KR20230016868A - Method for providing product discount coupons based on store demand forecasting and system thereof - Google Patents

Abstract

An operating method of a product discount providing system based on store demand forecasting comprises the steps of: obtaining existing sales data of a store including reservation history information and visit history information of customers by data of the store from a database; predicting demand for the store using the existing sales data of the store; determining the available number of product discount coupons based on the demand forecast result of the store; providing the product discount coupon to a user terminal of a customer; receiving request information for using the product discount coupon from the user terminal; and transmitting the request information for using the product discount coupon to the store. The present invention enables the store to predict the customer demand in advance.

Description

Method and system for providing product discount vouchers based on store demand forecast {METHOD FOR PROVIDING PRODUCT DISCOUNT COUPONS BASED ON STORE DEMAND FORECASTING AND SYSTEM THEREOF}

The present invention relates to a method for providing a product discount voucher, and more particularly, to a method and system for providing a product discount voucher based on predicted store demand based on store visit history and sales history of existing customers using an artificial neural network It is about.

Out of 660,000 restaurants, small restaurants with fewer than five employees account for 569,000, or 86.5%. Therefore, restaurants in the restaurant business, where most restaurants are self-employed, are vulnerable to external factors such as fierce competition and rapidly changing environments because they do not have a systematic system for store management.

Accordingly, stores such as restaurants may perform marketing by providing product discount coupons to consumers. Here, stores generally predict demand and use discount coupons only for unsatisfied demand. However, even if the discount coupon is provided through the above method, the operating profit rate is not high.

Accordingly, there is a need for a method of efficiently utilizing product discount coupons issued based on demand forecast information as well as providing scientific customer analysis and sales and demand forecast information. The present invention relates to this.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a method and system for providing a product discount voucher activated through a consumer's reservation in issuing a product discount voucher based on the predicted store demand. is to do

The technical problems to be achieved through the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned are clearly understood by those skilled in the art from the description below. It could be.

According to an embodiment of the present invention, a method of operating a product discount providing system based on store demand prediction obtains existing sales data of a store, including reservation history information and visit history information of customers by date of the store, from a database. doing; predicting demand for the store using existing sales data of the store; determining a usable quantity of the product discount vouchers based on a demand forecast result of the store; providing the product discount coupon to a consumer's user terminal; Receiving the product discount voucher use request information from the user terminal; and transmitting information requesting use of the product discount voucher to the store.

Here, the step of predicting the demand of the store using the existing sales data of the store may include preprocessing the existing sales data of the store; generating a plurality of store demand prediction models by using a plurality of preset algorithms learned from the preprocessed store's existing sales data; assigning a hyperparameter of each of the generated store demand prediction models to each of the plurality of preset algorithms; and determining a store demand prediction model based on evaluation results of each of the plurality of store demand prediction models.

Here, the step of predicting the demand of the store may include: predicting the demand of the store in the first time period; and predicting store demand in a second time period including the first time period.

Here, the operation method further includes determining the quantity of product discount coupons provided by the store based on the demand prediction result of the store, wherein the quantity of product discount coupons provided by the store corresponds to the demand of the store in the second time period. basis, and the usable quantity of the product discount coupon may be determined based on the demand of the store in the first time period.

Here, the operation method may include obtaining store visit record information of customers based on existing sales data of the store; obtaining sales information of the store generated by the customers; and analyzing store loyalty of each of the customers based on store visit record information of the customers and sales information of the store, wherein the consumer is filtered based on loyalty to the store among the customers. It may be characterized as being at least one or more customers.

Here, the receiving of the product discount voucher use request information may include transmitting information on the usable quantity of the product discount voucher in the store indicated by the product discount voucher use request information to the user terminal; Receiving reservation request information of the store using the product discount coupon from the user terminal; and activating the product discount voucher based on the reservation request information of the store.

A server of a product discount providing system based on store demand prediction according to an embodiment of the present invention includes a memory in which at least one program command is stored; a processor to execute the at least one program instruction; and a transmitting/receiving device connected to a network to perform communication, wherein the at least one program command obtains, from a database, existing sales data of the store, including reservation history information and visit history information of customers of the store by date; predicting the demand of the store using the existing sales data of the store; determining the number of product discount coupons provided by the store based on the demand prediction result of the store; determining a usable quantity of the merchandise discount vouchers based on a demand prediction result of the store; providing the product discount coupon to a user terminal of the consumer; receiving a request for using the product discount voucher from the user terminal; receiving the product discount voucher use request information from the user terminal; And it may be characterized in that it is executed to transmit the use request information of the product discount coupon to the store.

Here, the at least one program command, in executing to predict the demand of the store using the existing sales data of the store, pre-processes the existing sales data of the store; generating a plurality of store demand prediction models using a plurality of preset algorithms learned from the preprocessed store's existing sales data; assigning a hyperparameter of each of the generated store demand prediction models to each of the plurality of preset algorithms; And, it may be further executed to determine a store demand prediction model based on the evaluation result of each of the plurality of store demand prediction models.

Here, the at least one program command, when executed to predict the demand of the store, predicts the demand of the store in the first time period; And it may be further executed to predict the demand of the store in the second time period including the first time period.

Here, the at least one program command is further executed to determine the quantity of product discount coupons provided by the store based on the demand prediction result of the store, wherein the quantity of product discount coupons provided by the store is determined by the demand of the store in the second time period. , and the usable quantity of the product discount coupon may be determined based on the demand of the store in the first time period.

Here, the at least one program instruction obtains store visit record information of customers based on existing sales data of the store; obtaining sales information of the store generated by the customers; And, based on the store visit record information of the customers and the sales information of the store, it is further executed to analyze the loyalty of each of the customers to the store, and the consumer is filtered based on the loyalty to the store, among the customers. It may be characterized as being at least one customer.

Here, the at least one command is executed to receive the product discount voucher use request information, and transmits available quantity information of the product discount voucher in the store indicated by the product discount voucher use request information to the user terminal. do; Receiving reservation request information of the store using the product discount coupon from the user terminal; And it may be executed to activate the product discount coupon based on the reservation request information of the store.

The above-described aspects of the present disclosure are only some of the preferred embodiments of the present disclosure, and various embodiments in which the technical features of the present disclosure are reflected are detailed descriptions of the present disclosure to be detailed below by those of ordinary skill in the art. It can be derived and understood based on.

According to an embodiment of the present invention, the product discount coupon providing system provides product discount coupons with relaxed restrictions on usage conditions to consumers by providing product discount coupons based on the demand for stores predicted based on the characteristics of each date and the reservation details for each date. can provide

According to an embodiment of the present invention, the product discount coupon providing system can provide an effect of optimizing the store's profit while predicting customer demand in advance by activating the product discount coupon when making a store reservation.

The effects obtainable through the present invention are not limited to the mentioned effects, and other effects not mentioned will be clearly understood by those skilled in the art from the description below. .

