KR102568422B1 - A method for calculating the supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same - Google Patents

Abstract

An apparatus for providing an online commerce brokerage service using an artificial neural network according to an embodiment includes a buyer information storage unit for registering and storing buyer information, and a seller information storage unit for registering and storing seller information and sales product information on products sold by the seller. , an evaluation information generator for generating buyer evaluation information for the buyer and seller evaluation information for the seller, and a first supply price based on information registered in the seller information storage unit for the purchase request product of the buyer. A first supply price calculation unit calculates an expected correction cost that can be added or reduced based on the buyer's requested items for the purchase requested product and the seller's evaluation information on the requested items, and calculates the estimated cost. A second supply price calculation unit configured to calculate a second supply price by reflecting the correction cost in the first supply price may be included.

Description

A method for calculating the supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same}

The present invention relates to a method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same. For example, it is an invention that relates to a technology that provides information on projected costs together with the cost of purchase.

The concept of open market in a general online commerce system is largely general B2C (Business to Consumer) shopping mall, mall-in-mall (store-in type) shopping mall, open (market) mall, B2B (Business to Business), B2B2C (Business to Business to Consumer) , B2G (Business to Government), M-commerce, C-commerce, and T-commerce shopping malls. At this time, the general B2C shopping mall is a form in which the operator directly trades products for general consumers, and the mall-in-mall (store-in type) shopping mall is a form in which the operator or a seller trades products for general consumers. )mall is a form in which general consumers purchase and sell products and consign sales to the shopping mall operator. A general combined shopping mall means a form that combines the consignment sales functions of a mall-in-mall (store-in type) shopping mall and an open (market) mall. .

B2B is a form of transaction between a shopping mall operator and a company, and B2B2C is a form of transaction between a shopping mall operator and a seller with a company or consumer.

In B2B, in particular, a B2B operator that brokers products opens a B2B e-commerce brokerage company as a shopping mall on the Internet, and mediates transactions by involving corporate consumers and corporate sellers as member companies, and pays transaction fees from corporate consumers or corporate sellers. It is a model that earns margin by collecting or generating direct purchase sales.

In such a B2B system, a seller who wants to purchase a specific product or a service transaction requests a quote through a limited number of business partners and searches for product transaction, and the seller (supplier) who receives the request sends a quote prepared to the ordering company. This is done by finding out the estimated unit price through the method of sending, or calculating the estimated unit price by using the information disclosed by the seller for product transaction.

However, such a conventional bidding system incurs a lot of cost and effort in issuing a quotation and establishing a contract by sharing information among limited companies, and in comparison, the so-called high cost that the transaction is not quickly and clearly accomplished. It has a low-efficiency structure.

In addition, in the case of transactions between companies, unlike transactions with general consumers, there are many cases where large-scale products are traded. Due to these characteristics, various additional costs such as management costs, selling costs, and maintenance costs are incurred in the future as well as the initial purchase cost. do.

And these additional costs vary depending on which seller you purchase from, which product you purchase, and furthermore, since they occur differently each time depending on what the buyer wants to buy, the additional cost cannot be measured uniformly. There was no way to do it. Therefore, since the additional cost was calculated simply by the buyer's feelings, the actual cost and the expected additional cost did not match, resulting in a lot of difficulties due to cost problems, which caused a problem that sometimes caused disputes between buyers and sellers. .

Republic of Korea Registered Patent Publication No. 10-2045477 B1 - E-commerce platform evaluation system (2019.11.15)

Therefore, a method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same according to an embodiment are inventions designed to solve the above-described problems. The purpose is to additionally calculate the expected cost that may actually occur and provide it to the buyer.

More specifically, in calculating the expected cost that may occur additionally, by using evaluation information based on the past transaction history of sellers for the items requested by the buyer, the estimated cost that may occur for each seller is more accurately provided to the buyer. That purpose exists.

An apparatus for providing an online commerce brokerage service using an artificial neural network according to an embodiment includes a buyer information storage unit for registering and storing buyer information, and a seller information storage unit for registering and storing seller information and sales product information on products sold by the seller. , an evaluation information generator for generating buyer evaluation information for the buyer and seller evaluation information for the seller, and a first supply price based on information registered in the seller information storage unit for the purchase request product of the buyer. A first supply price calculation unit calculates an expected correction cost that can be added or reduced based on the buyer's requested items for the purchase requested product and the seller's evaluation information on the requested items, and calculates the estimated cost. A second supply price calculation unit configured to calculate a second supply price by reflecting the correction cost in the first supply price may be included.

The second supply price calculation unit may calculate the estimated corrected cost after applying different weights to the purchase request product and the buyer's request items for the purchase request product according to preset criteria.

The second supply price calculation unit evaluates the degree of demand by the purchaser for each item requested by the purchaser for the product requested for purchase, calculates the requested complexity, and calculates the expected correction cost in proportion to the size of the requested complexity. can

The second supply price calculation unit may calculate the expected correction cost based on an evaluation score of a past fulfillment result of the seller with respect to the buyer's requested item for the purchase requested product.

The second supply price calculation unit is characterized in that the expected correction cost decreases as the evaluation score of the past transaction result increases, and the expected correction cost increases as the evaluation score of the past transaction result decreases. A device for providing online commerce brokerage services using intelligence.

The second supply price calculation unit includes a pre-learned expected price calculation artificial neural network module that takes the information on the purchase request product and the evaluation information of the seller as input information and the expected correction cost as output information, The expected price calculation artificial neural network module may perform learning based on a difference between the correction cost and the actual correction cost.

The loss function of the expected price calculation artificial neural network module includes a difference between the expected correction cost for each item of the expected correction cost and the correction cost actually incurred for each item, and the expected price calculation artificial neural network module calculates the loss function Parameters of the artificial neural network module for calculating expected prices may be updated to have a minimum value.

