KR102468827B1 - Method, device and system for sourcing products based on artificial intelligence and providing category-customized review analysis service - Google Patents

Abstract

According to one embodiment, a device can receive candidate products from a seller terminal of a seller, collect sales data for the candidate products sold in an online shopping mall, classify the candidate products into a category including at least one of fashion clothing, fashion miscellaneous goods, beauty, digital/home appliances, food, leisure, sports, maternity/parenting, and life/health, generate a competition index by extracting the number of registered sales registered products of the candidate products and the number of competitors of the candidate products based on the sales data, generate a review index by extracting the number of registered reviews in association with the candidate products, the length of the reviews, and the number of images included in the reviews from review data registered in association with the candidate products, generate an interest index by extracting the search number of the candidate products and the number of sales of the candidate products based on the sales data, generate a trend index by checking the degree of mention of the candidate product on social networks and the degree of mention in the community related to product sales, select recommended products for sale among the candidate product through a first artificial neural network based on the competition index, review index, interest index, and trend index, and transmit a guide message on the recommended products for sale to the seller terminal. In embodiments, the scores of the candidate products may be output through an artificial neural network and the prices of the recommended products for sale may be determined using the obtained scores and the expected sales volume of the recommended products for sale.

Description

Artificial intelligence-based product sourcing and category-customized review analysis service provision method, device and system

The following embodiments relate to technologies for providing product sourcing and category-customized review analysis services based on artificial intelligence.

BACKGROUND OF THE INVENTION Recently, with the development of Internet and network technologies, various information exchanges between users are being actively performed. In addition, along with the rapid increase in Internet users, users who purchase products through online shopping malls are also rapidly increasing.

Since there are countless products and product information registered to be sold in online shopping malls, technology that recommends products suitable for the user according to the user's taste and consumption pattern when a consumer enters a search term for the product they want to purchase has been developed

However, on the contrary, it is very cumbersome for the seller to analyze, select, and sell the product to be sold among numerous products. Therefore, it is required to develop a technology that analyzes a product through product sales data and review data and recommends a product to be sold to a seller.

Republic of Korea Patent Registration No. 10-1859055 (Announced on June 28, 2018) Republic of Korea Patent No. 10-2174717 (2020.11.05 announcement) Republic of Korea Patent Registration No. 10-2409717 (Announced on June 16, 2017) Republic of Korea Patent Registration No. 10-2135674 (2020.08.14 announcement)

Embodiments are intended to provide artificial intelligence-based product sourcing and category-customized review analysis services.

In the embodiments, the score of the candidate product is output through an artificial neural network, and the price of the recommended product for sale is determined using the obtained score and the estimated sales volume of the recommended product for sale.

Embodiments classify main keywords extracted from review data of recommended products for sale into positive keywords and negative keywords, and create improvements for recommended products for sale based on the classified keywords.

According to one embodiment, in a method performed by an apparatus, receiving a candidate product from a seller terminal of a seller; collecting sales data for the candidate product sold in an online shopping mall; classifying the candidate product into a category including at least one of fashion clothing, fashion accessories, beauty, digital/home appliances, food, leisure, sports, childbirth/parenting, and life/health; generating a competition index by extracting the number of registered sales products of the candidate product and the number of competitors of the candidate product based on the sales data; generating a review index by extracting the number of registered reviews associated with the candidate product, the length of the review, and the number of images included in the review from review data registered in association with the candidate product; generating an interest index by extracting the number of searches for the candidate product and the number of sales of the candidate product based on the sales data; generating a trend index by checking the degree of mention of the candidate product on social networks and the degree of mention in a community related to product sales; selecting a recommended product for sale from among the candidate products through a first artificial neural network based on the competition index, the review index, the interest index, and the trend index; and transmitting a guide message for the recommended product for sale to the seller terminal.

Selecting the recommended product for sale through the first artificial neural network may include applying the candidate product to an output node, a first value corresponding to the competition index, a second value corresponding to the review index, and the interest. Applying a third value corresponding to the index and a fourth value corresponding to the trend index to an input node, outputting a score of the candidate product using the first artificial neural network, and a score of the candidate product Based on the method, selecting the recommended product for sale may be included.

The method performed by the apparatus further includes determining a price of the product recommended for sale, wherein the step of determining the price comprises extracting a sales price of the product recommended for sale registered in the online shopping mall. calculating an average selling price of the product recommended for sale based on the extracted price, extracting the sales volume of the product recommended for sale registered in the online shopping mall, and recommending the product for sale based on the extracted sales volume Calculating the average sales volume of the product; Calculating the expected sales volume of the product recommended for sale by applying a first weight determined based on the score obtained through the first artificial neural network to the average sales volume; and and determining a price of the product recommended for sale by applying a second weight determined based on the calculated estimated sales volume.

