KR20220076013A - System for providing artificial intelligence based restaurant supplies automatic ordering service for business-to-business market - Google Patents

Abstract

An artificial intelligence-based automatic food material ordering service provision system for the B2B market is provided, which automatically uploads POS data and collects POS data from at least one merchant terminal and at least one merchant terminal that automatically pays for food materials ordered The collection unit, the identification unit that extracts POS sales product information and POS sales information included in the POS data to identify the type and quantity of food materials, collects POS sales information, weather information, and day information and consumes it with a pre-established consumption prediction algorithm A prediction unit that predicts the type and quantity of food materials to be prepared, the type of food material predicted by the prediction unit and and an automatic ordering service providing server including an ordering unit that performs automatic ordering based on the quantity and the learned purchase pattern.

Description

SYSTEM FOR PROVIDING ARTIFICIAL INTELLIGENCE BASED RESTAURANT SUPPLIES AUTOMATIC ORDERING SERVICE FOR BUSINESS-TO-BUSINESS MARKET}

The present invention relates to an artificial intelligence-based automatic food material ordering service provision system for the B2B market, and a platform in which artificial intelligence automatically places an order instead of a human by learning the order data and pattern of each restaurant to predict the type and quantity of food materials provides

As competition in the self-employed industry intensifies, small business owners have come to recognize the importance of supply chain management. Supply chain management of food materials improves the competitiveness of small businesses by linking with various functional strategies of restaurants and furthermore, connecting with external raw material suppliers, distributors, and customers. The link between supply chain management and functional strategy can be a very complex task. From a strategic point of view, a restaurant's structure and management strategy must be consistent in order to succeed in the market. Management strategies include cost advantage strategy, differentiation strategy, and concentration strategy, which are competitive strategies. The strategies that can achieve competitive advantage are cost advantage strategy and differentiation strategy. The cost advantage strategy is based on economies of scale in food production, the experience curve effect, the utilization of cooking production facilities, and the efficiency of the kitchen organization. On the other hand, a differentiation strategy is possible only by acquiring the exclusivity of services and menus. Both strategies interact with the food production strategy and are also linked to the supply chain.

At this time, a platform that can automatically place an order for food materials has been researched and developed. Notice on July 15, 2019), receives the order details of each school, compares the stock and order quantity for each food material in the entered order history, decides whether to place an order for each food material, and executes the order If it is determined, the meal management system performs an order using the supplier information registered in advance for each food material, and assigns the delivery details to each pre-registered delivery driver according to the order details of each school, and a preset time By automatically accessing the customer server every cycle, downloading the order list, and automatically printing the label paper for the ordered item, the distribution center of the food material supplier operates the labor more efficiently, and the human A configuration for minimizing human error during labeling operation by printing a warning mark recognizable to an operator on the label of the product for a product in which errors occur relatively frequently is disclosed.

However, among the above configurations, in the former case, the term automatic is used, but it is only calculating the difference between stock and order quantity, and it is not a configuration to automatically predict the order quantity and place an order, and in the latter case, a human attaching a label sticker There is no meaning other than replacing the Although the procurement of ingredients and menu production, service, profit, and customer satisfaction are connected, all the activities of information collection, supplier selection, evaluation, negotiation, contracting, purchasing, and maintaining relationships for food material procurement are carried out by individual self-employed and small business owners. In many cases, economies of scale are not achieved due to lack of human and material resources. Food material procurement is connected through the value chain, leading to a link between excellent suppliers - operators - knowledgeable employees - satisfied customers. In particular, the role of excellent suppliers is important. In most cases, it is not easy to even find and order ingredients one by one after finishing the meal. Accordingly, research and development of a platform that can predict the menu to be sold by accumulating the sales menus of each affiliate store, calculate the types and quantity of food materials required for the predicted menu, and automatically place an order is required.

An embodiment of the present invention identifies the type and quantity of food materials required for this by extracting product information and sales information sold during a preset period by collecting POS data from at least one affiliated store terminal, and consumes such as weather or day of the week. Predict the type and quantity of food materials to be consumed by inputting variables that affect Thus, it is possible to provide an artificial intelligence-based automatic food material ordering service provision method for the B2B market that can extract the type and quantity of at least one food material and process it automatically. However, the technical task to be achieved by the present embodiment is not limited to the technical task as described above, and other technical tasks may exist.

As a technical means for achieving the above-described technical problem, an embodiment of the present invention automatically uploads POS data, and POS data from at least one affiliated store terminal and at least one affiliated store terminal for automatically paying the ordered food material amount A collection unit that collects POS sales product information and POS sales information included in the POS data, an identification unit that identifies the type and quantity of food materials, and a pre-established consumption forecast by collecting POS sales information, weather information and day information The prediction unit that predicts the type and quantity of food materials to be consumed with an algorithm, the food material predicted by the prediction unit after learning the purchasing pattern through deep learning using the purchase data that has been ordered from at least one affiliated store terminal accumulated during the preset period and an automatic ordering service providing server including an ordering unit that performs automatic ordering based on the type and quantity of the purchase pattern and the learned purchase pattern.

According to any one of the above-described problem solving means of the present invention, by collecting POS data from at least one affiliated store terminal, product information and sales information sold for a preset period are extracted to identify the type and quantity of food materials required for this, It predicts the type and quantity of food materials to be consumed by inputting variables that affect consumption, such as weather or day of the week, and furthermore, after learning the purchase pattern based on the purchase data of food materials ordered from each affiliate store, purchase data and It is possible to extract the type and quantity of at least one food material based on the predicted consumption data and automatically process the order, so that time and manpower are not wasted by letting the artificial intelligence algorithm take over the order processing that should be done by humans.

