KR20240013689A - electronic device for generating comparative information and method thereof - Google Patents

Abstract

In the electronic device, the electronic device includes a memory, a communication unit, and a processor for processing data, wherein the processor processes big data in response to receiving a first input including a unique product number from a user terminal. , Based on processing the big data, identify product group information regarding the product type corresponding to the first input, and identify at least one product included in the identified product group information and corresponding to the first input. The degree of similarity between products may be determined, and comparison information may be generated between the first product corresponding to the first input and the product with the determined highest degree of similarity.

Description

Electronic device for generating comparative information and method thereof}

The disclosed invention relates to an electronic device and method for generating comparison information, and more specifically, to an electronic device and method for generating comparison information through accumulation of big data regarding home appliance information.

Recently, as the number of Internet users has rapidly increased, online products related to various events and promotions, including shopping mall products provided through various types of product sales sites, have become the mainstay, and as a result, purchases through online shopping malls are becoming more active.

In other words, the distribution channel was shortened to 'business-consumer', there were no restrictions on transaction area and transaction time, and the sales base was moved to a virtual space, making normal commerce possible without a sales space. For such e-commerce, large-scale shopping malls are thriving through the Internet or PC communication networks.

However, in the case of multi-item markets such as home appliances, it is an offline market where various types of set purchases are made, and due to the structure in which the final seller determines the price, consumers face limitations in efficient consumption due to limitations in information supply and demand. are facing Accordingly, accumulation of data on product type, price, basic product information, detailed product information, etc. is very important.

However, these specific products and the data related to them are manually entered separately or by manually parsing search results to classify the information. In other words, the conventional method of managing product data, etc. is that people manually input it separately and then manually visit and search multiple sites to match identifiers. This is inefficient, and not only does it cost a lot of money and time to process the work. , there is a problem that the possibility of mistakes due to manual input is always present.

Additionally, due to the nature of home appliances, even if they have the same shape and product name, there are versions with different functions. In order to compare information about these functions, consumers need to judge the presence or absence and difference of each function one by one, and there is a wasteful problem of having to directly find out the difference in function and the price difference arising from the difference in function. .

In addition, in the home appliance market where various types of set purchases are centered, there are users who lack experience in knowing what type of home appliances are appropriate to purchase, so the development of a recommendation system for home appliances for these users is required. It's a situation.

One aspect of the disclosed invention seeks to provide an electronic device and method for providing home appliance information.

Additionally, one aspect of the disclosed invention seeks to provide an electronic device and method for providing home appliance information by identifying information about a product disclosed to a product seller, extracting metadata, and generating big data.

In the electronic device for generating big data according to one aspect of the disclosed invention, the electronic device includes a memory, a communication unit for transmitting and receiving data to and from an external server, and a processor for processing the data, wherein the processor includes information received from the external server. Based on processing the data, product information can be identified, at least one product group information can be generated according to the product type included in the identified product information, and big data can be generated based on the product group information. .

Additionally, identifying the product information can identify a unique product number, product type, price, basic product information, and detailed product information based on processing data received from the external server.

Additionally, identifying the product information may identify the product information based on a product information identification model learned to identify product information based on at least one of an image or text.

Additionally, the processor may, in response to identifying the product information, determine whether an unidentified data item is present in the product information and, in response to the presence of the identified data item, drive a data replenishment process. there is.

Additionally, the data replenishment process may include transmitting the identified product information and the de-identified data item to an external server, and

It may include receiving unidentified data corresponding to the unidentified data item from the external server.

Additionally, the external server may execute a learned information supplementation model to output the identified data by taking the identified product information and the identified data item as input.

Additionally, the processor may update the big data based on the received unidentified data.

Additionally, the external server may include a first server operated by a product seller, a second server that outputs metadata about keywords, and a third server that executes an information supplementation model learned to output unidentified data. .

Additionally, the data replenishment process may include sending the product information to the second server, obtaining metadata about a keyword from the second server, and transmitting the metadata and the identifying data item to the third server. It may include transmitting to and receiving unidentified data corresponding to the unidentified data item from the third server.

Additionally, in response to receiving the identification data, the processor controls the communication unit to transmit the identification data to an administrator terminal, and in response to receiving an approval message from the administrator terminal, the processor controls the communication unit to transmit the identification data to the administrator terminal. The big data can be updated based on identification data.

In addition, the processor, in response to identifying the product information, in response to the price corresponding to the product information being less than the price corresponding to the product information stored in the big data, selects the product corresponding to the product information. The big data can be updated by identifying the lowest price information.

Additionally, in response to receiving quotation information from the user terminal, the processor may process at least one unique product number and the big data included in the quotation information to generate lowest price quotation information.

Additionally, the processor may control the communication unit to transmit a notification message including whether the quotation information is at an abnormal price to the user terminal based on processing the quotation information and the lowest price quotation information.

Additionally, in response to receiving sales information from a seller terminal, the processor processes at least one unique product number included in the sales information and the big data to generate price information for each unique product number corresponding to the seller terminal. can be created.

In addition, in response to identifying the product information, the processor accumulates prices of unique identification numbers corresponding to the product information to identify an average price corresponding to the unique identification number, and identifies the identified average price and the In response to the fact that the price difference corresponding to the identified product information is greater than a predetermined value, a notification message may be transmitted to the manager terminal.

In the big data generation method according to one aspect of the disclosed invention, receiving data from an external server, identifying product information based on processing the received data, and products included in the identified product information. It may include generating at least one product group information depending on the type and generating big data based on the product group information.

A big data generation system comprising an external server that transmits data according to a request from an electronic device according to an aspect of the disclosed invention, and an electronic device that receives and processes data from the external server to generate big data, the electronic device identifies product information based on processing data received from the external server, generates at least one product group information according to the product type included in the identified product information, and based on the product group information Big data can be created.

In the electronic device for generating comparison information according to one aspect of the disclosed invention, the electronic device includes a memory, a communication unit, and a processor for processing data, wherein the processor receives a first input including a unique product number from a user terminal. In response to processing the big data, and based on processing the big data, identify product group information regarding the product type corresponding to the first input, and at least one product group information included in the identified product group information. The similarity between the product and the product corresponding to the first input may be determined, and comparison information may be generated between the first product corresponding to the first input and the product with the highest determined similarity.

In addition, the similarity may be determined based on at least one of a first value related to the price difference between products, a second value related to the number of different information among basic product information, or a third value related to the number of different information among detailed product information. You can.

In addition, the similarity is determined by assigning different weights to the first value, the second value, and the third value, and the different weights are scaled in the order of the third value, the second value, and the first value. may be determined based on a preset weight that becomes smaller.

In addition, the processor, in response to receiving a second input including a unique product number different from the first input from the user terminal, processes big data to create a product corresponding to the first product and the second input. 2 Comparison information between products can be created.

Additionally, the comparison information may be generated based on a price difference between the first product and the second product, other information among basic product information, and other information among detailed product information.

In addition, the processor generates the comparison information based on a natural language processing model that outputs a description phrase by inputting the price difference between the first product and the second product, other information among the basic products, and other information among the detailed product information. You can.

In addition, the processor transmits the wish list information to the seller terminal in response to receiving wish list information including at least one unique product number from the user terminal, and sends a message corresponding to the wish list information from the seller terminal. In response to receiving the quotation information, the quotation information may be transmitted to the user terminal.

Additionally, the processor may identify whether there is a match between at least one unique product number corresponding to the wish list information and at least one unique product number corresponding to the quotation information.

In addition, identifying whether the unique identification number matches can be identified depending on the product type corresponding to the wish list information and the quotation information.

In addition, the processor may respond to the fact that the unique identification number does not match according to the product type corresponding to the wish list information and the quotation information, and generate comparison information corresponding to the unmatched unique identification number. .

Additionally, the comparison information may include comparison list information regarding at least one of basic product information or detailed product information between the first product and the second product.

In addition, the processor transmits list information according to the similarity to the user terminal based on processing the determined similarity and the identified product group information, and receives from the user terminal a product related to one of the list information according to the similarity. In response to receiving the second input, big data may be processed to generate comparison information between the first product and the second product corresponding to the second input.

In addition, in response to receiving wish list information including at least one unique product number from the user terminal, the processor performs a comparison between the products corresponding to each unique product number and the products with the highest degree of similarity. A comparison information list containing information can be created.

In a method for generating comparative information according to an aspect of the disclosed invention, in response to receiving a first input including a unique product number from a user terminal, processing big data, based on processing the big data, Identifying product group information regarding the product type corresponding to the first input; Determining a degree of similarity between at least one product included in the identified product group information and the product corresponding to the first input; and It may include generating comparison information between the first product corresponding to the input and the product with the determined highest similarity.

A comparison information generating system comprising an external server that transmits data according to a request from an electronic device according to an aspect of the disclosed invention, and an electronic device that receives and processes data from the external server to generate comparison information, the electronic device Processes big data in response to receiving a first input including a unique product number from a user terminal, and based on processing the big data, a product group related to the product type corresponding to the first input. Identifying information, determining a degree of similarity between at least one product included in the identified product group information and a product corresponding to the first input, and determining the degree of similarity between the first product corresponding to the first input and the determined degree of similarity is the highest. Comparison information between products can be created.

