this distance can be applied to future rest days.

[0096] The future demand for the target product can be predicted by applying the permutation data to the forecast interval (S420).

[0097] The application of permutation data can be as follows: applying ratios or slopes to adjust demand. For example, applying ratios or slopes to the initial forecasts of future demand to calculate the adjusted value can be considered an application of permutation data.

[0098] Demand forecasting refers to predicting the future demand for a particular commodity, and may refer to the future demand for that commodity. This demand forecasting can be carried out using the various date units included in the forecasting period. The method of demand forecasting can be basically based on the past demand for the commodity. As the simplest method, demand forecasting can be carried out based on the past average demand for the commodity. The effect of rest days due to rest days can be used to forecast the demand for the commodity by applying arrangement data to this average demand. In addition to the method based on the above average demand, future demand can also be predicted in various ways by referring to known techniques related to demand forecasting. Any technique that corrects the effect of rest days by applying arrangement data to a known technique for predicting future demand based on past demand is included within the scope of the invention application of this invention.

[0099] According to the method described so far with reference to FIG4 (S400), the arrangement data calculated based on past rest day information can be used to predict future demand. The accuracy of demand forecasting can be improved by applying arrangement data that reflects the effect of rest days.

[0100] Figure 5 is a sequence diagram illustrating method S500 according to one embodiment of the present invention. Method S500 shown in Figure 5 may include the following action: preprocessing the requirements that form the basis of method S300 shown in Figure 3. The detailed actions of Figure 5 will be described below.

[0101] It is permissible to perform pre-processing to eliminate changes in demand caused by promotion of the target product (S510).

[0102] Promotion can refer to various offers provided by the operator of e-commerce services to users for target products. For example, promotion may include discounts, free shipping, and free gifts for target products. Since the promotions mentioned above are factors that can temporarily change the demand for target products, the method S500 shown in Figure 5 can be implemented before implementing the method S300 shown in Figure 3.

[0103] This action S510 may include the following actions: calculating the increase in demand caused by the promotion; and performing pre-processing to eliminate the change in demand caused by the promotion based on the calculated result.

[0104] The increase in demand caused by promotion can be calculated by comparing the average demand during periods without promotion with the average demand during periods with promotion.

[0105] Based on the method S400 described so far with reference to FIG5, as a prerequisite for more accurately exploring the effect of rest days, the change in demand caused by promotion can be effectively eliminated.

[0106] Figure 6 is a diagram used to explain the action S320 of determining the influence range as described with reference to Figure 3.

[0107] Figure 6 shows a curve 600 illustrating the demand curve 611 for the target product in terms of date units. Here, the first curve 612 can be a curve that uses the first parameter to smooth the demand curve 611, and the second curve 613 can be a curve that uses the second parameter to smooth the demand curve 611. In this case, the first parameter can be a value greater than the second parameter.

[0108] The starting point 632 and the ending point 633 of the influence interval 631 can be determined by the intersection of the first curve 612 and the second curve 613. Here, the influence interval 631 can be an interval that includes the base date 620.

[0109] Figure 7 is a diagram used to explain the action S330 of generating arrangement data as described with reference to Figure 3.

[0110] Figure 7 shows the demand curve 722 for the target commodity.

[0111] The influence interval 711 shown in Figure 7 may have a starting point 712 and an ending point 713.

[0112] The interpolated line 721 can be a line that linearly connects the demand at the starting point 712 and the demand at the ending point 713. A benchmark demand can be determined based on the interpolated line 721 to compare with the demand in the affected interval 711.

[0113] In the sequence diagram of the present invention, the actions of the method or algorithm are described sequentially. However, in addition to performing the actions sequentially, the actions may also be performed in any order that can be combined. The description of the sequence diagram of the present invention does not preclude changes or modifications to the method or algorithm, and does not imply that any action is necessary or preferred. In one embodiment, at least some actions may be performed in parallel, repeatedly, or heuristically. In another embodiment, at least some actions may be omitted, or other actions may be added.

[0114] Various embodiments of the present invention can be implemented in software form in a machine-readable storage medium (MRSM). The software can be any software used to implement the various embodiments of the present invention. Programmers in the art to which this invention pertains can deduce the software based on the various embodiments of the present invention. For example, the software can be a program including machine-readable commands. A computing device, as a means of operating according to commands invoked from the storage medium, can be referred to interchangeably with an electronic device. In one embodiment, the processor of the computing device executes the invoked command, thereby enabling the components of the computing device to perform functions corresponding to the command. Storage medium can refer to all kinds of recording media capable of being read by a machine and storing information. Storage media may include, for example, ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical information storage device, etc. In one embodiment, the storage medium can also be implemented in the form of a computer system connected to a network. In this case, the software can be executed by being distributed across the computer system. In another embodiment, the storage medium may be a non-transitory storage medium. A non-transitory storage medium refers to a medium that actually exists and is not related to the semi-permanent or temporary storage of information, excluding transiently transmitted signals.