1 is a view for explaining an artificial intelligence-based automatic food material ordering service provision system for a B2B market according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating an automatic ordering service providing server included in the system of FIG. 1 .
3 and 4 are diagrams for explaining an embodiment in which an artificial intelligence-based automatic food material ordering service for the B2B market according to an embodiment of the present invention is implemented.
5 is an operation flowchart illustrating a method of providing an artificial intelligence-based automatic food material ordering service for a B2B market according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated, and one or more other features However, it is to be understood that the existence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded in advance.

The terms "about", "substantially", etc. to the extent used throughout the specification are used in or close to the numerical value when manufacturing and material tolerances inherent in the stated meaning are presented, and are intended to enhance the understanding of the present invention. To help, precise or absolute figures are used to prevent unfair use by unconscionable infringers of the stated disclosure. As used throughout the specification of the present invention, the term "step of (to)" or "step of" does not mean "step for".

In this specification, a "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. In addition, one unit may be implemented using two or more hardware, and two or more units may be implemented by one hardware. Meanwhile, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or to reproduce one or more processors. Thus, as an example, '~' denotes components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'. In addition, components and '~ units' may be implemented to play one or more CPUs in a device or secure multimedia card.

Some of the operations or functions described as being performed by the terminal, apparatus, or device in the present specification may be performed instead of by a server connected to the terminal, apparatus, or device. Similarly, some of the operations or functions described as being performed by the server may also be performed in a terminal, apparatus, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal means mapping or matching the terminal's unique number or personal identification information, which is the identification data of the terminal. can be interpreted as

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a view for explaining a system for providing an artificial intelligence-based automatic food material ordering service for a B2B market according to an embodiment of the present invention. Referring to FIG. 1 , the artificial intelligence-based automatic food material ordering service providing system 1 for the B2B market includes at least one affiliated store terminal 100 , an automatic ordering service providing server 300 , and at least one information providing server 400 . ) may be included. However, since the artificial intelligence-based automatic food material ordering service providing system 1 for the B2B market of FIG. 1 is only an embodiment of the present invention, the present invention is not limitedly interpreted through FIG. 1 .

At this time, each component of FIG. 1 is generally connected through a network 200 . For example, as shown in FIG. 1 , at least one affiliate store terminal 100 may be connected to the automatic ordering service providing server 300 through the network 200 . In addition, the automatic ordering service providing server 300 may be connected to at least one affiliated store terminal 100 and at least one information providing server 400 through the network 200 . In addition, the at least one information providing server 400 may be connected to the automatic ordering service providing server 300 through the network 200 .

Here, the network refers to a connection structure in which information exchange is possible between each node, such as a plurality of terminals and servers, and an example of such a network includes a local area network (LAN), a wide area network (WAN: Wide Area Network), the Internet (WWW: World Wide Web), wired and wireless data communication networks, telephone networks, wired and wireless television networks, and the like. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi (Wi-Fi) , Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth (Bluetooth) network, NFC ( Near-Field Communication) networks, satellite broadcast networks, analog broadcast networks, Digital Multimedia Broadcasting (DMB) networks, and the like are included, but are not limited thereto.

In the following, the term at least one is defined as a term including the singular and the plural, and even if the at least one term does not exist, each element may exist in the singular or plural, and may mean the singular or plural. it will be self-evident In addition, that each component is provided in singular or plural may be changed according to embodiments.

At least one affiliated store terminal 100 automatically sends POS (Point of Sale) data using an artificial intelligence-based automatic food material ordering service related web page, app page, program or application for the B2B market. 300) may be a terminal for uploading. At this time, when the merchant terminal 100 uploads the POS data to the automatic ordering service providing server 300 in real time or periodically, it is uploaded in the background mode, so that the actual data upload is not displayed in the foreground. It may be a terminal that prevents it. In addition, the affiliated store terminal 100 may be a terminal that pre-registers a payment method with the automatic ordering service providing server 300 so that an immediate payment can be made when an automatic order is placed. In addition, the affiliated store terminal 100 uploads a review to the automatic ordering service providing server 300 for the type and quantity of food materials ordered automatically or whether there is insufficient or leftover, or data on the sales of the day, that is, POS data. It may be a terminal that reduces errors by comparing and analyzing the types and quantities of food materials ordered from the automatic ordering service providing server 300 and the quantity of consumed food materials by continuously uploading them. Of course, it is clear that continuously uploading the POS data does not involve human intervention, but is continuously uploaded to the automatic ordering service providing server 300 in the background mode in the affiliated store terminal 100 . However, this does not preclude a person from manually entering or uploading.

Here, the at least one affiliated store terminal 100 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop. In this case, the at least one affiliated store terminal 100 may be implemented as a terminal capable of accessing a remote server or terminal through a network. The at least one affiliated store terminal 100 is, for example, a wireless communication device that ensures portability and mobility, including navigation, Personal Communication System (PCS), Global System for Mobile communications (GSM), Personal Digital Cellular (PDC), PHS(Personal Handyphone System), PDA(Personal Digital Assistant), IMT(International Mobile Telecommunication)-2000, CDMA(Code Division Multiple Access)-2000, W-CDMA(W-Code Division Multiple Access), Wibro(Wireless Broadband Internet) ) terminal, a smart phone, a smart pad, a tablet PC, etc. may include all kinds of handheld-based wireless communication devices.