1 is a conceptual diagram illustrating an embodiment of the configuration of a product discount coupon providing system.
2 is a block diagram showing an embodiment of the configuration of a device constituting a product discount coupon providing system.
3 is a conceptual diagram illustrating an embodiment of an artificial neural network included in a device of a product discount coupon providing system.
4 is a flowchart illustrating an embodiment of a product discount voucher providing method of a product discount voucher providing system.
5 is a conceptual diagram illustrating an embodiment of an artificial intelligence architecture of a product discount coupon providing system.
6 is a flowchart illustrating an embodiment of a store demand prediction method of a product discount coupon providing system.
7 is a flowchart illustrating an embodiment of a method for generating a demand prediction model using model establishment data and test data of a product discount coupon providing system.
8 is a flowchart illustrating an embodiment of a method of generating a store demand prediction model of a product discount coupon providing system.
9 is a flowchart illustrating an embodiment of a store demand prediction model evaluation method of a product discount coupon providing system.
10 is a flowchart illustrating an embodiment of a customer loyalty evaluation method of a product discount coupon providing system.
11 is a flowchart illustrating an embodiment of a product discount voucher activating method of a product discount voucher providing system.
12 is a flowchart illustrating an embodiment of providing and using product discount coupons between a product discount coupon providing system, a store, and a user terminal of a consumer.
13 is a conceptual diagram illustrating an embodiment of a product discount coupon providing system, a flow of money between a store and a consumer.

Since the present invention can make various changes and have various embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

Terms such as first and second may be used to describe various components, but the components should not be limited by the terms. These terms are only used for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The terms and/or include any combination of a plurality of related recited items or any of a plurality of related recited items.

It is understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, but other elements may exist in the middle. It should be. On the other hand, when an element is referred to as “directly connected” or “directly connected” to another element, it should be understood that no other element exists in the middle.

Terms used in this application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, the terms "include" or "have" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in this application, they should not be interpreted in an ideal or excessively formal meaning. don't

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present invention will be described in more detail. In order to facilitate overall understanding in the description of the present invention, the same reference numerals are used for the same components in the drawings, and redundant descriptions of the same components are omitted.

A product discount voucher providing system to which embodiments according to the present invention are applied will be described. The product discount voucher providing system to which embodiments according to the present invention are applied is not limited to the description below, and embodiments according to the present invention can be applied to various communication systems.

1 is a conceptual diagram illustrating an embodiment of the configuration of a product discount coupon providing system.

Referring to FIG. 1, the product discount voucher providing system is a store management terminal connected to the server 110 through a network 120 and a server 110 of the product discount voucher providing system for providing product discount vouchers based on store information ( 130) and the consumer's user terminal 140.

The server 110 manages customer-related information such as store visit history information and store sales history information of customers of each store, predicts store demand, and provides information necessary to provide services for providing product discount coupons. can be saved The server 110 may store in advance information including an algorithm for generating a demand prediction model for a store of customers and a store demand prediction model.

The network 120 may refer to a connection structure for exchanging information between the server 110 and the store management terminal 130 . The network 120 includes the Internet, a local area network (LAN), a wireless local area network (Wireless Local Area Network), a wide area network (WAN), a personal area network (PAN), 3G, 4G, and long term evolution (LTE). , VoLTE (voice over LTE), 5G NR (new radio) Wi-Fi (wireless-fidelity), Bluetooth, NFC, RFID (radio frequency identification) home network, IoT (internet of things), etc. can include

The store management terminal 130 may be connected to the server 110 through the network 120 . Also, the store management terminal 130 may include user equipment such as a computer, tablet, or smart phone. The store management terminal 130 may receive information provided through a service for customer management from the server 110 . The information received from the server 110 may include store demand prediction result data and the like.

The store management terminal 130 may display the received store demand prediction result data through an output device. Also, the store management terminal 130 may provide feedback on the service provided from the server 110 based on the information input from the user.

The user terminal 140 of the consumer may be connected to the server 110 through the network 120 . In addition, the user terminal 140 of the consumer may include user equipment such as a computer, tablet, or smart phone. The consumer's user terminal 140 may receive information provided through a service for purchasing and using product discount vouchers from the server 110 . Information related to product discount coupons from the server 110 may be included.

The consumer's user terminal 140 may display the received product discount coupon-related information through an output device. Further, the consumer's user terminal 140 may provide feedback on the service provided from the server 110 based on the information input from the user.

The server 110 of FIG. 1 , the store management terminal 130 and the consumer user terminal 140 may be referred to as a device, and the configuration of the device may be as described below.

2 is a block diagram showing an embodiment of the configuration of a device constituting a product discount coupon providing system.

The device 200 included in the system may include at least one processor 210 and a memory 220 storing instructions instructing the at least one processor 210 to perform at least one step. can

Here, the at least one processor 210 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor for performing methods according to embodiments of the present invention. can Each of the memory 220 and the storage device 260 may include at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory 220 may include at least one of a read only memory (ROM) and a random access memory (RAM).

Also, the device 200 included in the system may include a transceiver 230 that performs communication through a wireless network. In addition, the device 200 included in the system may further include an input interface device 240 , an output interface device 250 , a storage device 260 , and the like. Each component included in the device 200 may be connected by a bus 270 to communicate with each other.

In addition, the device 200 included in the system may further include hardware components or software components for implementing/driving an artificial neural network. For example, the hardware components may include a neural processor unit (NPU) and the like. For example, software components include a framework, a kernel or device driver, middleware, an application programming interface (API), an application program (or application), and the like. can include The artificial neural network implemented by the device is as described below.

3 is a conceptual diagram illustrating an embodiment of an artificial neural network included in a device of a product discount coupon providing system.

Referring to FIG. 3 , the artificial neural network may include an input layer IL, a plurality of hidden layers HL1, HL2, ..., HLn, and an output layer OL.

The input layer IL may include i (i is a natural number) input nodes x1, x2, ..., xi. And, vector input data of length i may be input to the input node.

The plurality of hidden layers HL1, HL2, ..., HLn include n (n is a natural number) hidden layers, and the hidden nodes h11, h12, h13, ..., h1m, h21, h22, h23, ..., h2m, hn1, hn2, hn3, ..., hnm). For example, the hidden layer HL1 may include m (m is a natural number) hidden nodes h11, h12, h13, ..., h1m, and the hidden layer HL2 may include m hidden nodes. (h21, h22, h23, ..., h2m), and the hidden layer HLn may include m hidden nodes (hn1, hn2, hn3, ..., hnm).

The output layer OL may include j (j is a natural number) number of output nodes (y1, y2, ..., yj) corresponding to the class to be classified, and results for each class (for example, For example, scores or class scores) can be output. The output layer OL may be referred to as a fully connected layer.

The artificial neural network shown in FIG. 3 may include a branch between two nodes shown as a straight line and weight values between the connected nodes. Here, nodes in one layer may not be connected to each other, and nodes included in different layers may be completely or partially connected.

Each node (eg, h11) in FIG. 3 may receive an output of a previous node (eg, x1), perform an operation, and transfer the result of the operation to a subsequent node (eg, h21). Here, each of the nodes may calculate a value to be output by applying the input value to a specific function (eg, a nonlinear function).