In the method performed by the apparatus, when the seller has the recommended product for sale, transmitting a guide message about a product for sale of the recommended product for sale to the seller terminal based on a review of the recommended product for sale. Further including;, wherein the step of transmitting a guide message about the product for sale of the recommended product for sale to the seller terminal based on the review of the recommended product for sale held includes extracting text information of the review data of the recommended product for sale and extracting keywords, setting keywords mentioned more than a predetermined reference number of times from the extracted keywords as main keywords, classifying the main keywords as positive keywords or negative keywords, and selecting the positive keywords among the keywords. Calculating a positive rate, which is a percentage occupied by the keyword, calculating a negative rate, which is a percentage occupied by the negative keyword among the keywords, comparing the negative rate with a first standard value and a second standard value set in advance; If it is lower than the first reference value, transmitting a guide message recommending continued sales of the held recommended product for sale to the seller terminal; If the negative rate is higher than the first reference value and lower than the second reference value, the negative keyword arranging the negative keywords in the order in which most of the negative keywords were extracted, generating improvements of the stocked recommendation product for sale based on the sorted negative keywords, and transmitting a guide message about the negative keywords and the improvement points to the seller terminal. , When the negative ratio is higher than the second reference value, transmitting a guide message recommending changing the retained sales recommendation product to one of the products included in the sales recommendation product to the seller terminal.

The generating of the review index may include calculating a first positive rate, which is a rate occupied by positive keywords during a preset first period, and a first negative rate, which is a rate occupied by negative keywords, and a second rate after the first period. Calculating a second positive rate, which is a rate occupied by positive keywords, and a second negative rate, which is a rate occupied by negative keywords, for a period of time; Calculating a positive rate gradient, which is a difference between the first positive rate and the second positive rate; Calculating a negative ratio gradient, which is a difference between the first negative ratio and the second negative ratio, calculating a gradient score of the review data by summing the positive ratio gradient and the negative ratio gradient It may include generating a consistency weight for reviews of the product recommended for sale in inverse proportion to the gradient score, and updating the review index by applying the generated consistency weight to the review index. .

In the method performed by the device, determining whether the seller has experience of selling a product belonging to the same category as the category in which the recommended product for sale is classified; obtaining storage conditions of the product recommended for sale when the experience is not available; Obtaining conditions of a warehouse in which the seller stores products from the seller terminal; comparing storage conditions of the product recommended for sale with conditions of the warehouse; comparing the comparison result, generating a warehouse improvement point necessary for storing the recommended product for sale; Transmitting a guide message for the warehouse improvement point to the seller terminal; and transmitting a guide message for guiding storage conditions of the recommended product for sale to the seller terminal when the experience is present.

A method performed by the device, comprising: grouping buyers having the same lifestyle as that of the buyer and classifying them into buyer groups; selecting a product expected to be preferred by buyers belonging to the buyer group as a recommended product, based on the preference of each buyer belonging to the buyer group; comparing a released color of the product with a personal color of the purchaser, and selecting a color expected to match the purchaser's personal color among the released colors of the product as a recommended color; selecting a store that is found to be located closest to the buyer's movement line among stores that sell the product, as a recommended store, based on the buyer's movement pattern; setting a consumption prediction period in which future consumption is expected to be made for the type of store of the store, based on the consumer's consumption pattern, if the type of store is confirmed; and providing recommendation information on the product, the color, and the store to the purchaser's terminal when it is confirmed that the consumption prediction time has arrived.

The step of providing recommendation information on the product, the color, and the store to the buyer's terminal may include: checking whether there is a payment method with benefits when making a payment at the store; and providing recommendation information on the specific payment method to the purchaser's terminal if it is determined that there is a benefit when paying with a specific payment method at the store.

An apparatus according to an embodiment may be combined with hardware and controlled by a computer program stored in a medium to execute any one of the methods described above.

Embodiments may provide artificial intelligence-based product sourcing and category-customized review analysis services.

In embodiments, scores of candidate products may be output through an artificial neural network, and prices of recommended products for sale may be determined using the obtained scores and expected sales volume of the recommended products for sale.

In embodiments, main keywords extracted from review data of products recommended for sale may be classified into positive keywords and negative keywords, and improvement points for recommended products for sale may be generated based on the classified keywords.

1 is a diagram for explaining the configuration of a system according to an embodiment.
2 is a flowchart illustrating a process of providing a product sourcing and category-customized review analysis service based on artificial intelligence according to an embodiment.
3 is a flowchart illustrating a process of selecting a recommended product for sale according to an exemplary embodiment.
4 is a diagram for explaining a first artificial neural network according to an embodiment.
5 is a flowchart illustrating a process of setting a price of a product recommended for sale according to an exemplary embodiment.
6 is a flowchart illustrating a process of transmitting a guide message according to an indefinite ratio to a seller terminal according to an embodiment.
7 is a flowchart illustrating a process of generating a review index according to an embodiment.
8 is a flowchart illustrating a process of transmitting a guide message for guiding storage conditions of recommended products for sale to a seller terminal according to an embodiment.
9 is a flowchart illustrating a process of analyzing behavioral data according to an exemplary embodiment.
10 is a flowchart illustrating a process of providing recommendation guide information according to an exemplary embodiment.
11 is an exemplary diagram of a configuration of a device according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, since various changes can be made to the embodiments, the scope of the patent application is not limited or limited by these embodiments. It should be understood that all changes, equivalents or substitutes to the embodiments are included within the scope of rights.

Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only, and may be modified and implemented in various forms. Therefore, the embodiments are not limited to the specific disclosed form, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical spirit.

Although terms such as first or second may be used to describe various components, such terms should only be construed for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element.

It should be understood that when an element is referred to as being “connected” to another element, it may be directly connected or connected to the other element, but other elements may exist in the middle.

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

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

In addition, in the description with reference to the accompanying drawings, the same reference numerals are given to the same components regardless of reference numerals, and overlapping descriptions thereof will be omitted. In describing the embodiment, if it is determined that a detailed description of a related known technology may unnecessarily obscure the gist of the embodiment, the detailed description will be omitted.

The embodiments may be implemented in various types of products such as personal computers, laptop computers, tablet computers, smart phones, televisions, smart home appliances, intelligent vehicles, kiosks, and wearable devices.

An artificial intelligence (AI) system is a computer system that implements human-level intelligence, and unlike existing rule-based smart systems, machines learn and judge on their own. The more AI systems are used, the higher the recognition rate and the more accurate understanding of user preferences. Conventional rule-based smart systems are gradually being replaced by deep learning-based AI systems.

Artificial intelligence technology consists of machine learning and element technologies using machine learning. Machine learning is an algorithm technology that classifies/learns the characteristics of input data by itself, and element technology is a technology that uses machine learning algorithms such as deep learning to mimic functions such as recognition and judgment of the human brain. It consists of technical fields such as understanding, inference/prediction, knowledge expression, and motion control.

The various fields where artificial intelligence technology is applied are as follows. Linguistic understanding is a technology for recognizing and applying/processing human language/characters, including natural language processing, machine translation, dialogue systems, question and answering, voice recognition/synthesis, and the like. Visual understanding is a technology for recognizing and processing objects like human vision, and includes object recognition, object tracking, image search, person recognition, scene understanding, space understanding, image improvement, and the like. Inference prediction is a technique of reasoning and predicting logically by judging information, and includes knowledge/probability-based reasoning, optimization prediction, preference-based planning, and recommendation. Knowledge expression is a technology that automatically processes human experience information into knowledge data, and includes knowledge construction (data creation/classification) and knowledge management (data utilization). Motion control is a technology for controlling the autonomous driving of a vehicle and the movement of a robot, and includes motion control (navigation, collision, driving), manipulation control (action control), and the like.

In general, in order to apply machine learning algorithms to real life, learning is performed in a trial and error method due to the nature of the basic methodology of machine learning. In particular, deep learning requires hundreds of thousands of iterations. It is impossible to execute this in an actual physical external environment, so instead, the actual physical external environment is virtually implemented on a computer and learning is performed through simulation.

1 is a diagram for explaining the configuration of a system according to an embodiment.

First, a communication network may be configured regardless of its communication mode, such as wired or wireless, and may be implemented in various forms so that communication between servers and communication between servers and terminals is performed.

The seller terminal 10 may be a terminal used by a seller who wants to sell products according to the present invention. The seller terminal 10 may be a desktop computer, a laptop computer, a tablet, a smart phone, or the like. For example, as shown in FIG. 1 , the seller terminal 10 may be a smart phone and may be employed differently depending on the embodiment.

The seller terminal 10 may be configured to perform all or part of an arithmetic function, a storage/reference function, an input/output function, and a control function of a normal computer. The seller terminal 10 may be configured to communicate with the device 30 by wire or wireless.

The seller terminal 10 uses the device 30 to access a web page operated by a service provider or organization, or an application developed and distributed by a service provider or organization using the device 30 can be installed The seller terminal 10 may be linked with the device 30 through a web page or application.

The seller terminal 10 may access the device 30 through a web page or application provided by the device 30 .

Singular expressions in the claims may be understood to include the plural. For example, a seller in a claim may refer to one seller or more than one seller.

The buyer terminal 20 may be a terminal used by a buyer who wants to purchase a product according to the present invention. The purchaser terminal 20 may be a desktop computer, a laptop computer, a tablet, a smart phone, or the like. For example, as shown in FIG. 1 , the purchaser terminal 20 may be a smart phone and may be employed differently depending on the embodiment.

The purchaser terminal 20 may be configured to perform all or part of an arithmetic function, a storage/reference function, an input/output function, and a control function of a typical computer. The purchaser terminal 20 may be configured to communicate with the device 30 by wire or wireless.

The purchaser terminal 20 uses the device 30 to access a web page operated by a service provider or organization, or an application developed and distributed by a service provider or organization using the device 30 can be installed The purchaser terminal 20 may be linked with the device 30 through a web page or application.