The automatic ordering service providing server 300 may be a server that provides an artificial intelligence-based automatic food material ordering service web page, an app page, a program, or an application for the B2B market. The automatic ordering service providing server 300 collects POS data from at least one affiliated store terminal 400, and then extracts POS sales product information and POS sales information of each affiliated store to identify the type and quantity of food materials. can In addition, the automatic ordering service providing server 300 collects at least one variable that affects consumption, such as weather or day of the week, collects purchase data for ordering food materials for each affiliate store, and then purchases data for each variable. After learning and testing at least one artificial intelligence algorithm to determine the causal relationship between It can be a server. If the initial data set is not sufficient, errors or errors may be large, so the automatic ordering service providing server 300 extracts the predicted order data and the type and quantity of food materials actually used from the sales information of the POS data for comparative analysis. After that, it may be a server that re-learns the artificial intelligence algorithm.

Here, the automatic ordering service providing server 300 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop.

The at least one information providing server 400 may affect consumption such as weather information, news, etc. with or without using a web page, an app page, a program or an application related to an artificial intelligence-based automatic food material ordering service for the B2B market. It may be a server that provides at least one type of information. Here, the at least one information providing server 400 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop.

2 is a block diagram for explaining an automatic ordering service providing server included in the system of FIG. 1, and FIGS. 3 and 4 are artificial intelligence-based automatic food material ordering services for the B2B market according to an embodiment of the present invention. It is a diagram for explaining an implemented embodiment.

Referring to FIG. 2 , the automatic ordering service providing server 300 includes a collection unit 310 , an identification unit 320 , a prediction unit 330 , an ordering unit 340 , an error correction unit 350 , and a modeling unit ( 360), a big data conversion unit 370, and an accuracy measurement unit 380 may be included.

The automatic ordering service providing server 300 according to an embodiment of the present invention or another server (not shown) operating in conjunction with at least one affiliated store terminal 100 is an artificial intelligence-based automatic food material ordering service application for the B2B market, When transmitting a program, an app page, a web page, etc., at least one affiliated store terminal 100 may install or open an artificial intelligence-based automatic food material ordering service application for the B2B market, a program, an app page, a web page, etc. . In addition, the service program may be driven in at least one affiliated store terminal 100 using a script executed in a web browser. Here, the web browser is a program that enables the use of a web (WWW: World Wide Web) service, and refers to a program that receives and displays hypertext written in HTML (Hyper Text Mark-up Language), for example, Netscape. , Explorer, Chrome, and the like. In addition, the application means an application on the terminal, for example, includes an app (App) executed in a mobile terminal (smartphone).

Referring to FIG. 2 , the collecting unit 310 may collect POS data from at least one affiliated store terminal 400 . At least one affiliated store terminal 400 may automatically upload POS data.

The identification unit 320 may identify the type and quantity of food materials by extracting POS sales product information and POS sales information included in the POS data. In this case, the POS data includes customer payment data corresponding to each date and time. For example, there is a restaurant that specializes in champon and sells sweet and sour pork, and the menu includes champon and sweet and sour pork (small, medium, large). Ingredients for 1 bowl of Jjambbong include 1 cup of pork (150g), 2/3 cup of green onion (50g), 1/2 cup of onion (40g), 1/2 cup of cabbage (40g), 1/3 cup of carrot (20g). , Zucchini 1/2 cup (60g), wheat flour 1 cup (150g), fine red pepper powder 1 tablespoon (5 g), coarse red pepper powder 1 tablespoon (5 g), beef kelp 1/2 tablespoon (5 g), flower salt 1/5 tablespoon (2 g) ), soy sauce 1 tbsp (10g), cooking oil 2 tbsp (15g), black pepper powder are required. In a small bowl of sweet and sour pork, 600g pork tenderloin, 1/2 onion, 1/2 cucumber, 1T minced garlic , Fine salt 1/2T, sesame oil 1T, pepper powder 1/6T, starch powder 1 cup (150g), sugar 6T, ketchup 2T, vinegar 6T, oyster sauce 2T, soy sauce 2T, starch powder 2T. At this time, assuming that 100 bowls of Champon and 50 bowls of sweet and sour pork were sold on Monday, November 30, 2020, the type and quantity of each ingredient used per day can be summed up.

In addition, since POS data, that is, payment data, is generated for each date and time zone, which menu is served the most by day of the week, which day of the week there are many customers, at what time of the week, how much is the amount of delivery increased due to Corona, and orders for delivery Data on the number of disposable containers to be used will be accumulated. Initially, there may not be enough datasets, but if POS data is continuously collected and the influence of news or weather is taken into account, the prediction accuracy can gradually increase.

On the other hand, each affiliate may be reluctant to disclose the recipe, that is, the amount of ingredients required to make one menu. This is because, when setting up a restaurant, they work as apprentices for several years to learn recipes or learn to cook at least 10 million won. Accordingly, the type and quantity of food ingredients for each menu are collected and stored in the private blockchain, used only when calculating the type and quantity of at least one ingredient for each order, and then deleted again from the server end. It can be configured to ensure thorough security.

A private blockchain may be based on Hyperledger Fabric. Hyperledger Fabric is a modular architecture, a platform for developing blockchain solutions and applications. It is one of the projects of Hyperledger established by the Linux Foundation, and is managed by various technical steering committees and maintainers of various organizations. Hyperledger Fabric supports components such as consensus and member services in a Plug & Play manner. It is a permissioned blockchain that only authorized users can participate in, and is a type of private blockchain.