In general, the structure of an artificial neural network is predetermined, and appropriate values of weights according to connections between nodes can be calculated using data for which the correct answer is already known to which class they belong to. Data for which the correct answer is already known may be referred to as 'learning data', and the process of determining weights may be referred to as 'learning'. In addition, a bundle of structures and weights that can be learned independently can be referred to as a 'model'.

That is, a device such as a server illustrated in FIG. 2 may perform operations supported by the system using the artificial neural network illustrated in FIG. 3 . Algorithms and operations executed by the server and store management terminals included in the product discount coupon providing system may be described below.

4 is a flowchart illustrating an embodiment of a product discount voucher providing method of a product discount voucher providing system.

In step S401 , the server may acquire existing sales data of the store, including information on reservation details and visit details of customers by date of the store, from the database. Here, the existing sales data of the store includes at least one of existing customers' reservation history information, visit history information, sales information date information, weather information, store size information including the number of orders and tables in the store, and type and region information of the store. information may be included.

In step S403, the server may predict the demand of the store by using the existing sales data of the store. Here, the server may predict store demand for a preset time period. According to one embodiment, store demand may be predicted for a long time period and/or a short time period.

Here, the long-term time period may be at least one of a year unit, a semi-annual unit, a quarter unit, and a month unit. Also, the short-term time period may be at least one of a weekly unit and a daily unit. The short-term time period may be a time period included in the long-term time period. A server that predicts store demand for a short period of time can predict store demand for each day and hour.

The structure and operation of the server for predicting store demand may be as described below.

5 is a conceptual diagram illustrating an embodiment of an artificial intelligence architecture of a product discount coupon providing system.

Referring to FIG. 5 , the artificial intelligence architecture of the product discount coupon providing system may include a cloud computing solution, a virtual computing machine including a virtual private cloud (VPC), a database, a data collection and processing engine, and the like. Components of the artificial intelligence architecture of the gift voucher providing system and operational flow of each component may be described below.

The event signal generation unit of the artificial intelligence architecture generates a signal at a specified time, specifically, calls a cloud computing solution at a specified time (eg, 7:00 AM) every day. A cloud computing solution may function as an instance-controller.

A cloud computing solution may boot up a virtual computing machine. Here, the virtual computing machine is a computing machine running on a cloud and may be a computing machine such as an EC2 instance. The virtual computing machine executes the registered startup program. Specifically, the virtual computing machine executes an engine startup script (start.py) among execution shell script files (start.sh). Options in the engine startup script are described below.

Updated: If there is no option in the engine startup script, the virtual computing machine updates the engine startup script code to the latest version, assigns the updated option, and then executes the engine startup script again. When the engine start script is executed for the first time, the update option may not be given to the engine start script.

Preprocessing: If the preprocessing option is given to the engine startup script, the virtual computing machine loads the data up to the previous day necessary for learning from the database that stores temporary weather data and other engine data storage, reservation, and store data, and preprocessing By executing the job, the fetched data can be converted into a form that can be processed as learning data.

Model type (model_type) = (LightGBM|XGBoost|Dense): The virtual computing machine selects a model to be trained with a machine learning engine (eg, TMEngine) based on the model type option given to the engine startup script.

Label: After training, the virtual computing machine updates the label data stored one day ago in the prediction result data store, such as data collection and processing engine, and gives the label option to the engine startup script. In general, a label option is given to the script that is executed last.

Forced start (forced_start): When the forced start option is given to the engine start script, the virtual computing machine can execute the engine start script even after a preset maintenance time (eg, MAINTENANCE_TIME) has elapsed.

Shutdown: When a shutdown option is given to the engine startup script, the virtual computing machine shuts down the virtual computing machine after learning is finished. Generally, the last script to be executed is given a shutdown option.

Not_update: If the not_update option is given to the engine startup script, the virtual computing machine may not upload the data generated as a prediction result to the data collection and processing engine after training is finished.

The engine startup script (start.py) runs the machine learning engine (eg TMEngine). A machine learning engine performs learning on data.

The machine learning engine can update training data and convert it into a data set usable in a Python environment by performing a preprocessing operation. The machine learning engine may generate input data and label data, which are training data, based on the transformed data set.

The machine learning engine may classify data for validation from among training data. Specifically, the machine learning engine divides learning data into training data and verification data, and typically the ratio of training data and verification data may be 9:1. Here, for proper evaluation in the model learning process, it is preferable that the data configurations of the training data and the verification data are similar to each other. The machine learning engine analyzes the statistical distribution of the label data and reflects the analysis result to find a point where the label data is divided by a specific ratio based on a specific value.

A machine learning engine defines and trains a model. The types of algorithms constituting the model may vary depending on execution options. Here, the algorithm constituting the model may include an XGBoost algorithm, a LightBGM algorithm, and a deep learning algorithm. Further, the deep learning algorithm may include a deep neural network (DNN)-based algorithm, a convolutional neural network (CNN)-based algorithm, a recurrent neural network (RNN)-based algorithm, and a long short term memory (LSTM)-based algorithm.

The machine learning engine generates prediction values based on input data for prediction prepared in advance through a model. The machine learning engine stores model information, prediction settings, input data to be predicted, and prediction values.

The virtual computing machine stores prediction result data in a distributed search engine. The virtual computing machine may visualize the prediction result data and/or the engine log by linking the prediction result data with the visualization tool, and output an analysis result of the prediction result data and/or the engine log.

The virtual computing machine may analyze the engine log, and may detect whether or not a system failure has occurred based on the engine log analysis result. When a failure of the system is detected, the virtual computing machine transmits a message through a service that generates a notification signal by referring to data stored in the distributed search engine. Based on the received message, the administrator may request the progress of a corresponding task or the like.

A method of predicting store demand of a product discount voucher providing system including a virtual computing machine will be described below.

6 is a flowchart illustrating an embodiment of a store demand prediction method of a product discount coupon providing system.

In step S601, the server may preprocess each information of the store's existing sales data. The server may pre-process each piece of information based on a predefined criterion according to the characteristics of each piece of information, and each piece of information may be pre-processed as follows.

Date data may be preprocessed to indicate the year, month, day, and day of the week, and may be preprocessed to indicate whether it is a weekend, holiday, and/or holiday. Weather data may be preprocessed to indicate the season, temperature, or precipitation of a specific date.

The reservation data includes store visit reservation information of customers, and the server may pre-process the store visit reservation information of customers to be divided into units of a certain time period (eg, week unit). For example, the server may divide the reservation data into the number of reservations from 34 days to 28 days ago, the number of reservations from 27 days to 21 days ago, the number of reservations from 20 days to 14 days ago, the number of reservations from 13 days to 7 days ago, and the number of reservations on the same day. there is.

The server may pre-process existing sales data of the store to indicate the size of the store. The server may calculate the average number of daily reservations for a certain long period (eg, 3 months) before the prediction time, and the average number of daily reservations for a certain short period (eg, 1 month) before the prediction time. can be calculated, and the existing sales data of the store can be pre-processed by calculating the average number of visitors in the previous month at the time of prediction.