The device 30 may be a server owned by a person or organization that provides services using the device 30, may be a cloud server, or may be a peer-to-peer (p2p) set of distributed nodes. may be The device 30 may be configured to perform all or part of an arithmetic function, a storage/reference function, an input/output function, and a control function of a typical computer.

Device 30 may be configured to communicate with the seller terminal 10 and the buyer terminal 20 by wire or wireless, control the operation of the seller terminal 10 and the buyer terminal 20, and the seller terminal 10 and the buyer It is possible to control which information is to be displayed on the screen of the terminal 20 .

Meanwhile, for convenience of description, only the seller terminal 10 and the buyer terminal 20 are shown in FIG. 1 , but the number of terminals may vary according to embodiments. As long as the processing capacity of the device 30 permits, the number of terminals is not particularly limited.

The device 30 may select a product recommended for sale using the first artificial neural network 100 . According to one embodiment, a database may be provided in the device 30, but is not limited thereto, and the database may be configured separately from the device 30. Apparatus 30 may include a number of artificial neural networks for performing machine learning algorithms.

Meanwhile, for convenience of explanation, only the seller terminal 10 and the buyer terminal 20 are illustrated in FIG. 1 , but the number of terminals may vary according to embodiments. As long as the processing capacity of the device 30 permits, the number of terminals and the number of two-wheeled vehicles are not particularly limited.

2 is a flowchart illustrating a process of providing a product sourcing and category-customized review analysis service based on artificial intelligence according to an embodiment.

Referring to FIG. 2 , first, in step S201, the device 30 may receive candidate products from the seller terminal 10 of the seller.

In this case, the candidate product may mean at least one or more products that the seller wants to sell. Also, the candidate product may refer to a product included in the selected category when the seller selects a product category desired to be sold.

In step S202, the device 30 may collect sales data for candidate products sold in the online shopping mall.

In this case, the sales data may include the number of registered sales registered products of the candidate product, the number of competitors of the candidate product, the number of searches for the candidate product, and the number of sales of the candidate product.

In step S203, the device 30 may classify candidate products into categories. In this case, the category may include at least one of fashion clothing, fashion accessories, beauty, digital/home appliances, food, leisure, sports, childbirth/parenting, and life/health.

In step S204, the apparatus 30 may generate a competition index by extracting the number of registered selling products of candidate products and the number of competitors of candidate products based on sales data.

According to one embodiment, the device 30 may extract the number of registered selling products of candidate products and the number of competitors of candidate products from sales data. The apparatus 30 may generate a competition index by summing the number of the extracted registered products for sale and the number of competitors.

In step S205, the device 30 may generate a review index from review data registered in association with the candidate product.

In this case, the review data may include the number of registered reviews associated with the candidate product, the length of the review, and the number of images included in the review.

According to one embodiment, the device 30 extracts the number of registered reviews associated with the candidate product, the length of the review, and the number of images included in the review from review data registered in association with the candidate product, and extracts the extracted reviews. The review index can be calculated by adding the number of reviews, the length of the review, and the number of images included in the review.

In step S206, the device 30 may generate an interest index by extracting the number of searches for candidate products and the number of sales of candidate products based on sales data.

For example, the device 30 may calculate the interest index by summing the number of searches for the candidate product and the number of sales of the candidate product.

In step S207, the device 30 may generate a trend index by checking the degree of mention of the candidate product on the social network and the degree of mention in the community related to product sales. In this case, the social network may include YouTube, Instagram, blog, cafe, Facebook, Twitter, and the like.

For example, the device 30 may generate a trend index based on the sum of the number of times the candidate product is mentioned on a social network and the number of times the candidate product is mentioned in a community related to product sales.

In step S208, the device 30 may select a product recommended for sale from candidate products through the first artificial neural network based on the competition index, review index, interest index, and trend index. At this time, a detailed description of the process of selecting the product recommended for sale will be described later with reference to FIGS. 3 and 4 .

In step S209 , the device 30 may transmit a guide message about recommended products for sale to the seller terminal 10 .

3 is a flowchart illustrating a process of selecting a recommended product for sale according to an exemplary embodiment.

Referring to FIG. 3 , first, in step S301, the device 30 may apply candidate products to an output node.

In step S302, the device 30 may apply the first value corresponding to the competition index, the second value corresponding to the review index, the third value corresponding to the interest index, and the fourth value corresponding to the trend index to the input node. have.

The first input signal may include a first value corresponding to a competition index, a second value corresponding to a review index, a third value corresponding to an interest index, and a fourth value corresponding to a trend index. The first input layer 110 of the first artificial neural network 100 may include input nodes respectively applied to the first value, the second value, the third value, and the fourth value.

In step S303, the device 30 may output scores of candidate products using the first artificial neural network.

The first output layer 120 of the first artificial neural network 100 may include an output node generating an output value.

In step S304, the device 30 may select a recommended product for sale based on the score of the candidate product.

