KR102552856B1 - Method, device and system for automating creation of content template and extracting keyword for platform service that provide content related to commerce - Google Patents

Abstract

The present invention relates to a method, a device, and a system for automating keyword extraction and content template creation for a commerce-linked content providing platform service, capable of easily receiving necessary product information through content. The method of the present invention, comprises the steps of: collecting posting data for a first location; selecting an expected recommended item; classifying recommended items into a first group; generating a first template; and generating first content.

Description

Keyword extraction and content template generation automation method, device and system for commerce-linked content providing platform service

The following embodiments relate to technologies for extracting keywords for platform services that provide content associated with commerce and automatically generating content templates.

BACKGROUND OF THE INVENTION With the development of the Internet, electronic commerce is being actively conducted on the Internet, and the types of products provided in such electronic commerce are becoming very diverse.

In particular, because fashion products are rapidly changing in fashion and tastes vary greatly, consumers have difficulty in selecting products and managers managing e-commerce have difficulty in proposing products.

Therefore, from the manager's point of view, there is an increasing demand to efficiently provide product information to consumers by classifying many types of products according to appropriate situations. The demand for this is increasing, and the implementation of related technologies is being demanded.

Korean Patent Registration No. 10-2504773 Korean Patent Registration No. 10-2473915 Korean Patent Registration No. 10-1422128 Korean Patent Publication No. 10-2016-0117678

According to one embodiment, an object of the present invention is to provide a method, apparatus, and system for extracting keywords for a platform service that provides content associated with commerce and automatically generating content templates.

The object of the present invention is not limited to the object mentioned above, and other objects not mentioned will be clearly understood from the description below.

According to one embodiment, in the method of automating keyword extraction and content template creation for a commerce-linked content providing platform service, performed by an apparatus, collecting posting data for a first place among postings uploaded on users' SNS step; When the posted material for the first place is collected during the first period, keywords related to the first category, which is any one of the fashion categories, are classified as first keyword information based on the hash tag of the posted material collected during the first period. extracting; The first keyword information is applied to a first artificial neural network, and based on the output of the first artificial neural network, items belonging to the first category will have the most interest in the first place during the first period. Selecting an expected recommendation item; classifying the recommended items selected for each category into a first group when recommended items expected to have the most interest in the first place during the first period are selected for each category; generating a first template for inputting information on recommended items classified into the first group; providing a work page generated based on the first template to a worker terminal; and when information on recommended products selected for each recommended item is input through a work page generated based on the first template, when visiting the first place during the first period based on the input recommended product information. There is provided a method for automating keyword extraction and content template generation for a commerce-linked content providing platform service, including the step of generating first content suggesting a first outfit that is the outfit of .

In the step of generating the first template, when items worn on a first part of the body are classified into the first category and the first item is selected as a recommended item of the first category, the first part is assigned to the first part. A step of generating a first field, wherein the step of generating the first content includes, when information on a first product, which is a recommended product of the first item, is input through the first field, the step of generating the first content is based on the information on the first product. In this way, it is checked whether the first product is a product corresponding to the first item, and if the first product is identified as a product corresponding to the first item, an image of the first product is displayed in the first part. It may include processing to be displayed.

In the method for automating keyword extraction and content template creation for the commerce-linked content providing platform service, when posting materials for the first place are collected during a second period after the first period, posting materials collected during the second period extracting keywords related to the first category as second keyword information based on the hash tag of; The second keyword information is applied to the first artificial neural network, and items belonging to the first category, based on the output of the first artificial neural network, are of the most interest in the first place during the second period. Selecting recommended items that are expected to be recommended; classifying the recommended items selected for each category into a second group when recommended items expected to have the most interest in the first place during the second period are selected for each category; calculating a degree of agreement between the recommended items classified as the first group and the recommended items classified as the second group at a first ratio; setting the first content as content suggesting attire for visiting the first place from the first period to the second period, if it is confirmed that the first ratio is higher than a preset first reference ratio; ; When it is determined that the first ratio is lower than the first reference ratio, classifying recommended items that do not match the recommended items classified as the first group among the recommended items classified as the second group into a third group. step; generating a second template for inputting information on recommended items classified into the third group; providing a work page generated based on the second template to the worker terminal; and when information on recommended products selected for each recommended item is input through a work page generated based on the second template, the first item among the products constituting the first outfit is based on the information on the input recommended products. The method may further include updating product information of items classified into three groups and generating second content suggesting second clothes that are clothes worn when visiting the first place during the second period.