A method for calculating a supplier's supply price using an artificial neural network according to another embodiment includes the steps of registering and storing buyer information, registering and storing seller information and sales product information for products sold by the seller, generating buyer evaluation information about the buyer and seller evaluation information about the seller; calculating a first supply price based on information registered in the seller information storage unit for the product requested by the buyer to purchase; Calculating an expected correction cost that can be added or reduced based on the buyer's requested items for the product requested for purchase and the seller's evaluation information on the requested item, and calculating the estimated correction cost to the first supply price. A step of calculating a second supply price by reflection may be included.

A method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same according to an embodiment provide a supply price for a product to a buyer, including not only the purchase cost but also the expected correction cost that may occur in the future. Since it provides the supply price including the price, there is an effect that the buyer can easily know the total amount that can actually be generated. Therefore, the buyer can more accurately estimate the expected expenditure for purchasing the product, thereby efficiently managing costs, and preventing disputes with the seller due to additional costs incurred in the transaction in advance.

In addition, a method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same according to an embodiment include a first supply price corresponding to a purchase cost to a buyer and a first supply price that reflects an estimated correction cost that may actually occur. 2Since information on the supply price is provided together, buyers have the advantage of being able to choose a seller that suits them more.

1 is a block diagram showing some components of an online commerce system according to an embodiment of the present invention.
2 is a block diagram showing some components of a supply price calculation device according to an embodiment.
3 is a block diagram showing some components of a second supply price calculator according to an embodiment.
4 is a diagram illustrating a screen of a service provided with first supply price information and second supply price information according to an embodiment.
5 is a diagram illustrating a screen of a service on which information on items affecting a second supply price is displayed according to an embodiment.
6 and 7 are diagrams illustrating input information input to the second supply price calculation artificial neural network module and output information output according to an embodiment.
8 and 9 are diagrams illustrating input information input to the buyer/seller matching artificial neural network module and output information output according to an embodiment.

Hereinafter, embodiments according to the present invention will be described with reference to the accompanying drawings. In adding reference numerals to the components of each drawing, it should be noted that the same components have the same numerals as much as possible, even if they are displayed on different drawings. In addition, in describing an embodiment of the present invention, if it is determined that a detailed description of a related known configuration or function hinders understanding of the embodiment of the present invention, the detailed description thereof will be omitted. In addition, embodiments of the present invention will be described below, but the technical idea of the present invention is not limited or limited thereto and can be modified and implemented in various ways by those skilled in the art.

In addition, terms used in this specification are used to describe embodiments, and are not intended to limit and/or limit the disclosed invention. Singular expressions include plural expressions unless the context clearly dictates otherwise.

In this specification, terms such as "include", "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 the existence or addition of more other features, numbers, steps, operations, components, parts, or combinations thereof is not excluded in advance.

In addition, throughout the specification, when a part is said to be “connected” to another part, this is not only the case where it is “directly connected”, but also the case where it is “indirectly connected” with another element in the middle. Terms including ordinal numbers, such as "first" and "second" used herein, may be used to describe various components, but the components are not limited by the terms.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted.

1 is a block diagram showing some components of a method for calculating a supplier's supply price using an artificial neural network and an online commerce brokerage system using the same according to an embodiment of the present invention.

Referring to FIG. 1 , the online commerce brokerage system 1 according to an embodiment includes a buyer terminal device 300 of a buyer 310, a seller terminal device 400 of a seller 410, and a supply price calculation device 100 It may include a mediation service providing device 200 that includes, and the mediation service providing device 200 may be connected and operated online with a plurality of buyers 300a and 300b and a plurality of sellers 400a and 400b. there is. In FIG. 1, a limited number of buyers and sellers are described due to space limitations, but the present invention is not limited thereto and a larger number of buyers and sellers may be included in the online commerce brokerage system 1.

The online commerce brokerage system 1 establishes a product transaction contract between a buyer 310 and a seller 410 through a buyer terminal device 300, a brokerage service providing device 200, and a seller terminal device 400. You can induce an agreement and monitor information about it.

According to an embodiment, when the purchaser 310 registers the quantity and delivery date of each required product to the mediation service providing device 200 through the purchaser terminal device 300, the mediation service providing device 200 classifies the corresponding product into a sales category. The request for quotation for the product purchase of the buyer 310 is transferred to the seller terminal device 400 of the seller 410 registered in advance with . Thereafter, if there is a seller capable of supply among the sellers 410, quote information corresponding to the purchase information requested by the buyer 310 is generated and transmitted to the mediation service providing device 200, and the mediation service providing device 200 ) may transmit it to the purchaser terminal device 300 again.

In addition, the mediation service providing device 200 does not transmit information on the buyer 310's request for quotation to the seller terminal device 400, but utilizes sales information about the seller registered in the mediation service providing device 200. Thus, a quotation for the buyer's request for quotation may be generated, and the generated quotation content may be transmitted to the purchaser terminal device 300 .

The purchaser terminal device 300 may access the mediation service providing device 200 under the control of the purchaser 310 . The buyer terminal device 310 accesses the mediation service providing device 200 and can enter various information about the buyer and purchase information on the product to be purchased into the corresponding system, and register them in the mediation service providing device 200. You can search the seller and the seller's information.

In addition, the buyer terminal device 300 may receive a recommendation for a seller selling a specific product from the mediation service providing device 200 . The buyer terminal device 300 may request a purchase request to the mediation service providing device 200 after selecting a seller to purchase a specific product under the control of the buyer 310 . When receiving a quotation for a product requested to purchase or various analysis results for a specific seller from the mediation service providing device 200, the buyer terminal device 300 displays the received information on the display screen of the buyer terminal device 300. can do.