An electronic device for generating recommendation list information according to an aspect of the disclosed invention includes a memory for storing big data, a communication unit, and a processor for processing the big data, wherein the processor receives quotation information and the quotation information from a seller terminal. In response to receiving buyer information corresponding to, update the big data, and in response to receiving user information from a user terminal, generate recommendation list information based on processing the big data and the user information. You can.

In addition, updating the big data includes, based on processing the quotation information, identifying at least one unique product number included in the quotation information, and updating the buyer information for each of the identified unique product numbers. You can.

Additionally, the recommendation list information may be generated based on a recommendation model learned to input the user information and output recommendation list information corresponding to the user information.

Additionally, the recommendation model may be learned by inputting at least one unique product number included in the quotation information and the buyer information corresponding to the unique product number.

In addition, the processor may, in response to receiving the user information from the user terminal, determine whether a null value exists in the user information, and calculate a replacement value for the null value in response to the presence of the null value. there is.

Additionally, the recommendation list information may be generated by inputting the user information and the replacement value into the recommendation model.

Calculating the replacement value includes analyzing the correlation between items of the buyer information based on the buyer information, selecting the item with the highest correlation influence with respect to the item having the null value, and adding to the selected item. Based on this, the first replacement value can be calculated.

Additionally, the processor may select an item with the second highest correlation influence with respect to the item having the null value, and calculate a second replacement value based on the selected item.

Additionally, the processor inputs the user information and the first replacement value into the recommendation model to generate first recommendation list information, and inputs the user information and the second replacement value into the recommendation model to create a first recommendation list. 2 Recommended list information can be created.

In addition, the processor, based on processing the first recommendation list information, the second recommendation list information, and big data, selects at least one same product included in the first recommendation list information and the second recommendation list information. Comparison list information including comparison information according to type can be created.

Additionally, the buyer information or the user information may include at least one of data on purchase reason items, residence type items, living environment items, preferred brand items, purchased product type items, lifestyle habits, and family items.

In the method of generating recommendation list information according to one aspect of the disclosed invention, in response to receiving quotation information and buyer information corresponding to the quotation information from a seller terminal, updating big data and user information from a user terminal. In response to receiving, it may include generating recommendation list information based on processing the big data and the user information.

In the recommendation list information generating system including an external server that transmits data according to a request from an electronic device according to an aspect of the disclosed invention, and an electronic device that receives and processes data from the external server to generate recommendation list information, The electronic device updates the big data in response to receiving quotation information and buyer information corresponding to the quotation information from a seller terminal, and in response to receiving user information from a user terminal, updates the big data and the user. Recommended list information can be created based on the information processed.

According to one aspect of the disclosed invention, an electronic device and method for providing home appliance information can be provided.

In addition, according to one aspect of the disclosed invention, an electronic device and method for providing home appliance information can be provided by identifying information about a product disclosed to a product seller, extracting metadata, and generating big data. .

Figure 1 is a diagram schematically showing the overall configuration of a home appliance information provision system according to an embodiment.
Figure 2 is a block diagram for explaining the configuration of a home appliance information providing system according to an embodiment.
Figure 3 is a table explaining basic product information according to one embodiment.
Figure 4 is a table explaining detailed product information according to an embodiment.
FIG. 5 is a conceptual diagram illustrating how an electronic device identifies product information according to an embodiment.
FIG. 6 is a conceptual diagram illustrating output of list information according to the similarity of electronic devices according to an embodiment.
FIG. 7 is a conceptual diagram illustrating comparison information generated by an electronic device according to an embodiment.
FIG. 8 is a conceptual diagram explaining how an electronic device generates comparison information according to an embodiment.
FIG. 9 is a conceptual diagram explaining how an electronic device generates comparison information according to an embodiment.
FIG. 10 is a conceptual diagram illustrating comparison list information generated by an electronic device according to an embodiment.
FIG. 11 is a conceptual diagram illustrating a UI output by an electronic device to receive user information, according to an embodiment.
FIG. 12 is a conceptual diagram illustrating a UI output by an electronic device to receive user information, according to an embodiment.
FIG. 13 is a conceptual diagram illustrating a UI output by an electronic device to receive user information, according to an embodiment.
Figure 14 is a flowchart explaining a method of generating big data using an electronic device according to an embodiment.
Figure 15 is a flowchart explaining a method of producing big data using an electronic device according to an embodiment.
Figure 16 is a flowchart explaining a big data update method using an electronic device according to an embodiment.
Figure 17 is a flow chart to explain a big data update method using an electronic device according to an embodiment.
Figure 18 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.
Figure 19 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.
Figure 20 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.
Figure 21 is a flowchart for explaining a method of learning a recommendation model using an electronic device according to an embodiment.
Figure 22 is a flow chart to explain a method of generating recommendation list information using an electronic device according to an embodiment.

Like reference numerals refer to like elements throughout the specification. This specification does not describe all elements of the embodiments, and general content or overlapping content between the embodiments in the technical field to which the disclosed invention pertains is omitted. The term 'unit, module, member, block' used in the specification may be implemented as software or hardware, and depending on the embodiment, a plurality of 'unit, module, member, block' may be implemented as a single component, or It is also possible for one 'part, module, member, or block' to include multiple components.

Throughout the specification, when a part is said to be “connected” to another part, this includes not only direct connection but also indirect connection, and indirect connection includes connection through a wireless communication network. do.

Additionally, when a part "includes" a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

Throughout the specification, when a member is said to be located “on” another member, this includes not only cases where a member is in contact with another member, but also cases where another member exists between the two members.

Terms such as first and second are used to distinguish one component from another component, and the components are not limited by the above-mentioned terms.

Singular expressions include plural expressions unless the context clearly makes an exception.

The identification code for each step is used for convenience of explanation. The identification code does not explain the order of each step, and each step may be performed differently from the specified order unless a specific order is clearly stated in the context. there is.

Hereinafter, the operating principle and embodiments of the disclosed invention will be described with reference to the attached drawings.

Recently, as the number of Internet users has rapidly increased, online products related to various events and promotions, including shopping mall products provided through various types of product sales sites, have become the mainstay, and as a result, purchases through online shopping malls are becoming more active.

In other words, the distribution channel was shortened to 'business-consumer', there were no restrictions on transaction area and transaction time, and the sales base was moved to a virtual space, making normal commerce possible without a sales space. For such e-commerce, large-scale shopping malls are thriving through the Internet or PC communication networks.

However, in the case of multi-item markets such as home appliances, it is an offline market where various types of set purchases are made, and due to the structure in which the final seller determines the price, consumers face limitations in efficient consumption due to limitations in information supply and demand. are facing Accordingly, accumulation of data on product type, price, basic product information, detailed product information, etc. is very important.

However, these specific products and the data related to them are manually entered separately or by manually parsing search results to classify the information. In other words, the conventional method of managing product data, etc. is that people manually input it separately and then manually visit and search multiple sites to match identifiers. This is inefficient, and not only does it cost a lot of money and time to process the work. , there is a problem that the possibility of mistakes due to manual input is always present.

Additionally, due to the nature of home appliances, even if they have the same shape and product name, there are versions with different functions. In order to compare information about these functions, consumers need to judge the presence or absence and difference of each function one by one, and there is a wasteful problem of having to directly find out the difference in function and the price difference arising from the difference in function. .

In addition, in the home appliance market where various types of set purchases are centered, there are users who lack experience in knowing what type of home appliances are appropriate to purchase, so the development of a recommendation system for home appliances for these users is required. It's a situation.

The present invention may be intended to provide an electronic device and a control method thereof that can solve the above-described problems.

Figure 1 is a diagram schematically showing the overall configuration of a home appliance information providing system (hereinafter referred to as 'this system') according to an embodiment of the present disclosure.

Referring to Figure 1, this system 1000 includes a home appliance information providing electronic device 100 (hereinafter referred to as 'this device'), a server 200, a user terminal 300, a seller terminal 400, and an administrator terminal. It may include (500).

The system 1000 according to an embodiment of the present disclosure can obtain and process data existing in the server 200. In other words, the system 1000 can request data from the server 200 and obtain and process the requested data so as to obtain data that is continuously and/or periodically updated within the server 200.

In this case, the system 1000 can identify product information based on processing the acquired data. Here, the product information may include, for example, a unique product number, product type, product price, basic product information, and detailed product information. More specifically, a unique product number may refer to a unique and identifiable product number assigned to individual hardware or software. Additionally, product type may refer to the largest category that separates products. For example, the product type may refer to a product category such as a washing machine, dryer, refrigerator, etc., but is not limited thereto.

Meanwhile, as will be described in detail below, basic product information may be information that includes information about at least one item that home appliances may have in common, such as a rated voltage. That is, basic product information may be information including the weight of the product, rated voltage, etc. Additionally, the detailed product information may be information that includes information about at least one item that may exist depending on the characteristics of the type (category) of the product. By way of example, the detailed product information of the washing machine may include washing capacity, presence of bubbles, etc. In other words, detailed product information may include information about items that exist according to product characteristics.