The server may pre-process existing sales data of the store by calculating the number of orders and the number of tables in the store. In addition, the server may pre-process existing sales data of the store to indicate not only store size information but also industry information and/or region information.

The server may pre-process the store's existing sales data to indicate the tendency of customers to visit the store. The server may calculate a walk-in ratio based on the total number of reservations in the store and the number of walk-in reservations. In addition, the server may pre-process the existing sales data of the store by classifying the total reservation information of the store according to the day of the week and calculating the number of reservations for each day of the week.

In step S603, the server may learn each information of the pre-processed store's existing sales data by using a plurality of preset algorithms, and generate a plurality of store demand prediction models. Existing sales data of stores may be divided into training data for model learning and test data for model evaluation, depending on their use in the process of creating and evaluating a demand prediction model. Further, the training data may be divided into model establishment data, which is data for establishing a model, and validation data for verifying the suitability of the model.

The server may learn at least some of the store's existing sales data (eg, model establishment data) using a plurality of different algorithms, and use other data (eg, verification data) to determine the learning result. suitability can be judged. Here, a method of extracting model establishment data used for learning from existing sales data and a method of performing learning using the extracted model establishment data will be described below.

7 is a flowchart illustrating an embodiment of a method for generating a demand prediction model using model establishment data and test data of a product discount coupon providing system.

In step S701, the server may classify the existing sales data of the store according to a preset ratio based on the label data. Specifically, the server may classify the training data according to a preset ratio based on the label data of each training data among the existing sales data of the store. Here, the label data may indicate the number of reservations for stores per day. That is, the server may analyze the distribution of the label data and classify the existing sales data of the store based on the distribution.

For example, if 80% of the sample has label data less than or equal to 20 and 20% of the sample has label data greater than 20, the server can classify the store's existing sales data into two groups based on the label data. there is. Here, the first group may be a data group including label data equal to or less than 20, and the second group may be a data group including label data greater than 20.

In step S703, the server may extract model establishment data from existing sales data of the store according to a preset ratio. For example, the server may extract 80% of store model establishment data from the first group and 20% from the second group. In addition, existing sales data excluding model establishment data may be verification data. Accordingly, the server may similarly set label distributions of the model establishment data and the verification data.

In step S705, the server may generate a model obtained by learning model establishment data from existing sales data of the store. The server may generate a plurality of demand prediction models by learning model establishment data using different algorithms. Here, the plurality of algorithms include an XGBoost algorithm, a LightBGM algorithm, a deep learning algorithm, a linear regression algorithm, a support vector machine (SVM) algorithm, a decision tree algorithm, and a random forest algorithm. can include Further, the deep learning algorithm may include a DNN-based deep learning algorithm, a CNN-based algorithm, an LSTM-based algorithm, and the like.

The server may improve the accuracy of the predicted value by generating a plurality of demand forecasting models and calculating an average of the predicted values of the plurality of models. Alternatively, the server may improve the accuracy of the predicted value by comparing the accuracy of each of the plurality of demand forecasting models and calculating the predicted value using one model having the highest accuracy.

In step S707 , the server may verify the generated demand forecasting model using verification data, which is data excluding model establishment data from existing sales data of the store. Here, the server may verify the suitability of the demand prediction model using a K-fold cross validation algorithm. The server may calculate the accuracy of the generated demand prediction model as a result of the verification using the verification data.

The server may generate one demand prediction model by using a plurality of models in learning model establishment data and generating a demand prediction model. An operation of generating a store demand prediction model, which is one of a plurality of models of the server, will be described below.

8 is a flowchart illustrating an embodiment of a method of generating a store demand prediction model of a product discount coupon providing system.

The server performs learning based on existing sales data of the store including at least one of store reservation history information, visit history information, date information, weather information, store size information, industry information, and/or regional information. can In particular, the server may generate a demand prediction model of the store by learning model establishment data, which is part of the training data, from existing sales data of the store.

In step S801, the server may input reservation history information and visit history information to the first artificial neural network. Information on the number of reservations per day may be input from among the reservation details information to the first artificial neural network, and information on the number of visitors among the visit history information may be input to the first artificial neural network. Here, the first artificial neural network may be an artificial neural network based on one of CNN, DNN, and RNN.

For example, if the first artificial neural network is a CNN-based artificial neural network, the server receives information on the number of reservations per day, performs a one-dimensional synthetic product operation, and inputs the value calculated as a result of the synthetic product operation into an activation function to obtain a final output value can be calculated. Here, the activation function may be one or more of a MaxPooling function, a Softmax function, and a ReLu function.

In step S803, the server may input the existing sales data of the store excluding the reservation detail information and the visit detail information to the second artificial neural network. Here, the existing sales data of the store, excluding the reservation history information and the visit history information, may include at least one of date information, weather information, store size information including the number of orders and tables in the store, and type and region information of the store. can And, the second artificial neural network may be one of DNN, CNN, and RNN. And, the second artificial neural network may be an artificial neural network separate from the first artificial neural network, but is not limited thereto.

For example, if the second artificial neural network is a DNN-based artificial neural network, the server inputs date-related data and time-related data to a dense layer, and inputs an output value of the dense layer to an activation function to calculate a final output value. there is. Here, the activation function may be one or more of a MaxPooling function, a Softmax function, and a ReLu function.

In step S805, the server may merge the output value of the first artificial neural network and the output value of the second artificial neural network. That is, the server may merge at least some of information of the number of reservations, number of visitors, date information, weather information, store size information, industry type information, and/or regional information output as a result of input to each of the artificial neural networks. Data generated as a result of merging at least some pieces of information among information on the number of reservations, information on the number of visitors, date information, weather information, store size information, type of business information, and/or regional information is referred to as merged data.

In step S807, the server may define a store demand prediction model based on the output value of the first artificial neural network and the output value of the second artificial neural network. The server may define a store demand prediction model based on the merged data by inputting the merged data into a model based on an artificial neural network.

Here, the artificial neural network may be one of DNN, CNN, and RNN. For example, if the second artificial neural network is a DNN-based artificial neural network, the store demand prediction model inputs the input merged data into the dense layer and inputs the output value of the dense layer into the activation function to calculate the final output value. there is. Here, the activation function may be one or more of a MaxPooling function, a Softmax function, and a ReLu function.

The server may train a store demand prediction model with existing sales data of a store including one or more pieces of information. In addition, the server may predict store demand using the store demand prediction model that has been trained.

Referring back to FIG. 6 , in step S605 , the server may assign hyperparameter settings to each of a plurality of algorithms used to generate each of a plurality of store demand prediction models. A hyperparameter means a variable that is a training criterion in a model, and includes at least one of the number of hidden layers, the number of nodes, an activation function, an optimization method, test epoch, batch, and dropout. can do. The server can assign algorithms using predefined optimal hyperparameter settings to generate the model. In addition, the server may store the settings of each hyperparameter in algorithms for generating a model on one library. Accordingly, the server can easily switch between a plurality of algorithms in performing a learning operation.