The seller terminal device 400 may access the mediation service providing device 200 under the control of the seller 410 and input basic information of the seller and various types of information about a product to be sold.

Accordingly, the buyer terminal device 300 and the seller terminal device 400 may include a mobile terminal device capable of accessing the mediation service providing device 200 through a dedicated application or a general purpose application, and include a computing device. For example, PCS (Personal Communication System), GSM (Global System for Mobile communication), PDC (Personal Digital Cellular), 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, smartphone, smartpad, tablet PC, smart watch ), smart glasses, wearable devices, and the like, may include all types of handheld-based wireless communication devices.

The brokerage service providing device 200 may serve to mediate the buyer 310 and the seller 410 so that a transaction contract is created by connecting the buyer 310 and the seller 410, and the brokerage service providing device 200 can generally be implemented as a server.

A server refers to a typical server. A server is a computer hardware on which a program is executed. It monitors and controls the entire network, such as printer control or file management, or connects to other networks through a mainframe or public network, data, and programs. , software resources such as files, or sharing of modems, faxes, or printers. It can support sharing of hardware resources such as other equipment.

This server is loaded with an app related to the online commerce brokerage system and is connected to the buyer terminal device 300 and the seller terminal device 400 through a network, thereby managing the Internet homepage corresponding to the URL, and the terminal devices 300 and 400 ), information on the Internet homepage linked to the URL, transaction details, and transaction results can be output.

The brokerage service providing device 200 calculates an estimated price based on information entered by the seller 410 for a product that the buyer 310 intends to purchase in order to conclude a transaction between the buyer 310 and the seller 410. The supply price calculation device 100 for transmitting the calculated price information to the buyer terminal device 300 may be included. Let's learn more about the supply price calculation device 100 through the following drawings.

Figure 2 is a block diagram showing some components of the supply price calculation device 100 according to an embodiment.

Referring to FIG. 2 , the supply price calculation device 100 may include a storage unit 110, an evaluation information generation unit 140, a supply price calculation unit 150, and a buyer/seller matching unit 180, The storage unit 110 includes a buyer information storage unit 120 and a seller information storage unit 130, and the price calculation unit 130 includes a first supply price calculation unit 160 and a second supply price calculation unit 170. ) and the like.

The buyer information storage unit 120 may receive and store basic information about the buyer and purchase information about products that the buyer intends to purchase from the buyer, and may store evaluation information about past transaction details of the buyer.

Basic information about the buyer may receive and store basic information from a buyer who requests purchaser registration, and basic information about the buyer may include personal and personal information of the buyer.

Purchase information on a product that the buyer intends to purchase may include information on the specific name of the product the buyer wants to purchase, the quantity of the product, a transportation method, a requested delivery deadline, and various requirements that the buyer specifically desires.

In addition, the information registered in the buyer information storage unit 120 may store information about a product that a buyer prefers or mainly purchases or information about a seller of a preferred style. This information may be used as basic information when the buyer-seller matching unit 180 provides a recommended seller list to a buyer.

The seller information storage unit 130 may receive basic information from a seller requesting seller registration in the mediation service providing device 200 and store seller information about the seller. Specifically, the seller information storage unit 130 includes seller information including basic personal information and personal information about the seller, and seller sales information (name of the product, price per quantity, stock, transportation method) for the product sold by the seller. etc.) can be stored after being input from the seller.

In addition, when purchase information (product name, quantity, shipping conditions, etc.) for a product that a buyer intends to purchase is transmitted to the mediation service providing device 200 and stored in the buyer information storage unit 120, the mediation service providing device 200 A plurality of sellers registered in may identify corresponding information, indicate whether or not a product the buyer intends to purchase is available for sale, and generate quote information including specific price and quantity information. The information generated in this way may be stored in the seller information storage unit 130 .

In the process of registering seller information in the seller information storage unit 130 for the first time, the seller needs to input basic information such as financial statements of the seller's company. The registered seller information is reflected immediately whenever company information is changed regardless of whether or not a transaction has occurred in the mediation service providing device 200, and when a transaction occurs, the seller evaluation information for the seller generated by the evaluation information generator 140 is also It may be stored in the seller information storage unit 130 .

In addition, the information registered in the seller information storage unit 130 may store information on a seller's preferred sales product or information on a preferred buyer, and the collected information is later stored in the buyer/seller matching unit 180. ) can be used as basic information when providing recommended buyer information to a seller.

In addition, information registered in the seller information storage unit 130 may store information about a seller's preference or a buyer of a style. This information may be used as basic information when the buyer-seller matching unit 140 provides a recommended buyer list to a seller later.

The evaluation information generation unit 170 may generate evaluation information obtained by evaluating evaluation item targets set in advance based on buyers and sellers, and the generated evaluation information is transmitted to the second supply price calculation unit 170. This can be used as basic information for the second supply price calculation unit 170 to calculate the second supply price. A specific method for the second supply price calculator 130 to calculate the second supply price will be described later.

Evaluation information may include static evaluation information and dynamic evaluation information. Static information refers to evaluation information for a buyer or seller who has not yet used an online commerce brokerage service, and dynamic information means a buyer or seller who has not yet used an online commerce brokerage service. If there is a result using , it refers to the information evaluated on the evaluation items by the online commerce brokerage service provider or the counterparty.

Static information can be divided into buyer static information and buyer dynamic information based on the buyer, and dynamic information can be divided into seller static information and seller dynamic information based on the seller. Evaluation items for evaluating buyers and evaluation items for evaluating sellers are the same, but there are also different items according to their characteristics.