For example, the device 30 may select a product corresponding to a candidate product having the highest score as a recommended product for sale.

4 is a diagram for explaining the first artificial neural network 100 according to an embodiment.

The device 30 may select a product recommended for sale using the first artificial neural network 100 . The device 30 may apply the first value corresponding to the competition index, the second value corresponding to the review index, the third value corresponding to the interest index, and the fourth value corresponding to the trend index to the input node. The first input layer 110 of the first artificial neural network 100 may include input nodes respectively applied to the first value, the second value, the third value, and the fourth value.

The device 30 may generate a first output signal by applying the first input signal to the first artificial neural network 100 . The first output layer 120 of the first artificial neural network 100 may include an output node generating an output value.

The device 30 may output scores of candidate products using the first artificial neural network 100 . The device 30 may select a recommended product for sale based on the score of the candidate product.

5 is a flowchart illustrating a process of setting a price of a product recommended for sale according to an exemplary embodiment.

Referring to FIG. 5 , first, in step S501, the device 30 may extract a sales price of a recommended product for sale registered in an online shopping mall.

For example, the device 30 may extract the price of the product recommended for sale as 1000 won, 12000 won, and 14000 won.

In step S502, the device 30 may calculate the average selling price of the recommended product for sale based on the extracted price.

For example, when the prices of products recommended for sale are extracted as 1000 won, 12000 won, and 14000 won, the device 30 may calculate the average selling price of the product recommended for sale as 12000 won.

In step S503, the device 30 may extract the sales volume of recommended products for sale registered in the online shopping mall.

For example, the device 30 may extract the sales volume of products recommended for sale as 3000, 5000, and 7000.

In step S504, the device 30 may calculate the average sales volume of recommended products for sale based on the extracted sales volume.

For example, when the sales volume of products recommended for sale is extracted as 3000, 5000, and 7000, the apparatus 30 may calculate the average sales volume of products recommended for sale as 5000.

In step S505, the device 30 may calculate the estimated sales volume of recommended products for sale by applying a first weight determined based on the score obtained through the first artificial neural network 100 to the average sales volume.

In this case, the first weight may be generated as a value proportional to the score acquired through the first artificial neural network 100 . That is, the first weight may be set higher as the score obtained through the first artificial neural network 100 increases, and set lower as the score obtained through the first artificial neural network 100 decreases.

The device 30 may calculate the expected sales volume of the product recommended for sale by applying the first weight to the average sales volume.

In step S506, the device 30 may determine the price of the product recommended for sale by applying the second weight determined based on the estimated sales volume calculated to the average sales price.

In this case, the second weight may be generated as a value inversely proportional to the expected sales volume. That is, the device 30 may set the second weight lower as the expected sales amount increases, and set the second weight higher as the expected sales amount decreases.

6 is a flowchart illustrating a process of transmitting a guide message according to a negative ratio to a seller terminal according to an embodiment.

When the seller has a recommended product for sale, the device 30 may transmit a guide message about the product for sale of the recommended product for sale to the seller terminal 10 based on the review of the recommended product for sale.

Referring to FIG. 6 , first, in step S601, the device 30 may extract text information of review data of recommended products for sale and extract keywords.

In step S602, the device 30 may set a keyword mentioned more than a predetermined standard number of times from the extracted keyword as a main keyword. In this case, the reference number of times may be set differently according to embodiments.

In step S603, the device 30 may classify the main keyword as a positive keyword or a negative keyword.

In step S604, the device 30 may calculate a positive ratio and a negative ratio. In this case, the positive rate may mean a rate occupied by positive keywords among keywords, and the negative rate may mean a rate occupied by negative keywords among keywords.

In step S605, the device 30 may compare the negative ratio with a preset first reference value. In this case, the first reference value may be set differently according to embodiments.

In step S606, when the negative ratio is lower than the first reference value, the device 30 may transmit a guide message recommending continued sales of the recommended product for sale to the seller terminal 10.

In step S607, the device 30 may compare the negative ratio with a preset second reference value. At this time, the second reference value may be set differently according to the embodiment.

In step S608, when the negative ratio is higher than the first reference value and lower than the second reference value, the device 30 may transmit a negative keyword and a guidance message about improvement points to the seller terminal 10.

According to an embodiment, the device 30 may sort the negative keywords in the order in which many negative keywords are extracted, and generate improvement points for recommended products for sale based on the sorted negative keywords.

For example, when the negative keywords are 'weak durability' and 'poor color', the device 30 sorts the negative keywords in the order in which many negative keywords are extracted, and 'durability is improved' and 'color is improved' based on the negative keywords. It is possible to create improvement points for recommended products for sale.

In step S609 , when the negative ratio is higher than the second reference value, the device 30 may transmit a guide message recommending changing the recommended product for sale to one of the products included in the recommended product for sale to the seller terminal 10 .

7 is a flowchart illustrating a process of generating a review index according to an embodiment.