In step S607, the server may evaluate each of a plurality of store demand prediction models, and determine a store demand prediction model based on the evaluation result. Specifically, the server may calculate an evaluation index of each of a plurality of store demand prediction models using test data excluding training data from existing store data. An operation of the server calculating evaluation indicators for evaluating each of the plurality of store demand prediction models and evaluating each of the plurality of store demand prediction models based on the evaluation indicators will be described below.

9 is a flowchart illustrating an embodiment of a store demand prediction model evaluation method of a product discount coupon providing system.

The server evaluates the store demand prediction model with existing sales data of the store, including at least one of store reservation history information, visit history information, date information, weather information, store size information, industry information, and/or region information. can In particular, the server may calculate a predicted value by inputting test data from the store's existing sales data to the store's demand forecasting model, and evaluate the store's demand forecasting model based on the comparison result between the calculated predicted value and the actual label data. can

In step S901, the server may calculate evaluation indicators of each of the store demand prediction models. Evaluation indicators may be calculated based on error values of the store demand prediction model. Evaluation indicators may include mean absolute error (MAE) and mean absolute percentage error (MAPE).

The server may calculate the MAE of the store demand prediction model by performing the operation of Equation 1.

In Equation 1, y _i indicates the predicted value of store demand through the model, and x _i indicates the label data value of the store test data. n may indicate the total number of test data.

In addition, the server may calculate MAPE of the store demand prediction model by performing the operation of Equation 2.

In Equation 2, A _t indicates an actual store demand value, and F _t indicates a predicted store demand value through a model. n may indicate the total number of test data.

In step S903, the server may calculate a distribution of evaluation indicators of each of the store demand prediction models. The server may calculate a bivariate normal distribution between max_APE and MAPE based on max average percentage error (max_APE) and MAPE, which are maximum values of average error rates of each of the predicted data. Therefore, the server can calculate the 2D normal distribution of max_APE and MAPE. In calculating the normal distribution of max_APE and MAPE, the server may calculate the 2D normal distribution of max_APE and MAPE by further reflecting at least one of MAE, Pearson's coefficient, and SSE.

In step S905, the server may extract a PCA (principal component analysis) vector between evaluation indicators of each of the store demand prediction models. The server may extract a PCA vector based on the 2D normal distribution of max_APE and MAPE.

In step S907, the server may calculate the accuracy of the store demand prediction model based on the distribution of evaluation indicators of each of the store demand prediction models and the PCA vector. The server may calculate the distribution of max_APE and MAPE. Here, the calculated distribution may be a distribution of Euclidean distances of max_APE and MAPE. The server may evaluate stability, accuracy, and/or reliability of the store demand prediction model based on at least one of the calculated skewness, mean, median, and standard deviation of the distribution. Further, the server may evaluate stability by a demand forecasting model based on the size and direction of the PCA vector.

Referring back to FIG. 6 , in step S609 , the server may predict store demand using the determined store demand prediction model. The server may obtain input data for predicting store demand from the database, and may predict store demand by inputting the acquired input data into the determined store demand prediction model.

The server may periodically update store sales data in the database. In addition, the server may continuously review whether the store demand prediction model is suitable using the updated sales data of the store. When accuracy greater than a preset value is calculated using the updated sales data of the store, the server may maintain the store demand prediction model. On the other hand, as a result of the test using the updated store sales data, if the accuracy is less than a preset value using the updated store sales data, the server may adjust the hyperparameters of the store demand prediction model (hyperparameter tuning). there is.

Long-term forecasts of demand may differ from the sum of short-term forecasts of demand. For example, although 10,000 product discount coupons were issued according to the long-term demand forecast for 3 months later, 11,000 coupons may actually be needed as a result of adding the short-term demand forecast for a week later. According to an embodiment of the present invention, the store demand prediction model may increase inference accuracy of the model by defining a difference between a long-term demand prediction value and a short-term demand prediction value as new input data (learning data).

The difference between the sum of long-term and short-term demand forecasts can be caused by various reasons. That is, the difference to be used as learning data for retraining the demand forecasting model may be influenced by various factors including weather, climate, economic activity, level of advertising and marketing, and existing input data during the corresponding period. Although the server operator may predict various factors for the difference, in the machine learning of the demand forecasting model according to an embodiment of the present invention, by configuring the learning data as input data, the model itself identifies the correlation to improve the accuracy of the demand forecasting power can be raised

Specifically, the difference between the sum of short-term demand forecasts, long-term demand forecasts, and short-term demand forecasts derived in the past and long-term forecasts of demand can be used as additional input data, and the actual number of visitors can be used as output to improve the accuracy of the existing model. there is.

According to one embodiment, the server may execute an operation of predicting the demand of the store at a specific time. The server may predict store demand at a predetermined time and store the prediction result. Here, the pre-designated time may be a time prior to the opening time of the store. The manager of the server may check the prediction result calculated by the server, and if the prediction result is normal, the prediction result data may be provided to the manager terminal of the store. Therefore, the server can reduce the time required to obtain input values and the prediction time of the model.

Referring back to FIG. 4 , in step S405 , the server may determine the quantity of product discount coupons provided by the store based on the demand prediction result of the store. According to an embodiment, the quantity of product discount coupons provided by the store may be determined based on a long-term demand prediction result of the store. That is, the number of product discount coupons provided may indicate the number of seats that can be reserved and used in a store for a long period of time.

According to an embodiment, the server may determine a difference between the number of available seats in a store for a long period of time for which demand forecasting is completed and a demand prediction result value for a long period of time as the number of product discount coupons provided.

In step S407, the server may determine the usable quantity of product discount coupons based on the demand prediction result of the store. According to an embodiment, the usable quantity of product discount coupons may be determined based on a short-term demand forecast result of the store. That is, the usable quantity of product discount coupons may indicate the quantity of seats that can be reserved and used in a store for a short period of time.

According to an embodiment, the server may determine a difference between the number of available seats in a store for a short period of time for which demand forecasting is completed and a demand forecast result value for a short period of time as the quantity of product discount coupons provided. Here, the server may determine the quantity of product discount coupons provided in units of days and units of time.

According to another embodiment, the usable quantity of product discount coupons may be determined as a preset minimum value. For example, in the case of a specific point in time when demand forecast for a store has not been completed, the quantity of product discount coupons that can be used at the specific point in time may be set to a preset minimum value. Afterwards, when demand forecasting is completed for a period including a specific point in time, the server may reflect the demand forecasting value to the available quantity of product discount coupons.

In step S409, the server may provide the product discount coupon to the consumer's user terminal. Here, the server may analyze the loyalty of each customer to each store based on the existing sales data of the store, and may preferentially provide product discount coupons or notify issuance of product discount coupons to filtered customers based on the customer loyalty analysis result. can An operation of the server for analyzing customer loyalty to each store based on existing sales data of the store may be described below.