Buyer static evaluation information refers to evaluation information about buyers who do not yet have past transaction history, so the provider providing online commerce brokerage service can use the buyer information entered by the buyer for the evaluation items set in advance and the purchaser It refers to the information evaluated by analyzing the requested items when doing so.

Seller Static Evaluation Information Also, evaluation information for a seller who has not yet had a past transaction history. A provider that provides online commerce brokerage services based on seller information entered by the seller and information about what the seller generally requires when selling It means the information evaluated for the evaluation items set in advance.

Buyer dynamic evaluation information is based on the evaluation information evaluated by the provider providing the online commerce brokerage service for evaluation items based on the results of using the online commerce brokerage service, and the contents of the transaction between the seller and the buyer. It means the information evaluated for the evaluation item.

Seller dynamic evaluation information is based on the evaluation information evaluated by the provider providing the online commerce brokerage service based on the results of using the online commerce brokerage service, and the contents of the transaction between the seller and the buyer. It means the information evaluated for the evaluation item. In the following specification, for convenience of description, buyer dynamic evaluation information will be referred to as buyer evaluation information, and seller dynamic evaluation information will be referred to as seller evaluation information.

Evaluation items refer to items that must be observed during transactions between buyers and sellers, or items that include factors that can add or reduce costs depending on the progress of the transaction. However, since not all items that can affect the price of a product are the same due to the characteristics of a buyer who wants to purchase a product and a seller who wants to sell a product, they may include items with different properties.

Specifically, in the case of the seller, whether or not the seller successfully delivered the products to be sold according to the items requested by the buyer may be a main evaluation item. If the seller complies with the items requested by the buyer, no additional cost will be incurred, but if not, there is a high possibility that additional cost will be incurred.

Therefore, seller evaluation items refer to items that must be observed during transactions between buyers and sellers, or items that can be added or reduced in cost depending on the progress of the transaction. It can be divided into , , and transportation condition items.

The expiry date item is an item about whether the expiration date has been observed. If the product the buyer intends to purchase has an expiration date or expiration date, or the products sold by the seller have a shelf life of at least 70 days % or more.

The cargo packing status item means an item on whether or not the packing configuration of the online commerce brokerage service provider or the purchaser complied with the request. Specifically, if a buyer purchases products in bulk, whether a detailed packing list is provided, whether information on the number of boxes for each product and the number of items inside the box is provided, whether products are provided in the same box with the same expiration date for each product, pallet Including items on whether products are provided by dividing them into product units when packing in units, whether products are provided in a form that can be immediately stacked in a distribution center, whether or not other online commerce brokerage service providers or buyers comply with requests, etc. can do.

The item of transportation conditions means whether the goods are delivered to the place requested by the buyer or the online commerce brokerage service provider at the time of request, whether additional transportation costs other than the cost of the goods are incurred at the time of transportation request, and whether the goods are delivered on the specified date and time. Whether or not they have been transported may be included in the evaluation items.

The above-described seller evaluation items are only one embodiment of the present invention, but are not limited thereto, and additional seller evaluation items may be generated according to a request of an online mediation service provider or a buyer.

The seller evaluation items are evaluated as True / False by the buyer or the online brokerage service provider, and accordingly, information on the compliance rate of evaluation items can be created as well. For example, if seller A has made 10 transactions but fails to satisfy the conditions required for a specific item once, the compliance rate of the item may be evaluated as 90%. Such evaluation information is continuously collected whenever a transaction between a buyer and a seller is terminated, and the collected seller evaluation information on a seller may be stored in the seller information storage unit 130 . Seller evaluation items are evaluated as points, and score information may also be stored in the seller information storage unit 130 .

In the case of evaluation items for the buyer, since the buyer is in a position to pay costs differently from the seller, the buyer's evaluation items are determined based on how much additional cost is generated by the buyer's requirements. , Even if there are buyers who have made purchase requests for the same product in the same quantity, if there are many requirements for the purchased product and if it is difficult, you will have to spend more for that much, so the buyer evaluation items are determined taking these points into consideration.

Therefore, the buyer evaluation items can be divided into expiration date items, quantity error tolerance items for each product, product damage tolerance items, packing form requirements items, and delivery condition items. This is referred to as the required complexity, and as the required complexity increases, the cost incurred increases accordingly even if the same product is purchased. Therefore, the required complexity is different for each item, and when the buyer requests strict conditions based on the average condition for each item entered by the online brokerage service provider, the size of the required complexity increases, When a buyer requests a condition that is more relaxed than the average condition, the size of the required complexity decreases.

As for the expiration date item, buyers generally request the earliest product from the production date for each product, but if they do not, if they request only the same product, if they request a product produced at a specific point in time, or if they request a product produced after or before a certain point in time, When a product is requested, when the expiration date quantity for each product is specified, or when there are many other requests made by the buyer based on date, the complexity of the request for expiration date items increases.

Since it is impossible to 100% identify the quantity of items within the tolerance range for each product when large-scale provision is made, it is common to compensate the error part financially after the fact. However, if this is not the case and only 100% quantity matching is allowed and returns are requested in case of discrepancies, the complexity of the requirements for the tolerance range of quantity error per product increases.

Regarding the product damage tolerance items, if there is no allowable range for product damage during transportation, or if all products are not 100% normal, the total amount of return is requested, the complexity of the requirements for the product damage tolerance items increases.

As for the packing type requirements, the complexity of the packing type requirements increases when the buyer requests a detailed packing type, when a specific box is requested, when a label on the box or a separate work required for identification is required. will do

The delivery condition items include when there is a restriction on the shape of the pallet, when the delivery date and delivery location are requested in detail, when the date and time specified as the delivery date is short-lived, and when the product is delivered to the place designated as the delivery location. In this case, if the transportation cost increases compared to the general transportation cost, the complexity of the requirements for the delivery condition item increases.