Additionally, in response to processing data obtained from server 200 to identify product information, the system 1000 may drive a data replenishment process based on the product information. For example, the data supplementation process may be a process of obtaining unidentified data, that is, when a null value exists among the identified product information. However, the naming of the data supplementation process is not limited, and in another embodiment, the data supplementation process may be a process of acquiring metadata corresponding to product information.

Additionally, the system 1000 may generate at least one product group information based on the identified product information and generate big data based on the product group information. Here, product group information may be, for example, data generated by dividing identified product information according to product types corresponding to the product information, and multiple product group information may be generated depending on the number of product types. .

Additionally, in response to obtaining a first input including the user's unique product number, the system 1000 may generate comparison information between the first product corresponding to the first input and the product with the highest degree of similarity. there is.

In addition, the system 1000 receives and stores quotation information and buyer information corresponding to the quotation information, accumulates the buyer's purchase history according to the buyer information, and generates product recommendation list information according to the user information. there is.

The device 100, server 200, user terminal 300, seller terminal 400, and administrator terminal 500 of the system 1000 according to an embodiment of the present disclosure may be connected to the network 1. there is. Here, being able to be connected to the network 1 can be understood as being able to be connected electrically or communicationally.

An example of a network 1 includes a 3rd Generation Partnership Project (3GPP) network (3G network, 4G or Long Term Evolution (LTE) network, 5G or New Radio (NR) network, or 6G network, etc.), World Interoperability for WIMAX (WIMAX), etc. Microwave Access) network, Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), Bluetooth network, satellite broadcasting network, Includes, but is not limited to, analog broadcasting networks, DMB (Digital Multimedia Broadcasting) networks, etc.

The device 100 can transmit and/or receive data to the server 200. In more detail, the device 100 may transmit a request message requesting data transmission to the server 200 and receive data corresponding to the request message from the server 200. Accordingly, the device 100 can identify product information based on processing the received data.

Meanwhile, the data that the device 100 requests from the server 200 may include, for example, web page data and/or image data and/or text data. In another embodiment, the data may be, for example, metadata containing search results corresponding to specific keywords and/or output data of an artificial intelligence model. However, it is not limited to this.

The server 200 according to an embodiment of the present disclosure can process and store various data, and transmit and/or receive the processed data. More specifically, the server 200 may be connected to the device 100, the user terminal 300, the seller terminal 400, and the administrator terminal 500 through the network 1. Accordingly, the server 2000 can transmit and/or receive data for linking data with each other based on the network 1. The server 200 may be, for example, a platform and/or a web page cloud server, but is not limited thereto.

For example, the server 200 may be configured integrally with a database and searches for data corresponding to the request according to a request from the device 100, generates the requested specific information based on the retrieved information, and generates the generated information. Specific information can be transmitted back to the device 100.

The server 200 may exist as a single server or may be composed of multiple servers. For example, the server 200 provides the first server 201, which runs a web page operated by a product seller and/or an online shopping mall web page, etc., and metadata (e.g., search results) about keywords. It may include a second server 202 that (outputs) and a third server 203 that executes an artificial intelligence model. That is, the server 200 may be composed of a plurality of servers depending on the operator. Meanwhile, for convenience of explanation, the first server 201 is described below as a server that runs a web page operated by a product seller, but is not limited thereto, and sells products including online shopping mall web pages or sells products. It may include a server running an intermediary platform. By way of example, the third server 203 may be a server that executes an interactive artificial intelligence model such as GPT-3, BERT, Transformer-based chatbot, etc.

Meanwhile, the server 200 may include at least one processor that processes various data. More specifically, the processor of the third server 203 executing the artificial intelligence model according to an embodiment of the present disclosure inputs metadata (e.g., search results) and unidentified data items regarding a specific keyword. The learned classification model can be run to output de-identified data about de-identified data items of product information.

The user terminal 300, seller terminal 400, and manager terminal 500 according to an embodiment of the present disclosure include, for example, Personal Communication System (PCS), Global System for Mobile communication (GSM), and Personal Communication System (PDC). Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (WCode Division Multiple Access), Wibro (Wireless) It may include, but is limited to, all types of wired and wireless communication devices capable of input and output, such as Broadband Internet terminals, smartphones, SmartPads, tablet PCs, laptops, wearable devices, and digital signage. It doesn't work.

Meanwhile, in this disclosure, the term user refers to a user who uses the device 100, that is, a customer who receives services through the device 100 (for example, a consumer who wants to purchase a product or find out the price of the product). As such, it should be interpreted correctly according to the context for which this disclosure is intended. Meanwhile, in the present disclosure, the term administrator may be interpreted to mean a user (may refer to an institution/business, an operator, or a direct user) who manages or uses the device 100. Accordingly, the user terminal 300 in this disclosure may be interpreted to mean a customer's terminal.

Figure 2 is a block diagram for explaining the configuration of the present system 1000 according to one embodiment.

Referring to FIG. 2, the communication unit 110 of the device 100 according to an embodiment of the present disclosure can transmit and receive data with the server 200. For example, the communication unit 110 may transmit a request message requesting data to the server 200 at the request of the control unit 120 and may receive data received from the server 200.

The communication unit 110 may communicate with the server 200 and/or the user terminal 300 and/or the seller terminal 400 and/or the administrator terminal 500 through the network 1, and It may be composed of a tangible wireless communication module.

The control unit 120 may include a processor 121 and a memory 122.

The processor 121 may generate a signal that controls the communication unit 110 to transmit a message requesting data to the first server 201 operated by the product seller. In this case, the processor 121 may identify product information based on processing data received from the first server 201. More specifically, the data received from the first server 201 may include image data and/or text data (hereinafter referred to as 'web page data') about a web page on which a product can be launched.

In another embodiment, the processor 121 may connect to the first server 201, search a web page, and perform crawling to extract data from the page. More specifically, the processor 121 can access the web page through a URL for searching the web page in an Internet environment and download the contents of the web page. To this end, the processor 121 may transmit an HTTP request to the first server 201, receive a response from the first server 201, and receive a web page in HTML format. However, it is not limited to this.

Accordingly, the processor 121 can extract necessary information from the downloaded web page, and various methods can be used for such extraction. In other words, it can be acquired through an advertising information identification model, which will be described in detail below, and product information can be identified by analyzing the structure of the web page using parsing technology.

Figures 3 and 4 are tables for explaining basic product information and detailed product information, respectively, according to one embodiment.

The processor 121 according to an embodiment of the present disclosure may process the web page data in response to receiving the web page data, and identify product information based on the processed results. Product information may include, for example, a unique product number, product type, price, basic product information (31), and detailed product information (41).

More specifically, referring to FIG. 3, the basic product information that the processor 121 identifies based on processing the web page data includes, for example, the weight of the product, the size of the product, the color of the product, and the energy of the product. This may include consumption efficiency, power of the product, country of manufacture of the product, year of release of the product, etc. Basic product information may refer to data items other than data that have null values depending on the type of product (category, for example, washing machine, dryer, air conditioner, etc.). That is, as for basic product information, data items included in the basic product information may not vary depending on the type of product.

Meanwhile, referring to FIG. 4, the detailed product information identified by the processor 121 based on processing the web page data includes items of data in which null values occur depending on the product type and information corresponding to the items. It can mean. For example, referring to (a) of FIG. 4, when the product type is a washing machine, detailed product information includes washing capacity, presence or absence of a bubble function, presence or absence of a reservation function, presence or absence of a bubble soaking function, presence or absence of a light function, and number of rinses. It may include information such as the presence or absence of an adjustment function, the presence or absence of the spin speed control function, the presence or absence of the laundry addition function, the presence or absence of the button lock function, etc. That is, in the case of the presence or absence of additional laundry functions, which is a detailed product item for the washing machine, if the product type is an air conditioner, a null value for the item is inevitable, so the item may be an item included in the detailed product information for the washing machine. there is.

For example, referring to (b) of FIG. 4, when the product type is a dryer, detailed product information includes drying capacity, drying method, presence or absence of space dehumidification function, presence or absence of micro safety filter, presence or absence of auto-open door function, It may include information such as the presence or absence of a light function. However, detailed product information corresponding to at least two product types may include the same item. In other words, although dryers and washing machines are different product types, items such as the presence or absence of a light function may overlap. However, for example, in the case of an air conditioner, which is a different product type, items such as the presence or absence of a light function may not be included, so the light function may not be included. The item about presence or absence may be an item of detailed product information for the dryer and/or washing machine.

Meanwhile, detailed product information according to an embodiment of the present disclosure may include essential data items, for example. For example, in the case where the product type is a washing machine, unlike items such as the presence or absence of the smart control function of the washing machine, the items related to washing capacity may be important data items in the product type called washing machine, so they are classified as essential detailed information. It can be. As will be described in detail below, if the above-mentioned essential detailed information item exists as a null value, for example, the processor 121 may drive a data supplementation process.