10 is a flowchart illustrating an embodiment of a customer loyalty evaluation method of a product discount coupon providing system.

Referring to FIG. 10 , the server may perform a customer loyalty analysis operation for the store, and in order to perform the customer loyalty analysis operation, the server may retrieve customer data of the store from a database. The server may retrieve store information, customer information, visit date and time for each customer, number of visits for each customer, and total purchase amount information for each customer from the database.

In addition, the server may determine whether the data retrieved from the database is valid. The server preferably fetches 'non-failing', 'complete' and 'intact' data from the database. The server can filter data through isValid: true of each order of data. In addition, since the schema of the data in the past changes and it is difficult to expect it to be complete, the server can only load data for about 3 months to 2 years. The server that retrieves data from the database can perform the following operations.

In step S1001, the server acquires store visit record information including visit date information and visit frequency information for each customer of the store. The server may calculate R variables and F variables, which are input values of the RFM algorithm, based on store visit record information for each customer.

Here, the R variable may be a variable calculated by quantifying a difference between the time of analyzing loyalty and the time of the customer's recent visit. The server may calculate the R variable using Equation 3.

Here, Ri indicates the time point of the customer's i-th visit, and t indicates the current time. Therefore, through Equation 3, the server may calculate the R variable by reflecting the recency of the data.

Further, the F variable may be a variable calculated by quantifying the number of visits to the store and/or the frequency of visits by customers during a certain period in the past from the time point of the loyalty analysis. The server may quantify the number and/or frequency of store visits as follows.

Due to the nature of the industry, customers who visit the store on a specific day (eg, Christmas, New Year's Day, etc.) or by chance may not visit the store again with a high probability. Therefore, if a customer has an F(original) variable of 1 (i.e., if the total number of transactions is 1), there is little need to differ from a customer with an F variable of 0. can be calculated Specifically, the server may calculate the F variable using Equation 4.

Here, Foriginal may indicate the number of store visits by customers, and H[x] may indicate a step threshold function. The step threshold function may be a discrete step threshold function or a step threshold function.

In step S1003, the server obtains store sales information generated by customers. The server may calculate M variables, which are input values of the RFM algorithm, based on sales information for each customer of the store. Here, the M variable may be a variable calculated by quantifying a customer's sales value for a certain period in the past from the point of time of analyzing the loyalty.

In step S1005, the server standardizes store visit record information of customers and sales information generated by customers. Each of the R, F, and M variables indicates information about the customer's transaction pattern. However, if each variable is simply added and combined into a single score, the unique size relationship of the variables cannot be reflected. For example, assume that a store has a customer who has made a transaction relatively recently (R = 300, F (original) = 1) and a customer who has made 300 transactions in the past (R = 1, F (original) = 300). do. Since each of the two customers has an RFM score of 301, it can be concluded that the importance of both customers is the same. However, the score calculated according to the above example may not be suitable for actual analysis. This is because summing and comparing data is meaningful only when the shape of the data is the same.

In addition, standardization of data is necessary to confirm the absolute position of a specific customer in the entire sample according to the characteristics of the store. For example, in a small store, a regular customer who has made 3 transactions may be of high importance, but in a large store, a regular customer who has made 3 transactions may be of low importance. That is, even if a customer made the same three transactions, the importance may be different depending on the characteristics of the store. When standardization is performed, the position of the customer in the sample is known. Therefore, it is possible to quantify the importance of a single customer appropriately to the store situation, and to confirm not only the relationship between size but also statistical position. In addition, since it has the effect of setting the same standard for different stores, it is possible to establish a more suitable sales strategy by comparing the number of good customers through comparison with other stores.

Therefore, it is desirable to standardize each of the variables in order to give statistical significance to the data and derive analysis results more suitable for the store. The server may standardize each of the R, F, and M variables input to the RFM algorithm. According to an embodiment, it may be assumed that each of the R, F, and M variables are distributed in the form of a normal distribution function, and the server may standardize each of the R, F, and M variables based on the normal distribution function.

Due to the industry characteristics of stores that are subject to customer loyalty analysis, the percentage of one-time visit customers may be high. Thus, the distribution of customers' F variable may follow a logarithmic distribution and/or a Pareto distribution. The server may assign an F variable value of 0.5 to 1.5 to a customer who visits once to secure a statistical significance of the F variable. In standardizing the F variable, the server may standardize the F variable on the assumption that the cumulative distribution function of the F variable is a Pareto distribution and/or a log distribution.

When the server analyzes the sales information, when a simple transaction amount is used as a quantified value, the sales information analysis model does not accurately reflect the actually extracted sales information. According to an embodiment, the sales information may include data without sales because flexible work methods such as walk-in customers, consultation reservations, and 0-won service menus exist. If data without sales is found in the data pre-processing step, data without sales may not be filtered out. Primarily, in analyzing sales information, the server may analyze sales information for each customer by reflecting payment history information of the customer.

Therefore, as a way to solve this problem, data binning can be applied to obtain sales information regardless of patterns. An operation of a server that standardizes sales information for each customer by applying data binning may be described below.

The server may classify sales information for each customer in units of time intervals. For example, if the sales indicated by the sales information of a specific customer for 3 months is 25,000 won, the server may classify the sales information of the specific customer in units of time intervals. The server may classify the sales amount indicated by the sales information of a specific customer for 3 months in units of one month.

The server may first standardize sales information for each customer in a unit of time period. In addition, the server may calculate an average value of sales information for each customer in a first standardized time period unit.

The server may secondarily standardize the firstly standardized sales information for each customer in the unit of time period by reflecting the average value of the sales information. Here, in standardizing store visit record information of customers and sales information generated by the customers, the server may standardize standardized visit date information, visit frequency information, and sales information for each customer for all customers. That is, the server may standardize visit date information, visit frequency information, and sales information for each customer based on averages and distributions of R variable values, F variable values, and M variable values of all customers.

In addition, the server may calculate a customer's grade based on R variable values, F variable values, and M variable values calculated based on visit date information, visit frequency information, and sales information for each customer standardized for all customers. According to an embodiment, the server may classify customers by grade using Equation 5. Equation 5 is as follows.

는 천장 함수를 지시하고, n은 고객을 분류하는 등급들의 개수를 지시하며, erfc(x)는 1-erf(x)(오차 함수)를 지시할 수 있다. In Equation 5, Gdata indicates the customer's rating according to the data,

indicates a ceiling function, n indicates the number of ratings for classifying customers, and erfc(x) may indicate 1-erf(x) (error function).

According to an embodiment, the server may classify customers by grade using Equation 6. Equation 6 is as follows.

In Equation 6, normalcdf(x) may indicate a normal distribution function.

According to another embodiment, the server may calculate a customer rating based on the F variable for all customers. If it is assumed that the distribution function of the F variable follows the Pareto distribution, the server may substitute data into the cumulative distribution function in order to calculate an alpha coefficient according to actual data. The alpha coefficient can be calculated through Equation 7. Equation 7 is as follows.