[0115] The technical concept of the present invention has been described above according to various embodiments, but the technical concept of the present invention includes various substitutions, variations, and modifications that can be understood by those skilled in the art within which the present invention pertains. Furthermore, it should be understood that such substitutions, variations, and modifications may be included within the scope of the appended patent applications. [Simplified Explanation of the Diagram]

[0029] Figure 1 illustrates an environment in which an embodiment of the present invention can be applied. Figure 2 illustrates a computing device in which an embodiment of the present invention can be implemented. Figure 3 illustrates a sequence diagram of a method according to an embodiment of the present invention. Figure 4 illustrates a sequence diagram of a method according to an embodiment of the present invention. Figure 5 illustrates a sequence diagram of a method according to an embodiment of the present invention. Figure 6 illustrates an explanation of the operation of determining the influence range as described with reference to Figure 3. Figure 7 illustrates an explanation of the operation of generating arrangement data as described with reference to Figure 3.

Claims

1. A method for predicting changes in demand for goods on rest days, implemented by an electronic device, comprising the following steps: Based on changes in demand for a target commodity, determining whether there exists a benchmark date within a benchmark interval from a past rest day date where an outlier in demand occurs, the past rest day date being determined based on rest day information; Based on the determination that the benchmark date exists, determining an influence interval including the benchmark date, the influence interval being the interval affected by the rest day effect of the change in demand for the target commodity; and generating arrangement data representing changes in demand within the influence interval, wherein the step of determining the influence interval comprises the following steps: Identifying a first intersection point and a second intersection point of a first curve generated by applying a smoothing function based on a first parameter to the demand curve of the target commodity and a second curve generated by applying a smoothing function based on a second parameter to the demand curve of the target commodity; and determining the first intersection point as the starting point of the influence interval and the second intersection point as the ending point of the influence interval; wherein the value of the first parameter is different from the value of the second parameter. The first intersection point is the point where the first curve and the second curve intersect before the reference date, and the second intersection point is the point where the first curve and the second curve intersect after the reference date.

2. The method of Request 1 further includes the following steps: Based on the number of dates included in the above-mentioned influence interval, including future rest days corresponding to the above-mentioned past rest day dates, determine the prediction interval for the occurrence of the above-mentioned rest day effect.

3. The method of claim 2 further includes the following steps: predicting the future demand for the object product by applying the above-mentioned arrangement data to the above-mentioned forecast interval.

4. The method of request item 1 further includes the following steps: performing pre-processing to eliminate changes in demand caused by the promotion of the aforementioned target product.

5. The method of claim 1, wherein the step of determining whether the aforementioned base date exists includes the following steps: determining whether there is a periodicity of occurrence of the aforementioned outlier in the past N years (N is a natural number of 2 or more); and determining whether the aforementioned base date exists based on the determination that the aforementioned periodicity exists.

6. The method of Request 1, wherein the step of determining whether the aforementioned benchmark date exists includes the following steps: determining whether the aforementioned benchmark date exists based on the difference between the demand for the aforementioned object product within the aforementioned benchmark interval and the average demand for the aforementioned object product.

7. The method of claim 6, wherein the step of determining whether the aforementioned base date exists based on the aforementioned difference includes the following steps: determining the aforementioned base date based on the determination that a date exists where the aforementioned difference is above a critical value.

8. The method of claim 7, wherein the above-mentioned threshold value is the same for the product group including the above-mentioned object product.

9. As in Request 1, wherein the first parameter of the first curve is as follows: so that the first curve represents the annual demand pattern of the object commodity; and the second parameter of the second curve is as follows: so that the second curve represents the monthly demand pattern of the object commodity.

10. The method of claim 1, wherein the step of generating the above-mentioned arrangement data includes the following steps: determining the baseline demand of the above-mentioned influence interval by interpolating the first demand corresponding to the beginning of the above-mentioned influence interval and the second demand corresponding to the end of the above-mentioned influence interval; and generating the above-mentioned arrangement data based on the comparison between the demand of the above-mentioned influence interval and the baseline demand.

11. The method of request item 10, wherein the step of generating the above-mentioned arrangement data based on the above comparison includes the following steps: generating the above-mentioned arrangement data based on the comparison between the above-mentioned demand and the above-mentioned baseline demand according to the dates of each of the above-mentioned influence intervals.

12. The method of claim 1, wherein the step of generating the aforementioned arrangement data includes the following steps: generating the aforementioned arrangement data of the target goods for each of the past N years (N being a natural number of 2 or more); and wherein the method further includes: Based on the average of the above-mentioned arrangement data for each of the past N years, representative arrangement data of rest days corresponding to the above-mentioned past rest days for the above-mentioned target products is generated.

13. A non-transitory computer-readable recording medium that records a computer program for execution by a processor, wherein the computer program is configured to cause the processor to perform any one of claims 1 to 12.

14. An electronic device comprising: The communication interface is configured to communicate with a network; the processor is configured to execute a computer program including one or more instructions; and the memory stores the computer program; and the electronic device is configured such that when the computer program is executed by the processor, the processor performs the method as described in any one of claims 1 to 12.

Images