Figure 5 is a conceptual diagram to explain how the device identifies product information according to an embodiment.

Referring to FIG. 5, the processor 121 includes a learning processor for machine learning and can train a model composed of an artificial neural network using training data. The artificial neural network learned here can be referred to as a learning model. A learning model can be used to infer a result value for new input data other than learning data, and the inferred value can be used as the basis for a decision to perform an operation.

More specifically, the processor 121 of the device 100 analyzes data based on raw data including image data and/or text data for a web page (here, raw data can be understood as learning data). and processing to build a machine learning model. To ensure reliability, a machine learning model can be built by using existing web page data as raw data. The machine learning model constructed in this way may be named, for example, a product information identification model. However, this is an example and is not limited thereto.

The product information identification model may be a model learned by inputting image data and/or text data for a web page and product information included therein. Meanwhile, here, the product information included in the web page may be the result of performing morphological analysis and data preprocessing, for example, through text mining of the text at the description position for the product included in the web page data. In another embodiment, the product information identification model, for example, takes web page data as input, extracts image or text data included in the web page data, and re-enters the extracted image or text data to provide product information. It may be a model trained to output.

The processor 121 may perform preprocessing on web page data to perform more efficient learning based on the acquired data. Here, preprocessing may involve the manager classifying data into predetermined categories. More specifically, classifying a predetermined category may mean classifying by product brand or product type. Web page data for each category classified in this way can be used to learn the neural network of the device 100.

For example, since the location and/or size of the web page on which the product is posted is different for each product brand, classifying product information by product brand in order to identify product information with greater accuracy is to select the web page for a specific product brand among the web page data. It may be classification based on data alone. In this case, the product information identification model can be learned separately according to the type of product brand. However, it is not limited to this.

Meanwhile, the device 100 can use an encoder and decoder of a neural network model. The apparatus 100 may input training data into at least one encoder and obtain at least one vector value as an encoded result. The encoder and decoder are types of neural network models, each of which is Deep Neural Network (DNN), Recurrent Neural Network (RNN), Long Short-Term Memory models (LSTM), Bidirectional Recurrent Deep Neural Network (BRDNN), and Convolutional Neural Network (CNN). Networks), etc., but is not limited to this.

In another embodiment, the processor of the device 100 may train a learning model through a deep learning algorithm.

Deep learning algorithm is one of the machine learning algorithms and refers to a modeling technique developed from artificial neural networks modeled after human neural networks. Artificial neural networks can be composed of multi-layered hierarchical structures.

An artificial neural network (ANN) includes an input layer, an output layer, and at least one intermediate layer (or hidden layer) (e.g., a kernel) between the input layer and the output layer. It may be structured in a hierarchical structure. Deep learning algorithms can ultimately produce highly reliable results through learning to optimize the weights of the activation function between layers based on this multi-layer structure.

A deep learning algorithm applicable to the processor of the device 100 according to an embodiment of the present disclosure may include, for example, a convolutional neural network (CNN). However, it is not limited to this. It can be understood that different deep learning algorithms may be applied depending on the embodiment.

Unlike techniques in which a learning process is performed by extracting knowledge from existing data, Convolutional Neural Network (CNN) is characterized by having a structure that extracts features of data and identifies patterns of features. The convolutional neural network (CNN) can be performed through a convolution process and a pooling process. In other words, the convolutional neural network (CNN) may include an algorithm composed of a complex convolutional layer and a pooling layer. Here, a process of extracting features of data (eg, a convolution process (convolution process)) is performed in the convolution layer. The convolution process is a process of identifying features by examining adjacent components of each component in the data and deriving the identified features into one piece. As a compression process, the number of parameters can be effectively reduced. In the pooling layer, a process (for example, a pooling process) is performed to reduce the size of the layer that has undergone the convolution process. The pooling process can reduce the size of data, cancel out noise, and provide consistent features in fine details. For example, the convolutional neural network (CNN) can be used in various fields such as information extraction, sentence classification, and face recognition. Meanwhile, since the convolutional neural network (CNN) is a known technology, detailed description will be omitted below.

Referring to FIG. 5, the processor 121 identifies basic product information 31, detailed product information 41, unique product number, and price based on the learned product information identification model and web page data 51. You can. In more detail, the processor 121 may identify product information 31 by inputting web page data 51 obtained through the communication unit 110 into the product information identification model 52. Product information 31 may include, for example, a unique product number, product type, price, basic product information 31, and detailed product information 41. However, it is not limited to this.

In summary, the processor 121 inputs web page data 51 including at least one of an image or text into the product information identification model 52, which is learned to identify product information based on at least one of an image or text. , information about product information can be identified.

Meanwhile, the processor 121 according to an embodiment of the present disclosure may generate at least one product group information according to the product type included in the identified product information in response to identifying product information. That is, the processor 121 accumulates identified product information based on processing web page data, and according to the product type included in the product information, in the form of metadata according to the same product type, the product type Product group information can be created depending on the number.

That is, the processor 121 may store big data generated based on at least one product group information in the form of a unique product number, price, basic product information, and detailed product information according to the same product type. However, it is not limited to this.

Meanwhile, in response to identifying product information, the processor 121 may determine whether an unidentified data item exists among the product information. In more detail, the processor 121 inputs web page data into an advertising information identification model and processes the output product information, including unique product number, product type, price, basic product information, and It is possible to determine whether at least one unidentified data item exists among detailed product information.

More specifically, the processor 121 identifies data items that are not secured among required detailed product information among the unique product number, product type, price, basic product information, and detailed product information included in the product information as unidentified data items. You can. In this case, processor 121 may drive a data replenishment process in response to the presence of an unidentified data item.

The processor 121 may obtain unidentified data corresponding to the unidentified data item obtained through the data supplementation process and update the big data based on the obtained unidentified data. That is, the processor 121 may update the identified data obtained for the unidentified data item and store the data for the unidentified data item.

Meanwhile, more specifically, the data supplementation process driven by the processor 121 according to an embodiment of the present disclosure transmits identified product information and an unidentified data item corresponding to the identified product information to the server 200. and receiving unidentified data corresponding to the unidentified data item from the server 200.

Here, the server 200 may mean, for example, a third server 203 that executes an artificial intelligence model. The third server 203 is an information supplement model trained to input an unidentified data item and product information excluding the unidentified data item and output unidentified data corresponding to the unidentified data item, or an unidentified data item and It may be a server that executes a learned information supplementation model to input metadata and output identified data corresponding to the identified data item.

In another embodiment, processor 121 may provide identified product information, more specifically, product information for which no data exists for an unidentified data item, and metadata about keywords (e.g., search results). 2 Can be transmitted to the server 202. In this case, the processor 121 may receive (obtain) metadata about search results corresponding to the product information from the second server 202.

In this case, the processor 121 may transmit an unidentified data item for which no data exists in the obtained metadata and advertisement information to the third server 203. Accordingly, the processor 121 may receive (obtain) unidentified data corresponding to the unidentified data item from the third server 203.

Meanwhile, the processor 121 may control the communication unit 110 to transmit the unidentified data to the manager terminal 500 in response to receiving (obtaining) the unidentified data from the third server 203. That is, the processor 121 transmits the identified data to the manager terminal 500 in order to verify the identified data transmitted by the third server 203, and receives the identified data from the manager terminal 500. In response to receiving an update approval message for big data, the big data may be updated based on the unidentified data. However, it is not limited to this.

The processor 121 may request web page data from the server 200 at predetermined intervals and identify product information based on processing the requested web page data. In this case, processor 121 may generate big data by running a data replenishment process in response to product information being identified. Here, the predetermined cycle may mean, for example, a data processing cycle of approximately 30 times per minute, but is not limited thereto.

The processor 121 according to an embodiment of the present disclosure, in response to identifying product information, determines that the price included in the product information is smaller than the price included in product information previously stored in big data, and determines that the product The big data can be updated by identifying the lowest price information for the product corresponding to the information.

More specifically, the processor 121 may request web page data from the server 200 at predetermined intervals and identify product information based on processing the requested web page data. In this case, the processor 121 determines that the unique product number included in the identified product information and the unique product number of the product information pre-stored in the big data are the same, and the price of the identified product information is set to the product information pre-stored in the big data. In response to being lower than the price, the big data may be updated with the price of the identified product information, and the price corresponding to the unique product number on the web page may be updated to the lowest price.

Meanwhile, in response to identifying product information, the processor 121 may store data by accumulating a unique identification number corresponding to the product information and information about the price corresponding to the unique identification number. Accordingly, the average price corresponding to the unique identification number can be identified. Accordingly, the processor 121 may transmit a notification message to the manager terminal 500 in response to the fact that the difference between the average price corresponding to the unique identification number and the price corresponding to the identified product information is greater than a predetermined value. .

In other words, the difference between the average price of products with the same unique identification number and the price corresponding to the recently identified product information is greater than a predetermined value when the web page data that identified the product information is incorrect and/or Since there is a possibility that advertising information has been incorrectly identified, this may be to provide an alarm about this to the administrator terminal. However, it is not limited to this. Meanwhile, the predetermined value here may be an error with the highest probability of occurrence of an error empirically or experimentally and/or a value set by an administrator.