In Equation 7, p indicates the proportion of one-time customers. The server may classify customers who have visited once as the lowest level. In addition, the server may divide the intervals excluding the normalized cumulative interval of p in the cumulative distribution function of the Pareto distribution into n-1 pieces and assign grades from 1 to n-1. The server may classify customers based on visit frequency data through Equation 8. Equation 8 is as follows.

In step S1007, the server may generate a loyalty analysis model by learning standardized store visit record information and standardized sales information. Here, the loyalty analysis model may be an artificial neural network model having, as input values, a difference value between a loyalty analysis time point and a customer's latest visit point, a customer's number of visits to a store, and a sales value generated by a customer in a unit of a preset time period.

In step S1009, the server may classify the customers into one of pre-classified groups based on loyalty by using the loyalty analysis model. According to one embodiment, the pre-classified groups may be regular customers, new customers, outgoing customers, outgoing customers, and stable customers.

Referring back to FIG. 4 , in step S411 , the server may receive product discount coupon use request information from the consumer's user terminal. Here, the use request information may include user account information, information about a store where the product discount coupon is to be used, identification information such as a code for specifying the product discount coupon, and the like. Further, the server may activate the product discount voucher based on the product discount voucher use request information. An operation of the server for activating the product discount coupon may be described below.

11 is a flowchart illustrating an embodiment of a product discount voucher activating method of a product discount voucher providing system.

In step S1101, the server receiving the product discount coupon usage request information may transmit information on the usable quantity of the product discount coupon in the store to the consumer's user terminal. Here, the usable quantity information of the product discount coupon may be quantity information determined based on a demand prediction result of the store. The quantity information value may be a difference value between the number of available seats in the store during the predicted demand time period and the demand prediction result value. Alternatively, the quantity information value may be a preset minimum value, which is a quantity that can be reserved during a time period in which the demand forecast of the store is not completed.

In step S1103, the server may receive reservation request information for a store where the product discount coupon is to be used from the consumer's user terminal. The consumer can check available quantity information of the product discount coupon through the user terminal and request a reservation corresponding to the available quantity information of the product discount coupon. And the user terminal may transmit the reservation request information generated by the consumer to the server.

In step S1105, the server may activate a product discount voucher based on the reservation request information of the store. When the store can be reserved based on the reservation request information received from the consumer's user terminal, the server completes the store reservation process and transmits the store reservation information to the manager terminal of the store. In addition, the server that has completed the store reservation process may activate the product discount coupon.

That is, the product discount coupon provided through the product discount coupon providing system is preferably activated by completing a reservation procedure according to a reservation request received from a user.

Referring back to FIG. 4 , in step S413 , the server may transmit information requesting use of product discount coupons to the store. Here, the information requesting use of the product discount coupon may include account information of the consumer who used the product discount coupon, reservation information by the consumer, and the like. Therefore, the store can check whether or not to make a store reservation based on the customer's reservation information, and allow the customer to use the gift certificate on the reserved date and time.

Operations of devices constituting the product discount voucher providing system for providing and using the product discount voucher described in the embodiments of FIGS. 4 to 11 may be as described below.

12 is a flowchart illustrating an embodiment of providing and using product discount coupons between a product discount coupon providing system, a store, and a user terminal of a consumer.

In step S1201, the server may predict store demand in a preset time period. According to one embodiment, the server may predict the demand of each store for a long time period and/or a short time period. Here, the long-term time period may be at least one of annual, semiannual, quarterly, and one month. Further, the short-term time period may be at least one of 1 week and 1 day. Also, a server that predicts store demand for a short period of time may predict store demand for each day and hour.

In step S1203, the server may determine the number of product discount coupons provided by the store. According to an embodiment, the quantity of product discount coupons provided by the store may be determined based on the estimated store demand value predicted in step S1201, and specifically, based on the result of the store demand forecast for a long time period.

In step S1205, the server may request issuance of a product discount voucher from the store manager terminal. In order to request issuance of the product discount voucher, the server may pay the store side an amount obtained by reflecting the product discount rate and commission in an amount based on the product price and the issued number of product discount vouchers.

In step S1207, the store manager terminal may approve issuance of the product discount voucher to the server. That is, the store manager terminal may transmit product discount voucher issuance approval information to the server. The server receiving the product discount voucher issuance approval information may issue the product discount voucher based on the product discount voucher provision approval information.

In step S1209, the server may determine the number of product discount coupons available for each period in the store. Here, the usable quantity of product discount coupons in the store may be determined based on the store demand forecast value predicted in step S1201, and according to an embodiment, may be determined based on the store demand forecast result for a short time period.

After determining the number of usable product discount coupons for each period, the server may transmit advertisement information for selling product discount coupons to consumers. Here, the server may calculate customer loyalty to the store from existing sales data of the store, and transmit advertisement information to the customers based on the loyalty information.

In step S1211, the consumer may request purchase of a product discount voucher from the server using the user terminal. The consumer may proceed with a payment procedure with the server using the user terminal and pay a price corresponding to the product discount coupon.

In step S1213, the server may approve the product discount voucher purchase request received from the consumer's user terminal. The server may transmit product discount voucher-related information based on the consumer's product discount voucher purchase request information. Here, the information related to the product discount coupon may include information on available stores and unique identification information of the product discount coupon.

In step S1215, the consumer's user terminal may request the server to use the product discount voucher. The consumer's user terminal requesting the use of the product discount voucher may transmit product discount voucher use request information including unique identification information of the product discount voucher to the server.

In step S1217, the server receiving the product discount coupon use request information may transmit information on the quantity of product discount coupons available for each period in the store derived in step S1209 to the user terminal of the consumer.

In step S1219, the consumer can check information on the quantity of product discount coupons available for each period displayed through the user terminal, and request a store reservation at a date and time when the product discount coupon can be used using the user terminal.

In step S1221, the server may perform a store reservation procedure based on the store summary request information received through the user terminal. The server may transmit store reservation request information to a manager terminal of the store. The manager terminal of the store may complete the store reservation procedure by approving the store reservation request received from the server and transmitting store reservation approval information to the server.

In addition, the server that has completed the store reservation procedure corresponding to the store summary request information of the user terminal may activate the product discount coupon. The server may transmit store reservation information to a manager terminal of the store.

In step S1223, the consumer can visit the store on the date and time indicated by the reservation information in step S1221 and use the product discount coupon when paying for the product.

In step S1225, the manager terminal of the store may notify the server that the product discount coupon is exhausted after approving the use of the consumer's gift certificate and completing the product payment procedure.

Among the operations shown in FIG. 12 , operations indicated by bold lines may be operations involving the movement of funds. That is, in step S1205, the store may obtain an amount for issuing a product discount voucher from a management entity that manages a server of the system. Further, in step S1211, the user may pay an amount for purchasing a product discount voucher to a management entity of the system. Through the method of providing product discount coupons described in the embodiments of FIGS. 4 to 12, each subject can obtain the following benefits.

13 is a conceptual diagram illustrating an embodiment of a product discount coupon providing system, a flow of money between a store and a consumer.