The above-described purchase evaluation items are only one embodiment of the present invention, but are not limited thereto, and additional evaluation items may be generated at the request of an online mediation service provider or a purchaser.

Buyer evaluation items are evaluated as True / False by the buyer or online brokerage service provider, and accordingly, information on the compliance rate of evaluation items can be created as well. Such buyer evaluation information is continuously collected whenever a transaction between a buyer and a seller is terminated, and the buyer evaluation information may be stored in the buyer information storage unit 120 . When the buyer evaluation items are evaluated as points, score information may also be stored in the buyer information storage unit 120 .

The evaluation of buyer evaluation items and the evaluation of seller evaluation items are simultaneously conducted by the provider, buyer, and seller who provide online commerce brokerage services. It can be used only when the supply price calculation unit 150 calculates the supply price. The supply price calculation unit 150 cannot utilize the dynamic evaluation information when the number of evaluations has not reached a preset minimum, and until then, static information input by the online commerce brokerage service provider is used to calculate the supply price.

The supply price calculator 150 may serve to calculate a price for a product supplied by a seller, and may include a first supply price calculator 160 and a second supply price calculator 170 .

The first supply price calculation unit 160 may calculate a price for a product that a buyer wants to purchase by reflecting prices suggested by a seller based on purchase information. Specifically, the first supply price calculation unit 160 may calculate the first supply price based on the price per unit price stored in the seller information storage unit 130 for the product requested by the buyer. The first supply price can be commonly used in the sense of purchase cost.

For example, when a buyer wants to purchase 1,000 units of a specific product X, the first supply price calculation unit 160 selects sellers selling product X from among sellers stored in the seller information storage unit 130. After the search, the first supply price is calculated based on the information on the selling price per product X stored in the seller information storage unit 130 by the sellers selling the product X. For example, if the first seller sells product X at 1,000 won per piece, the first supply price is calculated as the total amount for the first seller at 1,000,000 won, and if the second seller sells product X at 900 won per piece, the second seller For , the first supply price is calculated with a total amount of 900,000 won.

The second supply price calculation unit 170 calculates an expected correction cost that can actually be added or reduced based on seller evaluation information for the buyer's requested items for the purchase request product of the buyer, and calculates the estimated correction cost. The second supply price may be calculated by reflecting the first supply price.

In the case of business-to-business transactions, unlike transactions with general consumers, there are many cases where large-scale products are traded, so the cost incurred in purchasing the actual product is not only the initial purchase cost, but also various costs later such as management costs, selling costs, and maintenance costs. This extra happens a lot.

However, these additional costs vary depending on which seller you purchase from, and since buyers have different requirements for each product, if you calculate the total cost based only on the purchase price calculated by simply multiplying the quantity by the price per item, There was a problem that was very different from the total cost incurred.

Therefore, the second supply price calculation unit 170 according to an exemplary embodiment predicts that the first supply price corresponding to the purchase cost may actually be additionally generated or reduced based on the buyer's requested complexity and the seller's evaluation information. The correction cost is calculated, and the second supply price, in which the expected correction cost is reflected in the first supply price, is provided to the purchaser. Thus, there is an advantage in that the purchaser can be provided with information on costs that can be spent more realistically. The method for calculating the second supply price will be described in detail with reference to FIG. 3 below.

3 is a block diagram showing some components of a second supply price calculator according to an embodiment.

Referring to FIG. 3 , the second supply price calculator 170 according to an embodiment includes a request item determiner 171, a weight adjuster 172, a required complexity calculator 173, and an estimated correction cost calculator 174. ) and the like. In FIG. 3, for convenience of explanation, the components of the second supply price calculation unit 170 are shown separated into four parts, but the embodiment of the present invention is not limited thereto, and the four parts are one part, and the second supply price It may be implemented by the calculation unit 170.

The demand item setting unit 171 may analyze purchase request information for a product the buyer intends to purchase, and may determine and determine whether there are additionally requested items by the buyer in addition to items generally required.

The required items may include evaluation items described in the buyer evaluation items previously described with reference to FIG. 2 . A detailed description of this will be omitted as it has been described above.

The weight adjustment unit 172 sets weights for the purchaser's request items based on the purchaser's information stored in the purchaser information storage unit 120 for the purchaser's request items determined by the request item determination unit 171. do. For example, if a buyer made many demands on shipping conditions according to the buyer's past transaction information, the buyer can be classified as a buyer who requires strict conditions on shipping conditions, so the shipping conditions item has a high weight. grant Increasing the weight means that the expected cost for a specific item is calculated higher than in the general case.

Conversely, in light of the buyer's past transaction information, when the buyer has few or no requirements for the packing condition, a low weight may be assigned to the packing condition. The meaning of lowering the weight means that the expected cost of a specific item is calculated lower than in the general case.

The required complexity calculation unit 173 calculates the required complexity by evaluating the degree of the purchaser's request for each item requested by the purchaser.

Comparing the supplier's requirements with the general criteria set in advance for each requirement item, if more stringent conditions than the general requirements are required, the complexity of the requirements is calculated to be high. In this case, the required complexity is calculated low.

For example, among the delivery condition items, it usually takes 1 month for product A on the delivery date, but if 3 weeks is requested, the required complexity is calculated higher than the average, and if 2 weeks is requested, the required complexity is further increased. do. Conversely, if the delivery date is requested to be 2 months, the required complexity is calculated lower than the average.

The expected correction cost calculation unit 174 calculates an expected correction cost that can be added or reduced based on the information generated by the weight adjustment unit 172 and the required complexity calculation unit 173, and calculates the calculated expected correction cost as a first correction cost. The second supply price, which is the final supply price, can be calculated by reflecting it in the supply price.