Accordingly, the processor 121 may store or delete the identified advertising information based on the response to the notification message from the manager terminal 500.

Meanwhile, the processor 121 may receive sales information from the seller terminal 400. Here, the sales information may include at least one unique product number actually sold and a price corresponding to the unique product number. That is, sales information may mean, for example, information purchased by the buyer among the quotation information generated by the seller for the buyer.

In this case, the processor 121 may generate price information for each unique product number corresponding to the seller terminal 500 based on processing at least one unique product number and big data included in the sales information. there is. In other words, the processor 121 can obtain and store the price corresponding to the unique product number actually sold by the seller, in addition to the product information obtained by processing the web page data. Accordingly, the processor 121 obtains and accumulates information about the actual discount rate for each store by acquiring information about the price corresponding to the unique product number according to the store (seller terminal) sold in the actual offline store. You can.

Meanwhile, the processor 121 according to an embodiment of the present disclosure may receive quotation information from the user terminal 300 through the communication unit 110. Here, the quotation information may include, for example, information about price quotations corresponding to lists of home appliance products, and more specifically, includes at least one unique product number and information about the price corresponding to the unique product number. can do.

In this case, the processor 121 may process at least one unique product number and big data included in the quotation information in response to receiving quotation information from the user terminal 300 to generate the lowest price quotation information. . Since the big data stored in the memory 122 is updated to the lowest price according to a predetermined cycle by the processor 121, the processor 121 provides quotation information based on the price corresponding to the unique product number included in the big data. The lowest price quote information corresponding to at least one unique product number included can be generated.

The processor 121 according to an embodiment of the present disclosure processes the quotation information received from the user terminal 300 and the lowest price quotation information generated based on processing the quotation information and big data, The communication unit 110 can be controlled to transmit a notification message to (300). Here, the notification message may be, for example, a message that includes a unique product number included in the quotation information and information about whether the price corresponding to the unique product number is an abnormal price.

More specifically, the processor 121 determines that the total sum of prices corresponding to at least one unique product information included in the quotation information and the sum of prices corresponding to at least one unique product information included in the lowest price quotation information are predetermined. In response to being greater than the value, the quotation information may be determined to be an abnormal price. In other words, if the price included in the quotation information is greater than a predetermined value than the price included in the lowest price quotation information, the processor 121 may determine whether the quotation information is a bubble price that is higher than the actual lowest price in the industry. there is.

However, it is not limited to this, and in another embodiment, the processor 121 configures the communication unit 110 to send a notification message when the price included in the quotation information is less than a predetermined value than the price included in the lowest price quotation information. You can control it. That is, when a price in the quotation information that is less than the lowest price is identified, the processor 121 may determine whether the quotation information is a special price that is less than the industry's actual lowest price. Here, the notification message may be, for example, a message that includes a unique product number included in the quotation information and information about whether the price corresponding to the unique product number is a special price.

In response to the case where the unique product number corresponding to the identified product information is not stored in big data, the processor 121 according to an embodiment of the present disclosure provides a memory (memory) in which the product information corresponding to the unique product number is stored. 122) You can set the usage area. In more detail, the processor 121 may set one or more use areas for the allocated volume for the function of checking the status of the memory 122.

In addition, the processor 121 according to an embodiment of the present disclosure determines a diagnosis target by comparing the volume size of the total used area and the unused area of the memory 122, and performs periodic diagnosis on the determined diagnosis target. You can.

For example, the processor 121 may compare the total used area including the set used area with the unused area, and if the total used area is greater than the unused area, determine all areas to be diagnosed, and select the set used area. If the total used area included is smaller than the unused area, only the used area can be determined as the diagnosis target. However, it is not limited to this.

If the total used area is smaller than the unused area, for example, if the size of the unused area is larger than the minimum volume size, there is a high possibility that additional volumes will be registered (set to use) in the unused area, so check data in the unused area in advance. This may be done so that failures can be avoided even if a volume is set in an unused area (set in a used area).

On the other hand, in the opposite case, if the total used area is larger than the unused area, the possibility of registering additional volumes in the unused area is lower than in the opposite case, so for quick data check and load reduction, the unused area is subject to data check. Instead, only the area of the volume in use may be subject to data check. Meanwhile, in the processor 121 according to an embodiment of the present disclosure, for example, the administrator of the device 100 sets the minimum used area volume and compares the minimum used area volume with the unused area to determine the diagnosis target. It may be a decision. However, it is not limited to this.

That is, the processor 121 according to an embodiment of the present disclosure may periodically determine a target area for data check and periodically check data on the determined target area. By periodically checking data, the occurrence of a failure of the memory 122 when an unused area is used may be reduced.

FIG. 6 is a conceptual diagram illustrating output of list information according to the similarity of electronic devices according to an embodiment.

Referring to FIG. 6, the processor 121 according to an embodiment of the present disclosure may process big data in response to receiving a first input including a unique product number from the user terminal 300. Accordingly, the processor 121 can identify product group information about the product type corresponding to the first input, based on processing the big data. That is, the processor 121 can identify the product type of the unique product number corresponding to the first input and identify product group information regarding the same product type as the identified product type. Meanwhile, here, the product with the unique product number corresponding to the first input is named the first product 61 for convenience of explanation.

In this case, the processor 121 determines the degree of similarity between at least one product information included in the identified product group information and the first product 61 corresponding to the first input, and Comparison information between the product and the product with the highest degree of similarity determined can be generated. That is, the processor 121 according to an embodiment of the present disclosure generates product group information about the same product type as the product type of the first product 61 corresponding to the first input, based on processing big data. and generate comparison information with the product with the highest degree of similarity to the first product corresponding to the first input within the product group information.

The processor 121 according to an embodiment of the present disclosure, based on processing big data, provides a first value regarding the price difference between products of the same product type and a second value regarding the number of other information among basic product information. The similarity may be determined based on at least one of a value or a third value related to the number of other information among detailed product information.

That is, based on the first input obtained from the user terminal 300, the processor 121 determines the product type and price corresponding to the unique product number based on the unique product number and big data included in the first input. , identify basic product information and detailed product information, and identify product group information having the same product type as the first product 62 corresponding to the first input, wherein at least one product included in the product group information and the The degree of similarity between the first products 62 can be determined.

More specifically, the processor 121 may determine the degree of similarity between at least one product included in product group information having the same product type as the first product 61, respectively. The processor 121 calculates the similarity between a first value regarding the price difference between the first product 61 and the at least one product, a second value regarding the number of different information among the basic product information, and other information among the detailed product information. Determination may be made by assigning different weights to the third value regarding the number, and the different weights may be determined based on preset weights such that the sizes of the third value, the second value, and the first value decrease in that order.

As an example, the greater the price difference between the first product 61 and the product to be compared, the smaller the weight given to the first value may be, and the basic product information items between the first product 61 and the product to be compared. The greater the number of items with different data, the smaller the weight may be assigned, and the greater the number of items with different data on the detailed product information items between the first product 61 and the product to be compared, the smaller the weight may be assigned. .

As another example, in the case where there are two different basic product information items and two different detailed product information items between the first product 61 and the comparison target product, the processor 121 sets the second value to The weight assigned may be smaller compared to the weight assigned to the third value. In other words, the greater the price difference, the smaller the similarity between the first product 61 and the product being compared, and since consumers place high importance on the basic product information of home appliances, the weight given to the second value is greater than the third value. By giving it a smaller value, the degree of similarity may be determined. That is, the first value, second value, and third value may have an inverse relationship in which the larger the size, the smaller the similarity.

However, it is not limited to this, and in another embodiment, the sizes of different weights assigned to the second value and the third value may be the same or the order may be changed.

Accordingly, referring to (b) of FIG. 6, the processor 121 according to an embodiment of the present disclosure includes the first product 61 and product group information corresponding to the product type of the first product 61. The similarity between at least one included product may be determined, and the similarity list 63 may be output in order of highest similarity among the at least one product included in the product group information. The similarity list 63 may, by way of example, include a unique product number of the same product as the product type of the first product 61 and list information according to the similarity corresponding to the unique product number. However, it is not limited to this.

In more detail, referring to (a) of FIG. 6, the processor 121 according to an embodiment of the present disclosure, in response to obtaining the first input from the user terminal 300, Comparison information between the first product 61 corresponding to the first input and the product with the highest similarity among product group information can be transmitted. In another embodiment, the processor 121, in response to obtaining the first input from the user terminal 300, outputs a UI 62 for selecting a product to be compared to the user terminal 300. ) can generate a signal that controls. Accordingly, in response to obtaining the selection of the UI 62 from the user terminal 300, the processor 121 sends a signal to control the communication unit 110 to output the similarity list 63 from the user terminal 300. can be created.

In this case, in response to receiving a second input regarding one product included in the similarity list 63 from the user terminal 300, the processor 121 selects a second input corresponding to the first product and the second input. 2 Comparison information between products can be created.