At this time, Smart Contracts of Hyperledger Fabric are written in chaincode and are called by applications outside the blockchain when the application needs to interact with the ledger. In most cases, chaincode only interacts with the database component of the ledger, the world state, not the transaction log. Chaincode is implemented in several programming languages. The currently supported chaincode language is Java, and more languages may be supported in the future.

The prediction unit 330 may collect POS sales information, weather information, and day information to predict the type and quantity of food materials to be consumed using a pre-established consumption prediction algorithm. For example, it is assumed that a restaurant specializing in naengmyeon sells naengmyeon and steamed dumplings, and it is assumed that daily sales are hardly affected by other socio-economic factors such as changes in the business district besides the weather. In this case, assuming that meteorological data is used as an independent variable of the menu and the correlation analysis and prediction model, the information providing server 400 may be a synoptic meteorological observation server of the Korea Meteorological Administration. At this time, factors that are estimated to have low correlation with sales volume are excluded, factors that are difficult for users to obtain when the developed forecasting model is actually operated are excluded, and only representative factors are selected and forecasted for factors with overlap between meteorological factors. Assume

Meteorological factors may be average temperature, precipitation duration, daily precipitation, average wind speed, average relative humidity, average sea level pressure, insolation, maximum daily snow cover, average total cloudiness, fog duration, and the like. In general, daily sales of naengmyeon show a clear trend of seasonal fluctuations in summer and small in winter. If this trend can also be seen in the distribution of weekly and monthly average fluctuations, it can be inferred that the sales volume of naengmyeon has a strong correlation with temperature fluctuations. However, if it is estimated that other weather factors or other variables besides temperature are also affected by the large daily fluctuations at the same time, other variables other than the meteorological factors may be considered.

The daily sales forecasting model of naengmyeon can be grouped into weekdays and weekends. After analyzing the correlation coefficient between the sales volume of naengmyeon and the total sales volume of other items and weather factors, a multiple regression model can be developed focusing on variables with large correlation coefficient values, After determining the coefficient of determination and the coefficient of determination of the model divided into weekdays and weekends, the better one can be set as the model. Among the independent variables of the weekday and weekend forecasting models, the meteorological factors may be temperature, solar radiation, and clouds, and may include day-of-week variables. In the weekend model, the sales volume of the previous weekend was included in the independent variable, but in the weekday model, the increase in the coefficient of determination of the sales variable one day before is small and the response time is short when operating the forecasting model, so it can be excluded because of the low availability. At this time, predictive models for other items can be developed, and different coefficients of determination and independent variables can be shown for each item.

Alternatively, since data that changes according to time series are analyzed, the ARIMA time series research model can be used in consideration of this. For example, it is possible to forecast food and beverage sales for today, this week, or this month based on POS data, which is food and beverage sales by item generated at affiliate A for the past 5 years. For this, one of the time series analysis techniques, ARIMA-Intervention, is used to classify major time-series influence variables that are expected to affect food and beverage sales and estimate the extent of their change. Based on the results, future occurrences can occur. It is possible to estimate the expected monthly or daily sales of food and beverages.

In the time series analysis for the ARIMA-intervention model, major events that are judged to have affected the sales of each franchise by time in the past are applied as an intervention variable, and the effect can be reflected in the overall demand. The ARIMA-intervention model can use a variety of observations for each time point. In addition, it is evaluated as a multivariate time series forecasting model that can structurally measure external shocks and demands such as specific events. For this reason, it is possible to solve the problem that it is difficult to measure the structural and influence of demand, which has been pointed out as a disadvantage of the time series analysis method. The ARIMA-intervention model is evaluated as a model that overcomes the shortcomings of the time series model by presenting the influence coefficient for each intervention event (event), and its predictive power is high.

Regression analysis is one of the statistical techniques mainly used in data prediction. However, in explaining or forecasting future demand for each restaurant using regression analysis, income variables (such as real or nominal income or disposable income or GDP), exchange rates (such as KRW/USD or KRW/YEN or KRW/CNY), or international It is recognized that independent variables such as oil price do not satisfactorily explain market demand fluctuations. For example, if the demand for each restaurant is adversely affected due to Corona (COVID-19), etc., or if China issues a Korean-Korean ordinance that forcibly bans visits to Korea due to South Korea's deployment of THAAD (high-altitude missile) This is because disaster or artificial national policy matters are difficult to set as a causal function formula and difficult to explain with data, so regression analysis cannot provide satisfactory predictive data.

The ARIMA time series research model analyzes time series data divided by time lag units, constructs the relationship into a model, and predicts future demand or data through the time series model (classical model) and stochastic model. Time series models can be divided into two types. Classical time series analysis methods include moving average method, exponential smoothing method, and decomposition method. The probabilistic time series model, which can be called a modern model, is a research modeled by introducing a stochastic method to time series data (original data), and is also known as the ARIMA analysis method. ARIMA analysis model refers to an analysis model constructed by combining A.R. (Auto Regression) and M.A. (Moving Average).

The autoregression model is a research model that assumes that there is a correlation between the data at every moment for the data x(t) for each lag unit. However, white noise exists in the correlation between data. There is no temporal correlation in white noise. The autoregressive analysis model was estimated through the Yule-Walker equation under the assumption that the observed data at the present time can be expressed in the form of a function of their past observed data values.