According to one embodiment, by extracting keywords necessary for suggesting products necessary to consumers, generating templates for generating content through the extracted keywords, and generating content based on information input through the template, keywords There is an effect of easily providing content related to commerce through extraction and template creation automation.

In addition, according to one embodiment, by generating content suggesting attire when visiting a specific place during a specific period, a manager classifies many types of products according to appropriate situations and efficiently distributes product information through content. It can be provided to consumers, and from the consumer's point of view, there is an effect of being able to easily receive necessary product information through content according to a desired situation.

On the other hand, the effects according to the embodiments are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

1 is a diagram schematically showing the configuration of a system according to an embodiment.
2 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a first place during a first period from among items belonging to a first category according to an embodiment.
3 is a flowchart illustrating a process of generating first content according to an exemplary embodiment.
4 is a diagram for explaining a process of generating a field in a body part according to an embodiment.
5 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a first place during a second period from among items belonging to a first category according to an exemplary embodiment.
6 is a flowchart illustrating a process of calculating a degree of agreement between recommended items selected for each period according to an embodiment and setting content according to the degree of agreement.
7 is a flowchart illustrating a process of generating second content according to an exemplary embodiment.
8 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a second place during a first period from among items belonging to a first category according to an exemplary embodiment.
9 is a flowchart illustrating a process of calculating a degree of agreement between recommended items selected for each location and setting content according to the degree of agreement according to an embodiment.
10 is a flowchart illustrating a process of generating third content according to an exemplary embodiment.
11 is a flowchart illustrating a process of determining a marketing cost of a product according to an embodiment.
12 is a flowchart illustrating a process of providing a product marketing strategy according to an embodiment.
13 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.

In an embodiment, an artificial intelligence (AI) system is a computer system that implements human-level intelligence, and unlike existing rule-based smart systems, a machine learns and judges on its own. The more the AI system is used, the higher the recognition rate and the more accurately the seller's taste can be understood, so the existing rule-based smart system is gradually being replaced by a deep learning-based AI system.

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 schematically showing the configuration of a system according to an embodiment.

Referring to FIG. 1 , a system according to an embodiment may include a worker terminal 100 and an apparatus 200 .

First, the worker terminal 100 and the device 200 may be connected through a communication network, and the communication network is configured regardless of its communication mode, such as wired and wireless, so that communication between servers and communication between servers and terminals is performed It can be implemented in various forms.

The worker terminal 100 may be implemented as a computing device having a communication function, and may be implemented as, for example, a mobile phone, a desktop PC, a laptop PC, a tablet PC, a smart phone, etc., but is not limited thereto, and is not limited thereto. It may be implemented in various types of communication devices that can be connected to the server.

The worker terminal 100 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 operator terminal 100 may be configured to communicate with the device 200 by wire or wireless.

The worker terminal 100 accesses a web page operated by a service provider or organization using the device 200, or an application developed and distributed by a service provider or organization using the device 200 can be installed The worker terminal 100 may be linked with the device 200 through a web page or application.

The device 200 may be a server owned by a person or organization that provides services using the device 200, may be a cloud server, or may be a peer-to-peer (p2p) set of distributed nodes. may be The device 200 may be configured to perform all or some of the arithmetic function, storage/reference function, input/output function, and control function of a normal computer. The device 200 may include at least one artificial intelligence model performing an inference function.

The device 200 may be configured to communicate with the worker terminal 100 by wire or wirelessly, control the operation of the worker terminal 100, and control which information is to be displayed on the screen of the worker terminal 100. .

The device 200 may be implemented as a server that extracts keywords for a commerce-related content providing platform service and automatically generates a content template, and may operate and manage a content providing platform linked to commerce.

According to an embodiment, the content associated with commerce is content suggesting attire when visiting a specific place during a specific period, and may be implemented in the form of a post, magazine, or the like. If a product included in the clothing is selected from the contents, it is accessed to a sales page selling the selected product, and commerce and contents can be linked.

Meanwhile, for convenience of description, only the worker terminal 100 is shown in FIG. 1 , but the number of terminals may vary according to embodiments. As long as the processing capacity of the device 200 allows, the number of terminals is not particularly limited.

According to an embodiment, the device 200 may select a recommended item expected to have the most interest in a specific place during a specific period by using artificial intelligence.