FIG. 7 is a conceptual diagram illustrating comparison information generated by an electronic device according to an embodiment.

Referring to FIG. 7, the processor 121 according to an embodiment of the present disclosure generates big data in response to receiving a second input including a unique product number different from the first input from the user terminal 300. processing, and generate comparison information 72 between the first product corresponding to the first input and the second product 71 corresponding to the second input.

More specifically, the comparison information 72 may include, for example, the price difference between the first product and the second product, based on the processor 121 processing data about the first product and the second product, the basic It can be created based on other information among product information and other information among detailed product information.

Exemplarily, the processor 121 inputs the price difference between the first product and the second product, other information among the basic product information, and other information among the detailed product information, and determines the price difference between the first product and the second product. Comparison information can be generated based on a natural language processing model that has been learned to summarize other information and output it in one sentence.

In more detail, referring to (b) of FIG. 7, the processor 121 produces comparison list information (based on processing the unique product number for the first product, the unique product number for the second product, and big data) 74) can be created. The comparison list information 74 may be information generated to compare, for example, price, data on items included in basic product information, and items included in detailed product information. Looking at the example shown, you can see that the first product (left) and the second product (right) have the same data for items A, B, and C, and the difference in capacity is 1.5L and the difference in power consumption is You can see that is 100W.

In this case, the processor 121 may input product information about the first product and product information about the second product into a natural language processing model and output comparison information 72 about a one-line summary. According to the above-mentioned example, the processor 121 inputs product information about the first product and product information about the second product into a natural language processing model and outputs the "amount based on the difference in power consumption of 1.5L capacity and 100W. Comparison information 72 including a one-line summary such as "The difference is 45,000 won" can be obtained, and the comparison information 72 can be transmitted to the user terminal 300 for output.

Meanwhile, the processor 121 may transmit information about the image to the user terminal 300 so that the image about the first product and the image about the second product are output together.

In addition, for the user's convenience, the processor 121 provides a detailed view UI (UI) so that the comparison list information 74 can be checked in the comparison information 72 including a one-line summary, as shown in (a) of FIG. The communication unit 110 can be controlled to output 73) to the user terminal 300. In addition, as shown in (b) of FIG. 7, the communication unit ( 110) can be controlled.

FIG. 8 is a conceptual diagram explaining how an electronic device generates comparison information according to an embodiment. FIG. 9 is a conceptual diagram explaining how an electronic device generates comparison information according to an embodiment.

Referring to FIG. 8, the processor 121 according to an embodiment of the present disclosure allows a product to be selected in order to provide comparison information between products even when the first input is not received from the user terminal 300. , the communication unit 110 can be controlled to output a UI 81 for selecting a product type, and a UI for selecting a list according to the product type. Accordingly, the consumer can transmit an input for selecting a product type to the device 100 through the user terminal 300 to compare specifications between products.

In this case, when the processor 121 receives the selection of a product to be compared from the user terminal 300 through the UI shown in FIG. 8, the processor 121 may generate and output comparison information between the products to be compared.

In more detail, referring to FIG. 9, the processor 121 selects a product type and two corresponding product types through the UI 81 for selecting a product type in order to provide comparative information between products to the user terminal 300. In response to receiving the unique product numbers, comparison information between the two unique product numbers can be generated and output. FIG. 9 illustrates a comparison information UI that the processor 121 outputs in response to receiving a selection of a product type for a dishwasher from the user terminal 300. However, this is an example and is not limited thereto.

Meanwhile, the processor 121 according to an embodiment of the present disclosure, in response to receiving wish list information including at least one unique product number from the user terminal 300, sends the wish list to the seller terminal 400. Information can be transmitted. Here, the wish list information may include, for example, unique product numbers of home appliances that the consumer (user) wants to purchase.

In this case, the processor 121 may receive quotation information corresponding to the wish list information from the seller terminal 400. That is, based on the buyer's wish list information received from the device 100, the seller displays quotation information including the prices of products with unique identification numbers corresponding to the wish list through the seller terminal 400 ( 100). In this case, the processor 121 may transmit the quotation information to the user terminal 300 in response to receiving quotation information corresponding to the wish list information from the seller terminal 400.

Meanwhile, the product type corresponding to at least one unique identification number included in the quotation information transmitted by the seller terminal 400 and the wish list information transmitted by the user terminal 300 may be the same, but the quotation information includes The unique identification number included in the unique identification number and wish list information may be different. This is because there is a possibility that the product that the consumer wants to purchase is not included among the products being sold by the seller, so the seller arbitrarily selects a product that has the same product type (category) but can be replaced with the product included in the wish list. You can change and transmit quotation information with at least one unique product number that is different.

Accordingly, the processor 121 according to an embodiment of the present disclosure corresponds to at least one unique product number corresponding to the wish list information transmitted by the user terminal 300 and the quotation information transmitted by the seller terminal 400. It is possible to identify whether there is a match between at least one unique product number. In other words, identifying whether or not there is a match may mean that the processor 121 identifies whether or not the unique identification number is matched according to the product type corresponding to the wish list information and quotation information.

By way of example, the wish list information transmitted by the user terminal 300 includes the unique product number of product A related to the dishwasher and the unique product number of product B related to the washing machine, and the quotation information transmitted by the seller terminal 400. If it includes the unique product number of product A related to the dishwasher and the unique product number of product C related to the washing machine, the processor 121, based on processing the wish list information and quotation information, sets out the tableware of the same product type. It is possible to identify whether the washing machine is matched and whether the washing machine is matched. Accordingly, the processor 121 can identify that the unique identification number for the washing machine does not match. However, this is an example and is not limited thereto.

In response to the fact that the unique identification number does not match the product type corresponding to the wish list information and the quotation information, the processor 121 may generate comparison information between products corresponding to the unmatched unique identification number.

FIG. 10 is a conceptual diagram illustrating comparison list information generated by an electronic device according to an embodiment.

Referring to FIG. 10, in another embodiment, the processor 121 responds to receiving wish list information including at least one unique product number from the user terminal 300, each of the wish list information included in the wish list information. Based on the processing of products and big data corresponding to unique product numbers, products with the highest similarity to each unique product number can be identified.

That is, in response to receiving wish list information including a plurality of unique product numbers, the processor 121 identifies product group information corresponding to the product type of each of the plurality of unique product numbers, and identifies the plurality of unique products. Multiple products with the highest similarity for each number can be identified. Illustratively, the processor 121 responds to receiving wish list information including the unique identification number of product A and the unique identification number of product B from the user terminal 300, and selects the same product type as that of product A. Product group information corresponding to and product group information corresponding to the same product type as the product type of product B can be identified. Accordingly, the processor 121 can identify product A', which has the highest similarity to product A, and product B', which has the highest similarity to product B.

In this case, the processor 121 generates a comparison information list including comparison information between product A and product A' with the highest similarity and comparison information between product B and product B' with the highest similarity, and generates a comparison information list containing comparison information between product A and product A' with the highest similarity, and ) can be sent. Accordingly, consumers can obtain comparative information between products with high similarity to the product corresponding to the unique product number they have selected and make a more rational decision.

The processor 121 according to an embodiment of the present disclosure may update big data in response to receiving quotation information and buyer information corresponding to the quotation information from the seller terminal 400.

More specifically, the processor 121 receives quotation information including information on at least one unique product number sold by the seller to the buyer and the corresponding price and buyer information corresponding to the quotation information from the seller terminal 400. You can receive it. Buyer information may include, for example, information obtained by the seller through consultation with the buyer and/or information entered directly by the buyer. Specifically, buyer information includes purchase reason items, residence type items, and living environment items. , may include at least one of data on preferred brand items, purchased product type items, lifestyle items, and family items.

FIG. 11 is a conceptual diagram illustrating a UI output by an electronic device to receive user information, according to an embodiment. FIG. 12 is a conceptual diagram illustrating a UI output by an electronic device to receive user information according to an embodiment. FIG. 13 is a conceptual diagram illustrating a UI output by an electronic device to receive user information, according to an embodiment.

The processor 121 according to an embodiment of the present disclosure identifies at least one unique product number included in the quotation information based on processing the quotation information received from the seller terminal 400, and The buyer information can be updated for each unique product number. In other words, big data can be updated by updating each unique product number of buyer information generated based on quotation information on which a transaction was actually made. Meanwhile, here, buyer information, for example, has the same items as user information, but is named to distinguish whether at least one product has already been purchased. However, since it is the same in terms of data content, it can be used interchangeably. You can.

Accordingly, the processor 121 may generate recommendation list information based on processing the updated big data and the user information in response to receiving user information from the user terminal 300.

Referring to FIGS. 11 to 13 , the processor 121 may output a selection UI to the user terminal 300 to obtain user information.

For example, user information, like the above-mentioned purchaser information items, includes purchase reason items (e.g., moving, interior design, etc.), residence type items (e.g., apartment, villa, etc.), and living environment items (e.g. (e.g., square footage), preferred brand items (e.g., product brands), product type items purchased (e.g., at least one product type), lifestyle items (e.g., residence time at home), and family items (e.g. For example, it may include at least one of data about a four-person household, etc.).