The moving average analysis model assumes that the current observation value can be expressed in the form of a linear combination of white noises. An Auto Regressive-Moving Average analysis model can also be used by combining autoregressive and moving average models. Using these models, a probabilistic time series analysis model was established as the most stable prediction method. It was Box-Jenkins' ARIMA time series analysis model that established this methodology. ARIMA time series analysis has been applied in various academic and industrial fields, and has established itself as a representative time series analysis method in the field of predicting future demand.

A three-stage research model exists for time series analysis. The first stage is the stage in which the researcher identifies a tentative model from the time series data. The second step is to estimate and test the coefficients of the previously identified time series analysis model (Estimation and Testing). In the parameter estimation and verification stage, iterative calculations and verification of the statistical program for the fit of models and coefficients are performed. This is because it is possible to prevent the intervention of a researcher in advance and improve the reliability of analysis results through repeated calculations and verification. Finally, the suitability of the selected model is diagnosed (Diagnostic Checking), and the diagnostic process is repeated until a satisfactory model is selected. In one embodiment of the present invention, it can be set based on such a three-step model.

<ARIMA Time Series-Intervention Model>

As mentioned above, time-series data related to restaurant demand are greatly affected by external political or economic events such as natural disasters (earthquake, tsunami, typhoon, corona (COVID-19), outbreak of war, and the Korean War). Events are called intervention variables in the ARIMA research model Intervention variables play a difficult role in designing the research model because they cause rapid changes in the normal flow and stochastic structure of time series data. The predictive power of future demand can be greatly improved by using the ARIMA intervention model that can include .

If there is time series data Yt, which is raw data, and external intervention occurs M times, the ARIMA intervention model is as Equation 1 below.

At this time, C = constant term, αi = coefficient of Xi,t, Nt = error term, and the error term can be expressed as ARIMA(p,d,q)(P,D,Q)12 as follows.

D is the number of seasonal differences, d is the number of non-seasonal differences, φp(B) is the non-seasonal AR model, Φp(Bs) is the seasonal AR model, θq(B) is the non-seasonal MA model, ΘQ(Bs) is the seasonal MA model , εt is an error term (White Noise) or the same as in Equation 4 below.

Substituting Equation 4 into Equation 1, it can be defined as Equation 5 below.

In general, intervention variables can be divided into two types: Pulse intervention, which is an event that only temporarily affects the dependent variable, and Step intervention, which is an event that continuously affects the dependent variable. If it is assumed that the pulse intervention occurred during time t = i, the pulse variable Xi,t is defined as

In this case, Xi,t is a dichotomous variable, meaning that Xi,t is intervened only at time i, Xi,t = 0 before time i, Xi,t =1 at time i, Xi,t = again after time i becomes 0. If it is assumed that the step intervention occurred during time t = i, the step variable Xi,t is defined as follows.

The step variable function Xi,t means that the intervention of an external event occurring at a certain point i continues to occur even after that. This is also a dichotomous variable, meaning that Xi,t intervenes at time i and thereafter. Therefore, Xi,t = 0 before time i, Xi,t=1 at time i, and Xi,t = 0 after time i.

The analysis procedure of the ARIMA-intervention model predicts the future using the model finally selected after the above-mentioned model identification, estimation and verification, and model diagnosis. In the identification step, it is possible to determine whether the data is normal or not through a sequential diagram of the data, an autocorrelation function (ACF), a partial autocorrelation function (PACF), and the like. To this end, the difference (d), AR(p), and MA(q) orders may be determined after determining whether the data is a difference or a log transformation.

At this stage, several tentative research models can be selected, and the most appropriate time series analysis model can be selected through estimation and validation and finally model diagnosis. It is possible to determine the parameters and coefficients of the model identified in the estimation step of the model. Determine the values of p autoregression (AR) parameters φ1, φ2,...,φp and q moving average (MA) parameters θ1, θ2,...,θq for the identified model, and calculate the t values of the parameters. It is a step to determine whether it is significant through

In the final stage, model diagnosis, it is verified whether the autocorrelation coefficient value of ACF and the autocorrelation coefficient value of PACF of the model that passed the goodness of fit sufficiently converge to zero. If there is a 5% significance level for both ACF and PACF, at least one of the lag autocorrelations may exceed this threshold. After the three-step analysis process is complete, predictions can be made.

<Data collection>

For example, suppose we want to collect monthly POS data from January 2015 to January 2018 of a member store A, a jjamppong restaurant located in Sadang-dong, using time series data. In this case, drinks may be present in addition to the meal menu, but beverages may be excluded if it is assumed that the food materials are mainly fruits and vegetables. Prior to ARIMA analysis, it is possible to identify whether the meal monthly sales data is normal. In order to predict future sales, it is necessary to find and remove abnormalities such as seasonality, and to find stable and fixed trends. For example, in raw data, there is a time series sequence chart of monthly meal sales at Affiliate A, and sales that are periodically repeated in July and August, peak season in summer, and December and January, peak season in winter If there is also a sharp decline, after confirming this trend, seasonal difference and non-seasonal difference can be implemented once. At this time, it can be confirmed that the time series data, which is the raw data, does not show an increase any more and shows a steady state. After performing data difference, autocorrelation function and partial autocorrelation function can be analyzed to determine the ARIMA p, d, and q order numbers. MA (moving average) order q can be determined through the auto correlation function, MA (moving average) order q can be determined through the auto correlation function, and AR (autoregression) order p can be discriminated through a partial autocorrelation function.