In the present invention, artificial intelligence (AI) means a technology that imitates human learning ability, reasoning ability, perception ability, etc., and implements it with a computer, and concepts such as machine learning and symbolic logic can include Machine learning (ML) is an algorithm technology that classifies or learns the characteristics of input data by itself. Artificial intelligence technology is a machine learning algorithm that analyzes input data, learns the result of the analysis, and can make judgments or predictions based on the result of the learning. In addition, technologies that mimic the functions of the human brain, such as recognition and judgment, using machine learning algorithms, can also be understood as artificial intelligence. For example, technical fields such as linguistic understanding, visual understanding, inference/prediction, knowledge expression, and motion control may be included.

Machine learning may refer to processing that trains a neural network model using experience of processing data. Through machine learning, computer software could mean improving its own data processing capabilities. A neural network model is constructed by modeling a correlation between data, and the correlation may be expressed by a plurality of parameters. The neural network model derives a correlation between data by extracting and analyzing features from given data, and optimizing the parameters of the neural network model by repeating this process can be referred to as machine learning. For example, a neural network model may learn a mapping (correlation) between an input and an output with respect to data given as an input/output pair. Alternatively, even when only input data is given, the neural network model may learn the relationship by deriving a regularity between given data.

An artificial intelligence learning model or a neural network model may be designed to implement a human brain structure on a computer, and may include a plurality of network nodes having weights while simulating neurons of a human neural network. A plurality of network nodes may have a connection relationship between them by simulating synaptic activities of neurons that transmit and receive signals through synapses. In the artificial intelligence learning model, a plurality of network nodes can send and receive data according to a convolutional connection relationship while being located in layers of different depths. The artificial intelligence learning model may be, for example, an artificial neural network model, a convolutional neural network model (CNN), and the like. As an embodiment, the artificial intelligence learning model may be machine-learned according to methods such as supervised learning, unsupervised learning, and reinforcement learning. Machine learning algorithms for performing machine learning include Decision Tree, Bayesian Network, Support Vector Machine, Artificial Neural Network, Ada-boost , Perceptron, Genetic Programming, Clustering, etc. may be used.

Of these, CNNs are a type of multilayer perceptrons designed to use minimal preprocessing. A CNN consists of one or several convolution layers and general artificial neural network layers on top, and additionally utilizes weights and pooling layers. Thanks to this structure, CNN can fully utilize the input data with a two-dimensional structure. Compared to other deep learning structures, CNN shows good performance in both video and audio fields. CNNs can also be trained via standard back-propagation. CNNs are easier to train than other feedforward artificial neural network techniques and have the advantage of using fewer parameters.

Convolutional networks are neural networks that contain sets of nodes with bound parameters. The increasing size of available training data and the availability of computational power, combined with algorithmic advances such as piecewise linear unit and dropout training, have greatly improved many computer vision tasks. With huge data sets, such as those available for many tasks today, overfitting is not critical, and increasing the size of the network improves test accuracy. Optimal use of computing resources becomes a limiting factor. To this end, a distributed, scalable implementation of deep neural networks can be used.

2 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a first place during a first period from among items belonging to a first category according to an embodiment.

Referring to FIG. 2 , first, in step S201, the device 200 may collect posting data for a first place among postings uploaded on users' SNS. Here, the first place may be set in various ways according to the operator's request, such as a specific country, city, tourist destination, or street.

The apparatus 200 may collect posting data for a first place among postings uploaded on SNS (Social Network Service) of users through various techniques such as crawling.

Specifically, the device 200 may collect posting data for a first place among posts uploaded on an SNS from a plurality of SNS servers (not shown) connected to the device 200 . In this case, the device 200 may collect posting materials in which the name of the first place is set as a hashtag as posting materials for the first place.

In step S202, the device 200 may check whether posting data for the first place has been collected during the first period. Here, the first period may be set in various ways according to operator requests such as an event period, week, month, season, and the like.

According to an embodiment, before step S201, the apparatus 200 may receive setting information about the first period and the first place from the worker terminal 100, and based on the received setting information, the first period and the first location may be respectively identified.

If it is determined in step S202 that posting material is not collected during the first period, the apparatus 200 may continue to collect posting material for the first place, returning to step S201.

If it is confirmed in step S202 that the posting material is collected during the first period, in step S203, the device 200 sets keywords related to the first category to the first keyword based on the hash tag of the posting material collected during the first period. information can be extracted. Here, the first category is any one of the fashion categories, and the fashion category may be divided into a hat category, a top category, a bottom category, a shoe category, a bag category, and the like.