In more detail, the processor 121 may control the communication unit 110 to output a selection UI ((a) of FIG. 11) for the purchase reason item among the user information to the user terminal 300. Accordingly, the user can transmit data on the purchase reason item to the device 100 by selecting the reason for purchasing the product through the user terminal 300.

Additionally, the processor 121 may control the communication unit 110 to output a selection UI ((b) in FIG. 11) for the residence type item among the user information to the user terminal 300. Accordingly, the user can select the type of residence through the user terminal 300 and transmit data on the known type item to the device 100.

As described above, the processor 121 may control the communication unit 110 to output a selection UI (FIGS. 11 to 13) for items corresponding to user information of the user terminal 300. Accordingly, the user can transmit data on items corresponding to user information to the device 100 by selecting the type of residence through the user terminal 300.

In response to receiving user information from the user terminal 300, the processor 121 may generate recommendation list information based on big data and processing the user information. More specifically, the processor 121 may generate recommendation list information based on a recommendation model learned by inputting quotation information received from the seller terminal 400 and buyer information corresponding to the quotation information.

The processor 121 according to an embodiment of the present disclosure provides raw data including quotation information received from the seller terminal 400 and buyer information corresponding to the quotation information (here, the raw data can be understood as learning data). Based on this, you can build a machine learning model by analyzing and processing data. To ensure reliability, a machine learning model can be built using quotation information and/or buyer information corresponding to the quotation information as raw data. The machine learning model constructed in this way may be named, for example, a recommendation model. However, this is an example and is not limited thereto.

The recommendation model may be a model learned by inputting at least one unique product number included in the quotation information and the buyer information corresponding to the unique product number. In other words, the recommendation model may be an artificial intelligence model learned to output a unique product number for a product to be recommended to a consumer by inputting buyer information. Accordingly, the processor 121 may input user information substantially the same as buyer information to the recommendation model and output at least one unique product number corresponding to the user information. In this case, the processor 121 recommends list information including at least one output unique product number and at least one product information corresponding thereto (e.g., price, product type, basic product information, detailed product information, etc.). can be created.

The processor 121 may perform preprocessing on quotation information and buyer information corresponding to the quotation information in order to perform more efficient learning based on the acquired data. Here, preprocessing may involve the manager classifying data into predetermined categories. More specifically, classifying a predetermined category may mean classifying data with the same data for a specific item included in the buyer information. Data for each category classified in this way can be used to learn the neural network of the device 100.

In summary, the processor 121 according to an embodiment of the present disclosure inputs user information received from the user terminal 300 into a recommendation model learned based on quotation information and corresponding buyer information, You can obtain printed recommendation list information.

Meanwhile, the processor 121 according to an embodiment of the present disclosure may determine whether a null value exists in the user information in response to receiving user information from the user terminal 300. Here, the presence of a null value in user information may mean that the user skips the item without selecting it through the user terminal 300, and thus there is no data for the corresponding item.

More specifically, the preprocessing process for data has a significant impact on improving the performance of prediction models using machine learning and deep learning. If the information in the data is inaccurate or the values are missing or ambiguous, no matter how large the amount of data is, Since it has no utility value, it requires the following series of purification and conversion operations. By way of example, the data preprocessing process includes data purification, data conversion and transformation, data reduction, and data cleaning. In other words, if the user information contains a null value and the user information containing the null value is input into the recommendation model, the accuracy of the output data is unlikely to be somewhat reduced, so artificial intelligence is used to find a replacement value for this null value. The accuracy of the model can be improved. This preprocessing process of obtaining replacement values can be seen as a part of data cleaning, for example.

The processor 121 according to an embodiment of the present disclosure can calculate a replacement value for a data item in which a null value exists.

Illustratively, calculating a replacement value for a null value uses representative values such as mean, variance, standard deviation, maximum value, minimum value, and quartile for the selected item. In this case, the processor 121 according to an embodiment of the present disclosure identifies the data mode for specific items of user information and buyer information as a replacement value based on processing big data, and sets the maximum value to It can be calculated as a replacement value.

For example, when the processor 121 identifies that a null value exists for the purchase reason item among the user information received from the user terminal 300, the processor 121 stores the user information and buyer information based on processing the big data. The mode of the data value for the purchase reason item can be calculated as a replacement value for the null value. Accordingly, the processor 121 may generate recommendation list information by inputting user information and the replacement value into the recommendation model. That is, in the case where the mode is newlyweds, the null value for the purchase reason item can be replaced with newlyweds and input into the recommendation model.

Meanwhile, in another embodiment, the processor 121 analyzes the correlation between items of the buyer information based on buyer information (same as user information, but pre-stored history information where purchases have been made), and selects items with null values. The item with the highest correlation influence can be selected, and a first replacement value can be calculated based on the selected item. Meanwhile, a plurality of methods may be used to evaluate the correlation between data values and/or items. Exemplarily, the processor 121 according to an embodiment of the present disclosure may analyze the correlation using at least one of a Pearson correlation coefficient, Spearman correlation coefficient, Kendall's rank correlation coefficient, and a correlation matrix. Meanwhile, the correlation analysis method is a known technique, and detailed description will be omitted below.

Exemplarily, the processor 121 may analyze the correlation between items included in the buyer information using at least one of the correlation analysis methods described above.

In this case, the processor 121 selects the item with the highest correlation influence with respect to the item with a null value through correlation analysis between items of the buyer information, and based on the data of the selected item, creates a first replacement value. can be calculated.

For example, in the case of the purchase product type item, when three or more products are selected, in the reason for purchase item, newlyweds are more suitable than moving. That is, since there is a correlation between the purchase product type and purchase reason items, the processor 121 can analyze the correlation between each item and each data type of the corresponding item.

Meanwhile, in another embodiment, the processor 121 may select an item with the second highest correlation influence with respect to an item with a null value and calculate a second replacement value based on the selected item.

Accordingly, the processor 121 replaces the null value with a first replacement value and inputs it into the recommendation model to generate first recommendation list information, and replaces the null value with a second replacement value and inputs it into the recommendation model to create a second recommendation list. Information can be generated. In order to provide recommendation list information that corresponds to the consumer's user information, a wider range of product groups can be recommended to the consumer by determining replacement values based on the item with the highest correlation influence and the item with the second highest correlation influence. .

Meanwhile, first recommendation list information generated based on the first replacement value and second recommendation list information generated based on the second replacement value may be the same. In this case, the processor 121 may control the communication unit 110 to provide only information about the first recommendation list information to the user through the user terminal 300.

In another embodiment, referring to FIG. 10, when the first recommendation list information and the second recommendation list information are different, the processor 121 processes the first recommendation list information, the second recommendation list information, and big data. Based on this, comparison list information can be generated that includes comparative information about two products where the product type included in the first recommendation list information and the product type included in the second recommendation list information are the same and have different unique product numbers. there is. As shown in FIG. 10, the first recommendation list information and the second recommendation list information include comparative information including products with different unique product numbers for all TVs, refrigerators, and washing machines, but are not limited to this. Alternatively, comparison list information containing only comparison information for one product type can be created.

The processor 121 may include a digital signal processor (DSP) and/or a micro control unit (MCU).

The memory 122 can store programs that perform the above-described operations and the operations described later, and the processor can execute the stored programs. In the case where there are multiple memories and processors, it is possible for them to be integrated into one chip, or they can be provided in physically separate locations. The memory may include volatile memory such as SRAM (Static Random Access Memory, S-RAM) and D-Lab (Dynamic Random Access Memory) for temporarily storing data. In addition, the memory 122 includes read only memory (ROM), erasable programmable read only memory (EPROM), and electrically erasable programmable read only memory (EEPROM) for long-term storage of control programs and control data. May include non-volatile memory. The processor may include various logic circuits and operation circuits, process data according to a program provided from memory, and generate control signals according to the processing results.

The method may be performed by the device 100 described above. Therefore, even if the content is omitted below, the content described with respect to the system 1000 and the device 100 can be equally applied to the description of the method.

Figure 14 is a flowchart explaining a method of generating big data using an electronic device according to an embodiment.

Referring to FIG. 14, the device 100 can identify product information based on processing data received from the external server 200 (S100).

Additionally, the device 100 may generate at least one product group information according to the product type included in the identified product information (S110).

Additionally, the device 100 may generate big data based on the product group information generated in step S110 (S120).

Figure 15 is a flowchart explaining a method of producing big data using an electronic device according to an embodiment.

Referring to FIG. 15, the device 100 can identify product information based on processing data received from the external server 200 (S100).

Additionally, the device 100 can determine whether unidentified data exists among the product information identified in step S100 (S101).

In this case, the device 100 may drive a data replenishment process in response to identifying unidentified data as present in step S101 (S102).

Meanwhile, the device 100 may generate at least one product group information according to the product type included in the identified product information in response to identifying that no identification data exists in step S101 (S110 ).

Additionally, in response to data supplementation being completed in step S102, the device 100 may generate at least one product group information according to the product type included in the identified product information (S110).