After examining the short-term lag of the graphs of the autocorrelation function and partial autocorrelation function, if it is not possible to confirm that the spike protrudes outside the confidence interval, and does not protrude outside the confidence interval of the graph of the autocorrelation function and partial autocorrelation function, MA The ARIMA(0,1,0)(0,1,0) model in which the order of (0) and AR(0) is determined and the scaffolding procedure and the seasonal difference are each 1 can be set as the final model. In addition, an intervention variable can be included in the ARIMA model, and the intervention variable can use a policy such as a phenomenon in which the number of tables decreases and increases due to the phased distance of Corona (COVID-19), and that only packing is possible after 9 o'clock. .

Also, the accuracy of the estimated value of the ARIMA intervention model can be measured through the process of reviewing the actual value and the prediction error. For example, you can use the MAPE goodness-of-fit coefficient, where MAPE means Mean Absolute Percentage Error. The MAPE goodness-of-fit coefficient can interpret the goodness-of-fit for each interval. If the MAPE value comparing the prediction error between the measured value and the predicted value is less than 5%, it appears and it can be judged that it is showing excellent predictive power. In addition, if the residuals of the autocorrelation function and the partial autocorrelation function are all within the 95% confidence interval, it can be analyzed as satisfying the white noise condition.

The ordering unit 340 learns a purchase pattern through deep learning using purchase data that has ordered food materials from at least one affiliated store terminal 400 accumulated for a preset period, and then predicts the type and quantity of food materials predicted by the prediction unit. , automatic ordering can be performed based on the learned purchase pattern. In this case, the at least one affiliated store terminal 400 may pay the automatically ordered food material amount. This is because if only POS data is used, it may not be accurate because it is only an estimate of the item to be ordered. POS data->Estimation of type and quantity of food materials->Learning with purchase data->Comparing the estimated value with the learning result, you can proceed to the process of re-learning.

The error correction unit 350 collects feedback data on the difference between the type and quantity of food materials automatically ordered from the at least one affiliated store terminal 400 and the type and quantity of food materials actually used, and for the error occurrence part It is possible to re-learn a pre-established consumption prediction algorithm.

The modeling unit 360 builds a database with POS data, weather information, day information and purchase data, and then performs database classification (DB Classification), opinion mining (Opinion Mining), and data processing (Data Processing) to perform data processing After creating a set (DataSet) and modeling the artificial intelligence algorithm through a learning segment and a judgment segment as a data set, a consumption prediction algorithm can be built. In this case, the learning segment may include a feature extraction and model learning, and the judgment segment may include a sensing element and a model judgment.

At this time, the sentiment score can be calculated by applying opinion mining to the preprocessed data, and at this time, the sentiment dictionary builder ( Sentiment Dictionary Builder). Based on this, sentiment analysis is carried out, and in this process, a detailed analysis algorithm for detailed analysis of sentiment may be added.

Online reviews are created in the form of multimedia such as text, images, and videos, and feature selection using machine learning may be used to select important words within the text. FS can filter out unnecessary words and improve classification performance in selecting important words. Alternatively, for an effective opinion mining model for sentiment analysis, a combination of a learning classifier and an attribute selection method FS may be used. Here, FS may be CFS (Correlation based Feature Selection), IG (Information gain), etc., and the classifier used for classification is Logistic Regression, Decision Tree, Neural Network, Support Vector. Machine (Support Vector Machine), Naive Bayesian Network, Random Forest, Bagging, Stacking, Random Subspace Any one or a combination of at least one can

FS, which appeared with the advent of data sets containing hundreds of thousands of attributes, can model the process of data creation and management and secure variables. to maintain or improve performance. For attribute selection, filter learning, in which the derivation algorithm learns as an independent preprocessing step, which depends on the general properties of the data, is used as a black box, and predictive performance is used to determine the relative usefulness of a subset of variables. Wrapper learning using a learning method as a subroutine that evaluates and calls a property selection algorithm to evaluate a subset of each feature (features) Subsets are displayed as included in the classification criteria construct, and may include embeddings that allow searching and finding dependencies in the combined space of subsets of properties and hypotheses.

Logistic regression is a regression analysis method used for the purpose of linear or non-linear classification. This method performs regression on the class by setting the output of the training instance belonging to each class to 1 and the output of the non-member instance being set to 0. A linear equation is derived from this result, and then, when performing unknown class verification, it is a classifier that calculates the result of each linear equation and classifies it by selecting the largest value. The goal of naive Bayesian is to accurately predict the class of a test instance by learning a training instance that includes class information. In this case, the naive Bayesian network can be specialized in the form of a naive because it depends on two important simplification assumptions. In particular, it is assumed that the prediction properties are conditionally independent for each given class, and the identified or latent properties influence the prediction process.

The neural network is an algorithm that mimics the neurons of the human brain. It is a multilayer perceptron and consists of an input layer, a hidden layer, and an output layer. It is composed of nodes. Next, the hidden layer processes the variable values transmitted from the input layer as a non-linear function and transmits them to the output layer. The output layer has nodes corresponding to the target variables. Random forest is an ensemble learning method for performing regression and other tasks, in which a number of decision trees are constructed during a training period, and can be operated by classification mode or class output, which is the average prediction of individual trees.

Random Forest is suitable for emotion classification because it is a more suitable classifier for sentiment classification compared to various ensemble classifiers due to the many overlapping values of text classification. In a random space, the training data set can be modified using algorithms such as bagging, where the modification can take place in the attribute space rather than the instance space. Random forest can use arbitrary subspaces to construct and aggregate base classifiers, and when the data set has a large number of redundant or unrelated properties, the base classifier is more effective in random subspaces compared to the original feature space. You can choose a classifier. The combined decision of this base classifier may outperform a single classifier built on the training data set of the entire set of attributes.