Specifically, when a high weight is assigned to a specific item by the weight adjustment unit 172, the expected correction cost calculation unit 174 measures a price higher than a cost that can normally occur for that item, and measures a price equal to the difference. Calculate the expected correction cost. Conversely, when a weight is given to another specific item and as a result, a cost that is less than a normal cost for that item can be incurred, an estimated correction cost is calculated by the difference. In the former case, the estimated correction cost is calculated as an additional cost, but in the latter case, the estimated correction cost can be a cost that reduces the final cost.

In addition, the expected correction cost calculation unit 174 may normally occur for the item when the required complexity value is calculated higher than the average because the buyer requests a high condition for a specific item by the required complexity calculation unit 173. Measure a price that is higher than the actual cost, and calculate the estimated corrected cost by the difference.

Conversely, if the required complexity value is calculated lower than the average because the purchaser requires a low condition for another specific item, and the cost can be less than the cost that can normally be incurred for that item, the estimated correction cost is equal to the difference yields In the former case, the estimated correction cost is calculated as an additional cost, but in the latter case, the estimated correction cost can be a cost that reduces the final cost.

As another example, if the buyer sets the condition that 100% quantity matches as a demand item, the seller can charge the buyer for the entire inspection cost in addition to the supply price, so the cost of purchase is likely to increase, and in this case, the seller can Failure to reflect this is likely to lead to disputes. Therefore, the expected correction cost calculator 174 calculates the expected correction cost in consideration of these conditions, and then includes the calculated expected correction cost in the first supply price to calculate the second supply price, which is the final amount. .Buyers can estimate not only the first supply price, which is the cost of purchase, but also the second supply price, which includes possible future costs, so that they can efficiently manage costs, and disputes that may arise with the seller due to future costs In addition, there is an effect that can be prevented in advance.

In addition, the expected correction cost calculation unit 174 may calculate the expected correction cost based on evaluation scores for past transaction results of the seller. If the evaluation of the seller's past transaction results is high, the possibility of additional costs is low, so the estimated correction cost is calculated low, and the lower the evaluation score of the past transaction results, the higher the possibility of additional costs. Costs can be measured high.

In addition, the expected correction cost calculation unit 174 may calculate the expected correction cost based on evaluation scores for past transaction results of the seller for each item requested by the buyer in more detail. For example, when a buyer has set strict requirements for delivery deadlines, seller A does not keep the delivery deadlines well, so the evaluation score for the delivery deadline item is low, whereas seller B adheres to the delivery deadline items well. If the evaluation score for is high, the estimated correction cost for seller A can be calculated higher than the estimated correction cost for seller B, since there is a high possibility that more costs will be incurred due to delivery problems when purchasing goods from seller A. there is.

4 is a diagram showing a screen of a service provided with information on a first supply price and a second supply price according to an embodiment, and FIG. 5 is an item affecting a second supply price according to an embodiment. It is a diagram showing one screen of a service on which information about a group is displayed.

FIG. 4 shows a screen 320 displayed on the buyer's terminal device 300 when the buyer requests a supply price estimate for a quantity of 1,000 for 100ML of ABC cream.

Referring to FIG. 4 , information 330 on estimated prices for a total of four sellers (SP01, SP02, SP03, and SP04) has been calculated for a product that a buyer intends to purchase, and in FIG. 4, the first supply price is requested. It means the price obtained by multiplying the supply price per unit by the requested quantity by the quantity, and the second supply price is the price calculated by reflecting the expected correction cost to the first supply possibility.

Looking at the quotation request list, the first supply price of seller SP04 is the highest, but the second supply price reflecting the expected correction cost is calculated to be the lowest. can proceed

Therefore, since the supply price calculation device 100 according to an embodiment provides the buyer with information on the first supply price corresponding to the purchase cost and the second supply price reflecting the expected correction cost that may actually occur, the buyer has the advantage of being able to choose a seller that is more suitable for oneself.

In addition, the service screen provided to the buyer is provided with specific information on the requirements requested by the buyer as shown in FIG. 4, and when the detailed view 340 screen is clicked, as shown in FIG. 5, the second supply price The factors influencing it can be identified in detail. The quotation detailed information includes comprehensive information 370 on the second supply price, information 380 on expected demand fulfillment ratios for each item, and information 390 for expected correction costs for each item. Accordingly, the purchaser may change the purchaser's requested information with reference to such information and receive a new quotation again.

Also, the second supply price calculation unit 170 may calculate the expected correction cost using an artificial neural network. Hereinafter, a detailed description will be made through FIGS. 6 and 7.

6 and 7 are diagrams illustrating input information input to the second supply price calculation artificial neural network module and output information output according to an embodiment.

Referring to FIGS. 6 and 7 , the second supply angle calculation unit 170 may include a second supply angle calculation artificial neural network module 171, and the second supply angle calculation artificial neural network module 181 may include purchase information. (10), seller information 20, and seller evaluation information 30 as input information, and output information as expected correction cost 30 may include a pre-learned artificial neural network module. Accordingly, the artificial neural network module 181 for calculating the second feed angle may include an inference session 173 for inferring output information from input information and a learning session 172 for performing learning based on the input information and feedback information. .

The purchase information 10 may include information on the name, quantity, and requirements for the product the buyer intends to purchase, and the seller information 20 may include the sale product stored by the seller in the seller information storage unit 130. The seller evaluation information 30 may include information evaluated by an online brokerage service provider and a buyer who has performed a transaction in the past with respect to the above-described seller evaluation items.