Referring to FIG. 13 , exchange of product discount vouchers may be performed on a product discount voucher providing platform. The product discount voucher providing platform may be provided by the product discount voucher providing system 1310 .

The product discount voucher providing platform may include a product discount voucher issuance/management platform 1301 and a product discount voucher transaction/management platform 1302 . Here, the product discount voucher issuance/management platform 1301 is provided between the product discount voucher providing system 1310 and the store 1320, and is a platform for issuing store product discount vouchers and managing whether or not the issued product discount vouchers are used. can In addition, the product discount coupon transaction/management platform 1302 is provided between the product discount coupon providing system 1310 and the consumer 1330 terminal, and is a platform for managing transaction information of the product discount coupon in the store and whether or not the traded product discount coupon is used. can The product discount voucher providing platform including the product discount voucher issuance/management platform 1301 and the product discount voucher transaction/management platform 1302 may be managed by the product discount voucher providing system 1310 .

Referring to FIG. 13 , the product discount voucher providing system 1310 may pay a price for issuing a product discount voucher to the store 1320 . Here, the price for issuing a product discount coupon may be an amount obtained by discounting a discount rate and an amount corresponding to a fee from the total product price. Accordingly, the store 1320 may secure the product discount voucher issuance price paid by the product discount voucher providing system 1310 regardless of operating conditions.

The product discount coupon providing system 1310 may advertise product discount coupons of the store 1320 to the consumer 1330 . The consumer 1330 may check the advertisement and purchase a product discount voucher of the store 1320 . Accordingly, the product discount voucher providing system may secure an amount corresponding to a sales commission, which is an amount obtained by subtracting a product discount voucher issuance price from the purchase price of the product discount voucher of the consumer 1330 .

It is preferable that the consumer 1330 proceeds with a reservation procedure of the store 1320 prior to using the product discount coupon. In addition, the consumer 1330 can visit the store 1320 on the reserved date and time, use the product, and use the product discount coupon. As a result, the consumer 1330 can use the product at a discounted price.

And, by managing the use of the product discount voucher through reservation, the product discount voucher providing system 1310 can provide the consumer 1330 with a product discount voucher that minimizes constraints on usage conditions (eg, use schedule, use store, etc.). can

The methods according to the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded on a computer readable medium. Computer readable media may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on a computer readable medium may be specially designed and configured for the present invention or may be known and usable to those skilled in computer software.

Examples of computer readable media include hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter or the like as well as machine language codes generated by a compiler. The hardware device described above may be configured to operate with at least one software module to perform the operations of the present invention, and vice versa.

Although described with reference to the above embodiments, those skilled in the art will understand that the present invention can be variously modified and changed without departing from the spirit and scope of the present invention described in the claims below. You will be able to.

Claims (12)

In the operating method of the product discount providing system based on the demand forecast of the store,
obtaining, from a database, existing sales data of the store including reservation history information and visit history information of customers for each date of the store;
predicting demand for the store using existing sales data of the store;
determining a usable quantity of the product discount vouchers based on a demand forecast result of the store;
providing the product discount coupon to a consumer's user terminal;
Receiving the product discount voucher use request information from the user terminal; and
And transmitting request information for use of the product discount voucher to the store. The method of claim 1,
The step of predicting the demand of the store using the existing sales data of the store,
preprocessing the existing sales data of the store;
generating a plurality of store demand prediction models by using a plurality of preset algorithms learned from the preprocessed store's existing sales data;
assigning a hyperparameter of each of the generated store demand prediction models to each of the plurality of preset algorithms; and
And determining a store demand prediction model based on evaluation results of each of a plurality of store demand prediction models. The method of claim 1,
The step of predicting the demand of the store,
predicting store demand in a first time period; and
A method of operation comprising predicting demand in a store in a second time period including a first time period. The method of claim 3,
Determining the quantity of product discount coupons provided by the store based on the demand prediction result of the store;
The quantity of product discount coupons provided by the store is determined based on the demand of the store in the second time period;
characterized in that the usable quantity of the product discount coupon is determined based on the demand of the store in the first time period. The method of claim 1,
obtaining store visit record information of customers based on existing sales data of the store;
obtaining sales information of the store generated by the customers; and
Analyzing loyalty to the store of each of the customers based on store visit record information of the customers and sales information of the store;
The consumer,
Among the customers, characterized in that at least one or more customers filtered on the basis of loyalty to the store, operating method. The method of claim 1,
Receiving the product discount voucher use request information,
transmitting available quantity information of the product discount voucher in the store indicated by the product discount voucher use request information to the user terminal;
Receiving reservation request information of the store using the product discount coupon from the user terminal; and
And activating the product discount voucher based on the reservation request information of the store. As a server of a product discount providing system based on store demand forecast,
a memory in which at least one program instruction is stored;
a processor to execute the at least one program instruction; and
Including a transmitting and receiving device that is connected to the network and performs communication,
The at least one program command,
Obtaining existing sales data of the store, including reservation history information and visit history information of customers for each date of the store, from the database;
predicting the demand of the store using the existing sales data of the store;
determining a usable quantity of the merchandise discount vouchers based on a demand prediction result of the store;
providing the product discount coupon to a user terminal of the consumer;
receiving a request for using the product discount voucher from the user terminal;
receiving the product discount voucher use request information from the user terminal; and
A server configured to transmit request information for use of the product discount voucher to the store. The method of claim 7,
The at least one program command,
In executing to predict the demand of the store using the existing sales data of the store,
preprocessing the existing sales data of the store;
generating a plurality of store demand prediction models by using a plurality of preset algorithms learned from the preprocessed store's existing sales data;
assigning a hyperparameter of each of the generated store demand prediction models to each of the plurality of preset algorithms; and,
The server further executed to determine a store demand prediction model based on evaluation results of each of the plurality of store demand prediction models. The method of claim 7,
The at least one program command,
In the implementation to predict the demand of the store,
predict demand in the store in the first time period; and
The server is further executed to predict demand of the store in a second time period including the first time period. The method of claim 9,
The at least one program command,
is further executed to determine a quantity of product discount coupons provided in the store based on the result of predicting demand in the store;
The quantity of product discount coupons provided by the store is determined based on the demand of the store in the second time period;
Characterized in that, the usable quantity of the product discount coupon is determined based on the demand of the store in the first time period. The method of claim 7,
The at least one program command,
obtaining store visit record information of customers based on the existing sales data of the store;
obtaining sales information of the store generated by the customers; and,
It is further executed to analyze the loyalty of each of the customers to the store based on the store visit record information of the customers and the sales information of the store;
The consumer,
Characterized in that, among the customers, at least one customer filtered based on loyalty to the store, the server. The method of claim 7,
The at least one command,
In executing to receive the product discount voucher usage request information,
transmitting available quantity information of the product discount voucher in the store indicated by the product discount voucher use request information to the user terminal;
Receiving reservation request information of the store using the product discount coupon from the user terminal; and
A server configured to activate the product discount voucher based on the reservation request information of the store.

Images