A decision tree is a non-parametric method that can be applied to classification or regression tasks. It is a hierarchical data structure for supervised learning in which the input space is divided into local regions to predict the dependent variable. The support vector machine provides an optimal separation hyperplane state for solving the classification task. Mathematical analysis can represent nonlinear problems related to the input space as linear problems in high-dimensional isolated hyperplanes.

Bagging is an abbreviation of Bootstrap aggregating. After repeatedly sampling according to a uniform probability distribution, classification prediction is made according to voting by combining a decision tree model. Bagging is mainly used in regression analysis and can improve stability and accuracy and reduce variance. Unlike bagging, which combines models of the same type, stacking is the task of combining predictions of different algorithms by learning various learning algorithms. In particular, it is possible to find and combine the classifier with the best performance after estimating the performance of which classifier is reliable using the meta learner.

The big data conversion unit 370 parallelly and distributedly stores raw data including POS data, weather information, day information, and purchase data, and stores unstructured data and structured data included in the stored raw data. ) data and semi-structured data are refined, pre-processing including classification as meta data is performed, pre-processed data is analyzed including data mining, and the analyzed data is visualized. can be printed out.

The accuracy measurement unit 380 may measure a difference between the type and quantity of food materials automatically ordered from the at least one affiliated store terminal 400 and the type and quantity of food materials actually used. In this case, the type and quantity of food materials actually used may be checked based on POS data collected from at least one affiliated store terminal 400 .

Hereinafter, an operation process according to the configuration of the automatic ordering service providing server of FIG. 2 will be described in detail with reference to FIGS. 3 and 4 as an example. However, it will be apparent that the embodiment is only one of various embodiments of the present invention and is not limited thereto.

Referring to Figure 3a, the platform according to an embodiment of the present invention is a wholesale market direct transaction artificial intelligence platform for alleviating the burden of purchasing food materials in the restaurant business, and one embodiment of the present invention among a plurality of systems is a third system, that is, an artificial intelligence algorithm. It is an automatic food material ordering system. It is possible to develop food material consumption patterns and learning data, link external API information of weather and day information, and model artificial intelligence algorithms based on historical data. 3b and 3c are system structural diagrams of the present invention. The data of FIG. 3c is information for deriving the results of the artificial intelligence algorithm, and is made based on the collection of internal information and external information, and the DB/DataSet stores the collected information. Classifying data for each visualization through mining after preprocessing, modeling is for deriving results through modeling by matching each algorithm with feature extraction, weights, and fixed variable values It is a visualization material that can be viewed by administrators or users (merchants) through pages, etc.

Referring to FIG. 3D , which shows the service of the present invention, food materials can be identified based on product sales history and sales information through POS information linkage of a restaurant, and a consumption prediction algorithm can be developed based on external information. In addition, by performing deep learning by combining the self-accumulated past order history and unique algorithm in the server end of the present invention, the three classification technologies are combined to finally implement an artificial intelligence algorithm-based automatic food material ordering system. have. At this time, for deep learning, POS information-based food materials are identified, consumption prediction algorithms are modeled based on external information and POS information, and mobile applications for customers (merchants) are developed through unique variable values and internal algorithm development based on past orders. You can build servers and databases in this platform.

Referring to FIG. 4A , according to an embodiment of the present invention, direct transaction in the wholesale market is possible through a mobile application. It enables reasonable purchase of wholesale market food materials that were only available for offline purchase through the mobile application, and the sale price compared to the auction market price on the day A transparent transaction environment can be provided through disclosure. In addition, the efficiency of the wholesale market and restaurant can be maximized through the artificial intelligence algorithm, and price and demand prediction can be made based on the artificial intelligence algorithm, so the price in the wholesale market can be stabilized, and automatic ordering and quotation comparison can be performed. Through this, restaurant owners can help them focus on restaurant operations, and the wholesale market reduces resources for guaranteeing margins and sales and customer service.

Referring to FIG. 4B , the service according to an embodiment of the present invention has know-how acquired while operating offline distribution and vegetable stores, price fluctuations by product, product cause information, information on major purchase products by restaurant, major sales products by industry Data, major sales product information by region, sales volume information by product, sales status information by wholesale market middleman, order pattern and amount information by restaurant, and actual efficiency information of logistics routes by time period have been used to secure and build big data, so it is modeled. In other words, if modeling is performed to create an artificial intelligence algorithm, an algorithm with an accurate predictive degree can be created.

Referring to Figure 4c, market price fluctuations and weather data, product recommendation data by category, order pattern data by company, relationship analysis algorithm for agricultural products past data and environmental variables, relationship analysis algorithm for agricultural production and growth characteristics, products used by customers and purchase variables Order prediction algorithm, other distribution channel selling price collection and price relationship analysis algorithm, sales information and environment variable relationship prediction algorithm can be derived, and price information according to customer industry classification, product recommendation for each customer, and supplier inventory burden can be derived. Demand forecasting, order pattern analysis, sales estimation, and demand forecasting for

Referring to FIG. 4D , since the platform according to an embodiment of the present invention brings goods directly through the public wholesale market, it is not possible to go through an intermediate distribution stage and builds an infrastructure that can receive various fresh food or industrial products. Therefore, immediate supply and direct supply are possible. In addition, as shown in FIG. 4E, logistics infrastructure efficiency was secured across the country using four public wholesale markets as logistics bases.