Specifically, the artificial neural network module 171 calculating the second supply angle, as shown in FIG. 7 , each of the attributes constituting the purchase information 10, the seller information 20, and the seller evaluation information 30 is x1, x2 Calculation of the second supply angle, such as x3, is input to each node of the input layer of the artificial neural network module 171, and computation of hidden layers such as h1, h2, and h3 based on weights such as w1 Afterwards, the estimated correction cost 30 predicted based on a cost function such as softmax or ReLU can be output from an output layer of y1. The estimated correction cost 60 refers to information about costs that may occur when transacting with a seller entered as seller information 20 .

As shown in FIG. 9, the second supply angle artificial neural network module 171 generates an error (error, -Sigma(yi log pi)) based on the recommended seller information 60 and the actual transaction evaluation information 70 with the recommended seller. ), back propagation can be performed to update the weight of the hidden layer.

In performing feedback, it is possible to perform feedback using a loss function. The loss function of the artificial neural network module 171 for calculating the second supply angle is the expected correction cost 40 when a transaction is made with a specific seller and the actual occurrence The absolute value of the difference in cost 50 can be a loss function. Accordingly, the artificial neural network module 171 for calculating the second feed angle may update parameters of the artificial neural network module 181 for calculating the second feed angle so that the value of the loss function is minimized.

For example, if the estimated correction cost inferred by the artificial neural network module 171 for calculating the second supply angle based on the buyer A and the seller B is 1 million won, but the actual correction cost incurred is 700 thousand won, the second supply angle calculation artificial The parameters of the artificial neural network module 171 for calculating the second supply angle are updated in a direction in which the expected correction cost 60 inferred by the neural network module 171 is further increased, and conversely, when the actual correction cost is 1.2 million won, the second Parameters of the second supply angle calculation artificial neural network module 171 may be updated in a direction to decrease the expected correction cost 60 inferred by the supply angle calculation artificial neural network module 171 .

As described in FIGS. 7 and 8, when calculating the expected correction cost using the artificial neural network module in calculating the second supply price, the expected correction cost is automatically calculated using the previously learned artificial neural network structure, There is an effect that the service provider does not have to manually calculate each of the required complexity items described above.

8 and 9 are diagrams illustrating input information input to the buyer/seller matching artificial neural network module and output information output according to an embodiment of the present invention.

8 and 9 , the buyer/seller matching unit 180 may include a buyer/seller matching artificial neural network module 181, and the buyer/seller matching artificial neural network module 181 provides buyer information 60 and a pre-learned artificial neural network module having seller information 70 as input information and output information as expected evaluation information 80. Accordingly, the buyer/seller matching artificial neural network module 181 may include an inference session 183 that infers output information from input information and a learning session 182 that performs learning based on input information and feedback information.

Specifically, as shown in FIG. 9, each attribute constituting the buyer information 60 and the seller information 70 is an input layer of the buyer/seller matching artificial neural network module 181 such as x1, x2, and x3. ), and predicted evaluation information (80) predicted based on a cost function such as softmax or ReLU after computing hidden layers such as h1, h2, and h3 based on weights such as w1 It can be output from the output layer of y1.

Buyer information 60, which is input information, may include basic information about the buyer and evaluation information of the buyer, and seller information 70 also includes basic information about the seller, sales product information about the seller, and evaluation of the seller. information may be included. The expected evaluation information 80, which is output information, refers to a score for transaction expectation satisfaction inferred based on past transaction information when a buyer and a seller, which are standards of input information, actually transacted.

The buyer/seller matching artificial neural network module 181 is an artificial neural network module that infers the optimal seller based on the buyer and optimally infers the buyer based on the seller. That is, when the buyer/seller matching artificial neural network module 181 infers the optimal seller based on the buyer, information on a specific buyer and information on all sellers are respectively input into the buyer/seller matching artificial neural network module 181. After that, expected evaluation information 80 may be output by calculating an expected satisfaction score.

For example, in a situation where there are 10 buyers and 10 sellers, if you want to match the best seller based on buyer A, information on 10 sellers individually together with buyer information about buyer A, together with buyer/seller information. After the information is input to the matching artificial neural network module 181, a seller who has obtained the highest score among the inferred expected evaluation information may be connected to buyer A.

Conversely, if you want to match the optimal buyer based on X buyers, information on 10 buyers is individually entered into the buyer/seller matching artificial neural network module 181 along with seller information about X sellers, and then the inferred A buyer who has obtained the highest score among expected evaluation information may be connected to seller X.

On the other hand, the buyer / seller artificial neural network module 181 reduces the error (-Sigma (yi log pi)) based on the expected evaluation information 60 and the actual transaction evaluation information 70 as shown in FIG. In this direction, back propagation can be performed to update the weight of the hidden layer of the artificial neural network module.

In performing feedback, feedback can be performed using a loss function. The loss function of the buyer/seller matching artificial neural network module 181 is included in the estimated evaluation information 80 inferred when a transaction is made with a specific seller. It can be the difference between the score for expected satisfaction and the score for actual transaction satisfaction included in the actual transaction evaluation information. Accordingly, the buyer/seller matching artificial neural network module 181 may update parameters of the buyer/seller matching artificial neural network module 181 such that the value of the loss function is minimized.

So far, the configuration and characteristics of the present invention have been studied through various embodiments.

A method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same according to an embodiment provide a quoted price for a product to a buyer, not only the purchase cost but also the estimated correction cost that may occur in the future. Since it provides the supply price including the price, there is an effect of easily knowing the total amount that can actually be generated. Therefore, it is possible to efficiently manage costs associated with the purchase of goods, and there is an effect of preventing disputes due to costs that may occur due to unexpected reasons in advance.

In addition, a method for calculating a supply price of a seller using an artificial neural network and an apparatus for providing an online commerce brokerage service using the same according to an embodiment include a first supply price corresponding to a purchase cost to a buyer and a first supply price that reflects an estimated correction cost that may actually occur. 2Since information on the supply price is provided together, buyers have the advantage of being able to choose a seller that suits them more.