Additionally, the device 100 can generate big data based on product group information (S120).

Figure 16 is a flowchart explaining a big data update method using an electronic device according to an embodiment.

Referring to FIG. 16, the device 100 can transmit product information to the external server 200 (S200).

Additionally, the device 100 can receive metadata about product information from the external server 200 (S210).

Additionally, the device 100 may transmit the metadata and unidentified data items received in step S210 to the external server 200 (S220).

Additionally, the device 100 may receive unidentified data corresponding to an unidentified data item from the external server 200 (S230).

Additionally, the device 100 may update big data based on the unidentified data received in step S230 (S240).

Figure 17 is a flow chart to explain a big data update method using an electronic device according to an embodiment.

Referring to FIG. 17, the device 100 may transmit the unidentified data to the manager terminal 500 in response to receiving the unidentified data (S231).

Additionally, the device 100 may receive an approval message from the manager terminal 500 (S232).

Additionally, the device 100 may update big data based on the identified data in response to receiving an approval message from the administrator terminal 500 (S240).

Figure 18 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.

Referring to FIG. 18, the device 100 may receive a first input including a unique product number from the user terminal 300 (S300).

Additionally, the device 100 can identify product group information regarding the product type corresponding to the first input (S310).

Additionally, the device 100 may determine the degree of similarity between at least one product included in the identified product group information and the product corresponding to the first input (S320).

Additionally, the device 100 may generate comparison information between the first product corresponding to the first input and the product with the determined highest similarity (S330).

Figure 19 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.

Referring to FIG. 19, the device 100 may receive a first input including a unique product number from the user terminal (S300).

Additionally, the device 100 may receive a second input including a unique product number different from the first input from the user terminal (S340).

Additionally, the device 100 may generate comparison information between a first product corresponding to the first input and a second product corresponding to the second input (S350).

Figure 20 is a flow chart to explain a method of generating comparison information using an electronic device according to an embodiment.

Referring to FIG. 20, the device 100 may receive wish list information including at least one unique product information from the user terminal (S400).

Additionally, the device 100 can transmit the wish list information to the seller's terminal (S410).

Additionally, the device 100 may receive quotation information corresponding to the wish list information from the seller terminal (S420).

Additionally, the device 100 can identify whether there is a mismatch between the quotation information and the wish list information by product type (S430).

Additionally, the device 100 can generate comparison information between unique identification numbers corresponding to unmatched product types (S440).

Figure 21 is a flowchart for explaining a method of learning a recommendation model using an electronic device according to an embodiment.

Referring to FIG. 21, the device 100 can receive quotation information and buyer information from the seller terminal (S500).

Additionally, the device 100 can identify at least one unique product number included in the quotation information (S510).

Additionally, the device 100 may update the buyer information to the identified unique product number (S520).

Additionally, the device 100 can learn a recommendation model by inputting the identified unique product number and buyer information corresponding to the unique product number (S530).

Figure 22 is a flow chart to explain a method of generating recommendation list information using an electronic device according to an embodiment.

Referring to FIG. 22, the device 100 can receive user information from the user terminal (S600).

Additionally, the device 100 can identify whether a null value exists in user information (S610).

In this case, the device 100 may calculate a replacement value for the null value in response to the fact that the null value of the user information is identified in step S610 (S620).

Additionally, in response to the replacement value being calculated in step S620, the device 100 may input user information and the replacement value into the recommendation model (S630).

Meanwhile, the device 100 may input user information into the recommendation model in response to identifying that the null value of the user information does not exist in step 610 (S640).

Accordingly, the device 100 may generate recommendation list information output in step S630 or step S640 (S650).

Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium that stores instructions executable by a computer. Instructions may be stored in the form of program code, and when executed by a processor, may create program modules to perform operations of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

Computer-readable recording media include all types of recording media storing instructions that can be decoded by a computer. For example, there may be read only memory (ROM), random access memory (RAM), magnetic tape, magnetic disk, flash memory, and optical data storage devices.

As described above, the disclosed embodiments have been described with reference to the attached drawings. A person skilled in the art to which the present invention pertains will understand that the present invention can be practiced in forms different from the disclosed embodiments without changing the technical idea or essential features of the present invention. The disclosed embodiments are illustrative and should not be construed as limiting.

1000: main system
100: This device
110: communication unit 120: control unit
200: server
300: User terminal 400: Seller terminal
500: Administrator terminal

Claims (20)

Memory;
Ministry of Communications; and
Including a processor that processes data,
The processor,
In response to receiving a first input including a unique product number from the user terminal, processing big data,
Based on processing the big data, identify product group information regarding the product type corresponding to the first input,
Determine the degree of similarity between at least one product included in the identified product group information and a product corresponding to the first input,
Generating comparison information between the first product corresponding to the first input and the product with the highest determined similarity,
Electronic devices. According to paragraph 1,
The similarity is,
Determined based on at least one of a first value regarding the price difference between products, a second value regarding the number of different information among the basic product information, or a third value regarding the number of different information among the detailed product information,
Electronic devices. According to paragraph 2,
The similarity is,
It is determined by assigning different weights to the first value, the second value, and the third value, wherein the different weights are preset weights that decrease in size in the order of the third value, the second value, and the first value. determined based on,
Electronic devices. According to paragraph 1,
The processor,
In response to receiving a second input including a unique product number different from the first input from the user terminal, big data is processed to generate comparative information between the first product and the second product corresponding to the second input. generating,
Electronic devices. According to paragraph 4,
The above comparative information is,
Generated based on the price difference between the first product and the second product, other information among the basic product information, and other information among the detailed product information,
Electronic devices. According to clause 5,
The processor is
Generating the comparison information based on a natural language processing model that outputs a description phrase by inputting the price difference between the first product and the second product, other information among the basic products, and other information among the detailed product information,
Electronic devices. According to clause 6,
The processor,
In response to receiving wish list information including at least one unique product number from the user terminal, the wish list information is transmitted to the seller terminal,
In response to receiving quotation information corresponding to the wish list information from the seller terminal, transmitting the quotation information to the user terminal,
Electronic devices. In clause 7,
The processor,
Identify whether there is a match between at least one unique product number corresponding to the wish list information and at least one unique product number corresponding to the quotation information,
Identifying whether or not the matching occurs,
Identifying whether the unique identification number matches according to the product type corresponding to the wish list information and the quotation information,
Electronic devices. According to clause 8,
The processor,
In response to the fact that the unique identification number does not match according to the product type corresponding to the wish list information and the quotation information, generating comparison information corresponding to the unmatched unique identification number,
Electronic devices. According to paragraph 4,
The above comparative information is,
Containing comparison list information about at least one of basic product information or detailed product information between the first product and the second product,
Electronic devices. According to paragraph 3,
The processor,
Based on the determined similarity and processing the identified product group information, transmitting list information according to the similarity to the user terminal,
In response to receiving a second input regarding one product among the list information according to the similarity from the user terminal, big data is processed to generate comparison information between the first product and a second product corresponding to the second input. doing,
Electronic devices. According to clause 6,
The processor,
In response to receiving wish list information including at least one unique product number from the user terminal, comparison information including comparison information between the product corresponding to each unique product number and each product with the highest degree of similarity. creating a list,
Electronic devices. In the method of generating comparative information,
In response to receiving a first input including a unique product number from a user terminal, processing big data;
Based on processing the big data, identifying product group information about a product type corresponding to the first input;
determining a degree of similarity between at least one product included in the identified product group information and a product corresponding to the first input; and
Including; generating comparison information between a first product corresponding to the first input and a product with the determined highest similarity.
method. According to clause 13,
The similarity is,
Determined based on at least one of a first value regarding the price difference between products, a second value regarding the number of different information among the basic product information, or a third value regarding the number of different information among the detailed product information,
method. According to clause 14,
The similarity is,
It is determined by assigning different weights to the first value, the second value, and the third value, wherein the different weights are preset weights that decrease in size in the order of the third value, the second value, and the first value. determined based on,
method. According to clause 13,
The above method is,
In response to receiving a second input including a unique product number different from the first input from the user terminal, big data is processed to generate comparative information between the first product and the second product corresponding to the second input. Further comprising the step of generating,
method. According to clause 16,
The above comparative information is,
Generated based on the price difference between the first product and the second product, other information among the basic product information, and other information among the detailed product information,
method. According to clause 17,
The step of generating the comparison information is,
Generating the comparison information based on a natural language processing model that outputs a description phrase by inputting the price difference between the first product and the second product, other information among the basic products, and other information among the detailed product information,
method. A comparison information generation system comprising an external server that transmits data according to a request from an electronic device and an electronic device that receives and processes data from the external server to generate comparison information,
The electronic device is,
In response to receiving a first input including a unique product number from the user terminal, processing big data,
Based on processing the big data, identify product group information regarding the product type corresponding to the first input,
Determine the degree of similarity between at least one product included in the identified product group information and a product corresponding to the first input,
Generating comparison information between the first product corresponding to the first input and the product with the highest determined similarity,
system. A computer-readable recording medium on which a program capable of executing the method of claim 13 is recorded.

Images