The matters that have not been described with respect to the method of providing an artificial intelligence-based automatic food material ordering service for the B2B market of FIGS. 2 to 4 are described above with respect to the method of providing an artificial intelligence-based automatic food material ordering service for the B2B market through FIG. 1 Since it is the same as the content or can be easily inferred from the described content, the following description will be omitted.

5 is a diagram illustrating a process in which data is transmitted/received between components included in the artificial intelligence-based automatic food material ordering service providing system for the B2B market of FIG. 1 according to an embodiment of the present invention. Hereinafter, an example of a process in which data is transmitted and received between each component will be described with reference to FIG. 5, but the present application is not limited to such an embodiment, and the example shown in FIG. 5 according to various embodiments described above will be described. It is apparent to those skilled in the art that the data transmission/reception process may be changed.

Referring to FIG. 5 , the automatic ordering service providing server collects POS data from at least one affiliated store terminal (S5100).

And, the automatic ordering service providing server extracts POS sales product information and POS sales information included in the POS data to identify the type and quantity of food materials (S5200), collects POS sales information, weather information, and day information Predict the type and quantity of food materials to be consumed with the constructed consumption prediction algorithm (S5300).

In addition, the automatic ordering service providing server learns a purchase pattern through deep learning using purchase data that has been ordered for food materials from at least one affiliated store terminal accumulated for a preset period, and then the type and quantity of food materials predicted by the prediction unit; Automatic ordering is performed based on the learned purchase pattern (S5400).

The order between the above-described steps ( S5100 to S5400 ) is merely an example and is not limited thereto. That is, the order between the above-described steps ( S5100 to S5400 ) may be mutually changed, and some of these steps may be simultaneously executed or deleted.

Matters that have not been explained about the method of providing an artificial intelligence-based automatic food material ordering service for the B2B market of FIG. Since it is the same as the content or can be easily inferred from the described content, the following description will be omitted.

The method for providing an artificial intelligence-based food material automatic ordering service for the B2B market according to an embodiment described with reference to FIG. 5 is in the form of a recording medium including instructions executable by a computer, such as an application or program module executed by a computer can also be implemented. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer-readable media may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The method for providing an artificial intelligence-based automatic food material ordering service for the B2B market according to an embodiment of the present invention described above may include an application basically installed in a terminal (which may include a program included in a platform or an operating system basically installed in the terminal) ), and may be executed by an application (ie, a program) installed directly on the master terminal by a user through an application providing server such as an application store server, an application, or a web server related to the corresponding service. In this sense, the AI-based automatic food material ordering service providing method for the B2B market according to an embodiment of the present invention described above is implemented as an application (ie, program) installed in a terminal or installed directly by a user and installed in the terminal. It may be recorded on a computer-readable recording medium, such as.

The above description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

Claims (7)

at least one affiliated store terminal that automatically uploads POS data and pays for food materials ordered automatically; and
A collection unit that collects the POS data from the at least one affiliated store terminal, an identification unit that extracts POS sales product information and POS sales information included in the POS data to identify types and quantities of food materials, the POS sales information, weather A purchase pattern using a prediction unit that collects information and day information and predicts the type and quantity of food materials to be consumed with a pre-established consumption prediction algorithm, and purchase data that has been accumulated for a preset period at the at least one affiliated store terminal to order food materials an automatic ordering service providing server including an ordering unit that automatically places an order based on the type and quantity of food materials predicted by the prediction unit after learning through deep learning and the learned purchase pattern;
Artificial intelligence-based food ingredient automatic ordering service provision system for the B2B market, including.
The method of claim 1,
The automatic ordering service providing server,
Collecting feedback data on the difference between the type and quantity of food materials automatically ordered from the at least one affiliated store terminal and the type and quantity of food materials actually used, and re-learning the pre-established consumption prediction algorithm for the error occurrence part error correction unit;
Artificial intelligence-based food material automatic ordering service provision system for the B2B market, characterized in that it further comprises.
The method of claim 1,
The automatic ordering service providing server,
After constructing a database with the POS data, weather information, day information, and purchase data, the database classification (DB Classification), opinion mining (Opinion Mining), and data processing (Data Processing) are performed to create a DataSet. a modeling unit that generates and performs modeling of an artificial intelligence algorithm through a learning segment and a judgment segment with the dataset, and then builds the consumption prediction algorithm;
Artificial intelligence-based food material automatic ordering service provision system for the B2B market, characterized in that it further comprises.
4. The method of claim 3,
The training segment includes a feature extractor and model learning,
The judgment segment is an AI-based automatic food material ordering service providing system for the B2B market, characterized in that it includes a sensing element and a model judgment.
The method of claim 1,
The automatic ordering service providing server,
Raw data including the POS data, weather information, day information, and purchase data are stored in parallel and distributed, and unstructured data, structured data and semi-structured data included in the stored raw data Big data that refines (Semi-structured), performs preprocessing including classification as meta data, analyzes the preprocessed data including data mining, and visualizes and outputs the analyzed data stoker;
Artificial intelligence-based food material automatic ordering service provision system for the B2B market, characterized in that it further comprises.
The method of claim 1,
The automatic ordering service providing server,
an accuracy measuring unit for measuring a difference between the type and quantity of food materials automatically ordered from the at least one affiliated store terminal and the type and quantity of food materials actually used;
Artificial intelligence-based food material automatic ordering service provision system for the B2B market, characterized in that it further comprises.
7. The method of claim 6,
The artificial intelligence-based food material automatic ordering service providing system for the B2B market, characterized in that the type and quantity of the actually used food material is checked based on POS data collected from the at least one affiliated store terminal.

Images