On the other hand, components, units, modules, components, etc. described in this specification described as “~ unit” may be implemented together or separately as interoperable logic devices. Depiction of different features for modules, units, etc. is intended to highlight different functional embodiments and does not necessarily mean that they must be realized by separate hardware or software components. Rather, functionality associated with one or more modules or units may be performed by separate hardware or software components or integrated within common or separate hardware or software components.

Although acts are depicted in the drawings in a particular order, it should not be understood that these acts need to be performed in the particular order shown, or in a sequential order, or that all depicted acts need to be performed to achieve a desired result. . In some circumstances, multitasking and parallel processing may be advantageous. Moreover, the division of various components in the above-described embodiments should not be understood as requiring such division in all embodiments, and the described components may generally be integrated together into a single software product or packaged into multiple software products. It should be understood that it is possible

Terminal device programs (also known as programs, software, software applications, scripts, or code) may be written in any form of programming language, including compiled or interpreted languages or a priori or procedural languages, and may be stand-alone programs or modules. , components, subroutines, or other units suitable for use in a terminal device environment.

Additionally, the logic flow and structural block diagrams described in this patent document describe corresponding actions and/or specific methods supported by corresponding functions and steps supported by the disclosed structural means. It can also be used to build software structures and algorithms and their equivalents.

The processes and logic flows described herein can be performed by one or more programmable processors executing one or more terminal device programs to perform functions by operating on input data and generating output.

The present description presents the best mode of the invention and provides examples to illustrate the invention and to enable those skilled in the art to make and use the invention. The specification thus prepared does not limit the invention to the specific terms presented.

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art or those having ordinary knowledge in the art do not deviate from the spirit and technical scope of the present invention described in the claims to be described later. It will be understood that the present invention can be variously modified and changed within the scope not specified.

Therefore, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification, but should be defined by the claims.

1: Online commerce brokerage system
100: supply price calculator
110: storage unit
120: buyer information storage unit
130: seller information storage unit
140: evaluation information generation unit
150: supply price calculation unit
160: first supply price calculation unit
170: second supply price calculation unit
180: buyer / seller matching unit

Claims (8)

Buyer information storage unit for registering and storing buyer information;
a seller information storage unit for registering and storing seller information and sales product information about products sold by the seller;
an evaluation information generating unit generating buyer evaluation information for the buyer and seller evaluation information for the seller;
a first supply price calculation unit that calculates a first supply price based on information registered in the seller information storage unit for the purchase request product of the buyer;
An expected correction cost that can be added or reduced is calculated based on the buyer's requested item for the purchase request product and the seller's evaluation information on the requested item, and the calculated expected correction cost is added to the first supply price. a second supply price calculation unit that calculates a second supply price by reflecting; and
A pre-learned buyer/seller matching artificial neural network module that takes the buyer information and the seller information as input information and outputs expected evaluation information including an expected satisfaction score when the buyer and the seller make a transaction. include
The second supply price calculator,
A pre-learned second supply price calculation artificial neural network module having purchase information on the product requested for purchase, sales information of the seller, and evaluation information of the seller as input information, and the expected correction cost as output information; ,
The second supply price calculation artificial neural network module,
Performing learning based on the difference between the expected correction cost and the actual correction cost;
The second supply price calculator,
Calculating the estimated correction cost based on evaluation scores of past fulfillment results of the seller for the buyer's requested item for the purchase request product;
The expected correction cost is calculated such that the higher the evaluation score of the past implementation result is, the lower the expected correction cost is, and the lower the evaluation score of the past implementation result is, the higher the expected correction cost is.
A device for providing online commerce brokerage service using an artificial neural network. According to claim 1,
The second supply price calculator,
Apparatus for providing an online commerce brokerage service using an artificial neural network, which calculates the expected correction cost after applying different weights according to preset criteria to the purchase request product and the buyer's request items for the purchase request product. According to claim 1,
The second supply price calculator,
Online commerce brokerage using an artificial neural network, which calculates the required complexity by evaluating the degree of demand by the buyer for each item requested by the buyer for the product requested for purchase, and calculates the expected correction cost in proportion to the size of the requested complexity. Service Provisioning Device. delete delete delete delete Registering and storing buyer information by a server;
registering and storing, by the server, seller information and product information for sale of the seller;
generating, by the server, buyer evaluation information for the buyer and seller evaluation information for the seller;
calculating, by the server, a first supply price based on information registered in a seller information storage unit for the purchase request product of the buyer;
calculating, by the server, an expected corrected cost that can be added or reduced based on the buyer's requested item for the purchase request product and the seller's evaluation information on the requested item;
calculating, by the server, a second supply price by reflecting the estimated correction cost to the first supply price; and
The server takes the buyer information and the seller information as input information, and the pre-learned buyer/seller matching artificial neural network having expected evaluation information including an expected satisfaction score when the buyer and the seller make a transaction as output information Including; outputting the expected evaluation information using a module;
In the step of calculating the expected correction cost,
Using a pre-learned second supply price calculation artificial neural network module having purchase information on the purchase requested product, sales information of the seller, and evaluation information of the seller as input information, and the expected correction cost as output information, Calculating an estimated correction cost; and
The second supply price calculation artificial neural network module performs learning based on the difference between the expected correction cost and the actual correction cost,
In the step of calculating the expected correction cost,
Based on the evaluation score of the seller's past performance results for the buyer's requested item for the purchase request product, the higher the evaluation score of the past performance result, the lower the expected correction cost, and the past performance result Calculating the expected correction cost so that the estimated correction cost increases as the evaluation score of is lower; further comprising
Seller's supply price calculation method using artificial neural network.