Referring to FIG. 7 , first, in step S701, the device 30 may calculate a first positive ratio and a first negative ratio for a preset first period. In this case, the first positive rate may mean a rate occupied by positive keywords during the first period, and the first negative rate may mean a rate occupied by negative keywords during the first period. Here, the first period may be set differently according to embodiments.

In step S702, the device 30 may calculate a second positive rate and a second negative rate for a second period after the first period.

In this case, the second positive rate may mean a rate occupied by positive keywords during the second period, and the second negative rate may mean a rate occupied by negative keywords during the second period. Here, the second period may be set differently according to embodiments.

In step S703, the device 30 may calculate a positive ratio gradient that is a difference between the first positive ratio and the second positive ratio.

In step S704, the device 30 may calculate a negative ratio gradient that is a difference between the first negative ratio and the second negative ratio.

In step S705, the device 30 may calculate a gradient score of the review data by summing the positive percentage gradient and the negative percentage gradient.

That is, the device 30 may calculate the gradation score of the review data based on the variance of the positive ratio for the first period and the second period and the variance of the negative ratio for the first period and the second period. have.

In step S706, the device 30 may generate a consistency weight for reviews of products recommended for sale in inverse proportion to the gradient score. At this time, the device 30 may set the consistency weight higher as the variance of the positive and negative keywords of the review data is smaller, and set the consistency weight lower as the variance of the positive and negative keywords is higher.

In step S707, the device 30 may update the review index by applying the generated consistency weight to the review index.

8 is a flowchart illustrating a process of transmitting a guide message for guiding storage conditions of recommended products for sale to a seller terminal according to an embodiment.

Referring to FIG. 8 , first, in step S801, the device 30 may determine whether or not the seller has experience of selling a product belonging to the same category as the category in which the recommended product for sale is classified.

In step S802, the device 30 may acquire the storage conditions of recommended products for sale if there is no experience.

For example, the device 30 may obtain refrigeration (0 to 10 °C) as the storage condition of the product recommended for sale.

In step S803 , the device 30 may obtain conditions of a warehouse in which the seller stores the product from the seller terminal 10 .

For example, the device 30 may obtain room temperature conditions (1 to 35°C) of a warehouse in which a seller stores products from the seller terminal 10 .

In step S804, the device 30 may compare storage conditions of recommended products for sale and warehouse conditions.

In step S805, the device 30 may compare the comparison result to generate a warehouse improvement point necessary for storing recommended products for sale.

For example, if the storage condition of the product recommended for sale is refrigeration (0 to 10 ° C) and the condition of the warehouse where the seller stores the product is room temperature (1 to 35 ° C), the device 30 is the product recommended for sale. It is possible to create a warehouse improvement that requires refrigeration (0-10 °C) conditions to be stored.

The device 30 may transmit a guide message about warehouse improvements to the seller terminal 10 . For example, the guide message for warehouse improvement points may include information on warehouse improvement points and items and equipment required to satisfy the warehouse improvement points.

In step S806, if there is experience, the device 30 may transmit a guide message for guiding storage conditions of recommended products for sale to the seller terminal 10.

9 is a flowchart illustrating a process of analyzing behavioral data according to an exemplary embodiment.

Referring to FIG. 9 , first, in step S901, the device 30 may analyze the buyer's preference based on the buyer's SNS data.

Specifically, the device 30 may analyze the preference of the purchaser by analyzing videos, images, texts, etc. included in posting materials uploaded to the purchaser's SNS account.

For example, when a number of pictures of clothes of a specific style are posted on the buyer's SNS account, the device 30 may analyze the clothes of the corresponding style as the buyer's preference, and upload a video of a specific makeup method to the buyer's SNS account. If many of these are posted, the corresponding makeup method can be analyzed according to the buyer's preference.

In step S902, the device 30 may analyze the purchaser's consumption pattern based on the purchaser's card payment details.

Specifically, the device 30 determines the location of the merchant using the card based on the purchaser's card payment history data, identifies in which region the purchaser mostly uses the card, and analyzes the purchaser's consumption pattern. can

In addition, the device 30 checks the types of merchants using the card, such as convenience stores, hospitals, pharmacies, shopping malls, and marts, based on the purchaser's card payment history data, and determines which types of merchants the buyer mostly used the card for. It is possible to understand and analyze the consumption patterns of buyers.

In addition, the device 30 can analyze the purchaser's consumption pattern by determining when the purchaser used the card most of the time, based on the purchaser's card payment detail data.

In step S903, the device 30 may analyze the buyer's movement pattern based on the buyer's location record data.

Specifically, based on the buyer's location record data, the device 30 analyzes the buyer's movement pattern by figuring out where the buyer visits, how long he stays at a certain point, and which movement he frequently moves to. can do.

In step S904 , the device 30 may collect the purchaser's preference, purchaser's consumption pattern, and purchaser's movement pattern, and analyze the purchaser's behavioral data.

That is, the device 30 may generate the buyer's behavior data by collecting all the results of analyzing the buyer's preference, the buyer's consumption pattern, and the buyer's movement pattern.

10 is a flowchart illustrating a process of providing recommendation guide information according to an exemplary embodiment.

Referring to FIG. 10 , first, in step S1001, the device 30 may group customers having the same lifestyle as that of the buyer and classify them into a buyer group.

For example, the device 30 may classify customers having the same lifestyle as that of the buyer into a buyer group.

In step S1002, the apparatus 30 may select a product expected to be preferred by customers belonging to the buyer group as a recommended product, based on the preferences of each customer belonging to the buyer group.

For example, when a second customer, a third customer, and a fourth customer belong to a buyer group, the device 30 determines a first list that is a product list that is expected to be preferred by the second customer based on the preference of the second customer. Extracts a second list, which is a product list that is expected to be preferred by the third customer through the preference propensity of the third customer, and a product list that is expected to be preferred by the fourth customer through the preference propensity of the fourth customer. A third list can be extracted, and as a result of comparing the first list, the second list, and the third list, if it is confirmed that the product is included in each of the first list, the second list, and the third list, the buyer group It can be expected that customers belonging to will prefer the product, and the product can be selected as a recommended product.

In step S1003, if the product is selected as a recommended product, the device 30 checks the product's release color, compares the product's release color with the buyer's personal color, and determines that the purchaser's personal color matches the purchaser's personal color among the product's release colors. An expected color may be selected as a recommended color.

For example, if the released colors of the product are the first color, the second color, and the third color, the device 30 analyzes whether the first color matches the buyer's personal color, and determines whether the second color matches the buyer's personal color. analyzes whether the third color matches the buyer's personal color, and compares the result of analyzing whether the product's launch color matches the buyer's personal color, and determines whether the first color matches the buyer's personal color. If it is analyzed to be most suitable, the first color may be selected as a recommended color.

In step S1004, if the product is selected as a recommended product, the device 30 identifies stores that sell the product, and based on the buyer's movement pattern, among stores that sell the product, is determined to be located closest to the buyer's movement line. The confirmed retailer may be selected as a recommended retailer.

For example, as a result of analyzing the buyer's movement pattern, if the movement pattern is analyzed as commuting from point A to point B, the device 30 determines the movement route from point A to point B and the locations of stores that sell products. is compared, and if it is confirmed that the location of the distribution point and the circulation moving from point A to point B are the closest, the sales point may be selected as the recommended sales point.

In step S1005, when the store is selected as a recommended store, the device 30 may check the type of business of the store. Here, the type of business may be divided into types of card-using merchants such as convenience stores, hospitals, pharmacies, shopping malls, and marts.

In step S1006, when the type of store is confirmed, the device 30 may set a consumption prediction period in which future consumption is expected to be made based on the purchaser's consumption pattern.

Specifically, the device 30 may identify the types of card-using member stores, such as convenience stores, hospitals, pharmacies, shopping malls, and marts, through the purchaser's card payment details, and analyze the purchaser's consumption pattern for each type of business.

For example, the device 30 may analyze the consumption pattern by visiting the convenience store once a day in the evening when it is confirmed that the card payment is made in the convenience store industry once a day on average every evening time, and on average Monday If it is confirmed that the card payment was made in the mart sector every month, the consumption pattern can be analyzed by visiting the mart once a week on Monday. A consumption pattern can be analyzed by visiting a shopping mall once at the beginning of the month.

Thereafter, when the type of business of the first store is confirmed, the device 30 determines the purchaser's consumption pattern for the type of business. It is possible to set a consumption forecast period in which future consumption is expected to occur for the industry.

For example, if the industry is a convenience store industry, the device 30 can set 6:00 pm as the consumption prediction time of the convenience store industry when it is analyzed that the consumption pattern of the convenience store industry is mainly consumed in the evening, and the industry is In the case of the mart industry, if the consumption pattern of the mart industry is analyzed to be mainly on Monday, Monday can be set as the consumption forecast period for the convenience store industry. When it is analyzed that consumption has been made on , 1 day can be set as a consumption prediction period for the shopping mall industry.

In step S1007 , the device 30 may check whether or not the consumption prediction time has arrived when the consumption prediction time for the type of business is set.

For example, if the device 30 determines that the consumption prediction time of the convenience store industry is set to 6:00 pm when the industry is a convenience store industry, it can be confirmed that the consumption prediction time has arrived after 6:00 pm, In the case of this mart industry, if it is confirmed that the consumption forecast period for the mart industry is set to Monday, it can be confirmed that the consumption forecast period has arrived after 9 am on Monday, and if the industry is a shopping mall industry, the consumption forecast for the shopping mall industry If it is confirmed that the time is set to 1st, it can be confirmed that the consumption prediction time has arrived only after 9 am on the 1st.

If it is determined that the consumption prediction time has arrived in step S1007, the device 30 may provide recommendation information on products, colors, and stores to the purchaser terminal 20 in step S1008. Here, recommendation information on products, colors, and stores may be provided as recommendation guide information for buyers. Through this, a product may be recommended as a recommended product for a buyer, a color may be recommended as a recommended color for a buyer, and a store may be recommended as a recommended store for a buyer.

11 is an exemplary diagram of a configuration of a device 30 according to an embodiment.

Device 30 according to one embodiment includes a processor 31 and a memory 32 . Device 30 according to an embodiment may be the above-described server or terminal. The processor may include at least one device described above with reference to FIGS. 1 to 10 or may perform at least one method described above with reference to FIGS. 1 to 10 . The memory 32 may store information related to the above method or store a program in which the above method is implemented. Memory 32 may be volatile memory or non-volatile memory.

The processor 31 may execute a program and control the device 30 . Program codes executed by the processor 31 may be stored in the memory 32 . The device 30 may be connected to an external device (eg, a personal computer or network) through an input/output device (not shown) and exchange data.

The embodiments described above may be implemented as hardware components, software components, and/or a combination of hardware components and software components. For example, the devices, methods and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). array), programmable logic units (PLUs), microprocessors, or any other device capable of executing and responding to instructions. A processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the embodiment or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. The device can be commanded. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

As described above, although the embodiments have been described with limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, the described techniques may be performed in an order different from the method described, and/or components of the described system, structure, device, circuit, etc. may be combined or combined in a different form than the method described, or other components may be used. Or even if it is replaced or substituted by equivalents, appropriate results can be achieved.

Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

Claims (3)

In the method performed by the device,
Receiving a candidate product from a seller terminal of a seller;
collecting sales data for the candidate product sold in an online shopping mall;
classifying the candidate product into a category including at least one of fashion clothing, fashion accessories, beauty, digital/home appliances, food, leisure, sports, childbirth/parenting, and life/health;
generating a competition index by extracting the number of registered sales products of the candidate product and the number of competitors of the candidate product based on the sales data;
generating a review index by extracting the number of registered reviews associated with the candidate product, the length of the review, and the number of images included in the review from review data registered in association with the candidate product;
generating an interest index by extracting the number of searches for the candidate product and the number of sales of the candidate product based on the sales data;
generating a trend index by checking the degree of mention of the candidate product on social networks and the degree of mention in a community related to product sales;
selecting a recommended product for sale from among the candidate products through a first artificial neural network based on the competition index, the review index, the interest index, and the trend index; and
Transmitting a guide message for the recommended product for sale to the seller terminal;
including,
The step of selecting the sales recommendation product through the first artificial neural network,
applying the candidate product to an output node;
Applying a first value corresponding to the competition index, a second value corresponding to the review index, a third value corresponding to the interest index, and a fourth value corresponding to the trend index to an input node;
outputting a score of the candidate product using the first artificial neural network; and
Selecting the recommended product for sale based on the score of the candidate product
including,
Determining the price of the product recommended for sale; further comprising,
The step of determining the price is,
Extracting a sales price of the recommended product for sale registered in the online shopping mall;
Calculating an average selling price of the product recommended for sale based on the extracted price;
Extracting the sales volume of the sales recommendation product registered in the online shopping mall;
Calculating an average sales volume of the product recommended for sale based on the extracted sales volume;
Calculating an estimated sales volume of the product recommended for sale by applying a first weight determined based on a score obtained through the first artificial neural network to the average sales volume; and
Determining the price of the recommended product for sale by applying a second weight determined based on the calculated expected sales volume to the average sales price,
How to provide AI-based product sourcing and category-customized review analysis services. delete According to claim 1,
Further comprising, when the seller possesses the product recommended for sale, transmitting a guide message about a product for sale of the recommended product for sale to the seller terminal based on a review of the product recommended for sale,
In the step of transmitting a guide message about a product for sale of the recommended product for sale to the seller terminal based on the review of the recommended product for sale,
Extracting text information of the review data of the recommended product for sale and extracting keywords;
Setting a keyword mentioned more than a predetermined standard number of times from the extracted keyword as a main keyword;
Classifying the main keywords as positive keywords or negative keywords;
Calculating a positive rate, which is a rate occupied by the positive keyword among the keywords;
Calculating a negative ratio, which is a ratio occupied by the negative keyword among the keywords;
Comparing the negative ratio with a first reference value and a second reference value set in advance;
When the negative ratio is lower than the first reference value, transmitting a guide message recommending continued sales of the owned sales recommendation product to the seller terminal;
When the negative ratio is higher than the first reference value and lower than the second reference value, the negative keywords are sorted in the order in which many of the negative keywords were extracted, and improvement points of the recommended product for sale are determined based on the sorted negative keywords. Generating and transmitting a guide message for the negative keyword and the improvement point to the seller terminal;
When the negative ratio is higher than the second reference value, transmitting a guide message recommending changing the held sales recommendation product to one of the products included in the sales recommendation product to the seller terminal;
containing
How to provide AI-based product sourcing and category-customized review analysis services.

Images