For example, during the first period, the first post and the second post were collected as posting materials for the first place, the hashtag of the first post was identified as “first place”, “shirt”, and the second post When the hashtags are identified as “first place” and “jeans,” the device 200 sets “shirt” identified based on the hashtag of the first post as the first keyword information when the first category is the top category. If the first category is a bottom category, "jeans" identified based on the hash tag of the second posting may be extracted as the first keyword information.

The database of the device 200 stores an important keyword list classified by category, and the important keyword list includes keywords mainly used in each category. The list of important keywords for each category can be manually set by a worker or automatically through big data analysis.

After acquiring the list of important keywords of the first category from the database, the device 200 compares the keywords included in the list of important keywords of the first category with hashtags of posted materials collected during the first period, and matches each other. A keyword may be extracted as first keyword information.

In step S204 , the device 200 may apply the first keyword information to the first artificial neural network trained in advance in the device 200 . Here, the first artificial neural network may be an algorithm that, after receiving keyword information, analyzes and outputs a recommended item that is expected to have the most interest in a specific place during a specific period among items belonging to a specific category.

According to an embodiment, in a relationship between a category and an item, a category is a higher category and an item is a lower category, and a plurality of items may be included in one category.

In step S205, the apparatus 200 may select a recommended item expected to have the most interest in the first place during the first period from among items belonging to the first category, based on the output of the first artificial neural network. . Here, items belonging to the first category may be set differently for each category. For example, when the first category is a top category, items belonging to the first category may be classified as shirts, knits, hoods, etc. When the first category is a bottom category, it may be classified into jeans, cotton pants, and shorts.

For example, as a result of checking the output of the first artificial neural network by applying the first keyword information to the first artificial neural network, the apparatus 200 determines that the output value is “”, the first of the items belonging to the first category If a recommended item that is expected to have the most interest in the first place during the period is selected as the first item and the output value is confirmed as “1,” the most popular item in the first place during the first period among the items belonging to the first category is selected. A recommended item expected to have a lot of interest may be selected as the second item.

According to an embodiment, the first artificial neural network may be an algorithm that, after receiving keyword information, analyzes and outputs a recommended item that is expected to have the most interest in a specific place during a specific period among items belonging to a specific category. .

The learning device in which the first artificial neural network is learned may be the same device as the device 200 or a separate device. Hereinafter, a process of learning the first artificial neural network will be described.

First, the learning device may generate an input to be input to the first artificial neural network. In this case, the learning device may generate an input based on keyword information extracted from keywords related to a specific category to be input to the first artificial neural network.

Specifically, the learning device may perform a process of pre-processing keyword information. The preprocessed keyword information may be used as an input of the first artificial neural network, or an input of the first artificial neural network may be generated through a normal process of removing unnecessary information.

Next, the learning device may apply the input to the first artificial neural network.

The first artificial neural network may be an artificial neural network trained according to reinforcement learning. The first artificial neural network may be a Q-Network, a Depp Q-Network (DQN), or a relational network (RL) structure suitable for outputting abstract inference through reinforcement learning.

The first artificial neural network learned through reinforcement learning may be updated and optimized by reflecting evaluations on various rewards. For example, the first artificial neural network may be updated and optimized through first and second compensation. Here, the first reward may increase as an item related to the keyword mentioned most in the keyword information is selected as a recommended item, and the second reward selects an item related to a keyword not mentioned in the keyword information as a recommended item. The more you don't, the higher it can go.

Next, the learning device may obtain an output from the first artificial neural network. Here, the output of the first artificial neural network may be information on a recommended item. At this time, based on the keyword information, the first artificial neural network selects an item that is expected to have the most interest in a specific place during a specific period among items belonging to a specific category as a recommended item, and provides information on the selected recommended item. can be printed out.

Next, the learning device may evaluate the output of the first artificial neural network and provide a reward. At this time, the evaluation of the output may be divided into first compensation, second compensation, and the like.

For example, as the learning device selects an item related to a keyword mentioned the most in keyword information as a recommended item, more first rewards are awarded, and an item related to a keyword not mentioned in keyword information is selected as a recommended item. The more the second reward is not selected, the more the second reward can be awarded.

Next, the learning device may update the first artificial neural network based on the evaluation.

Specifically, the learning device is configured in specific states so that the expectation of the sum of reward values is maximized in an environment in which the first artificial neural network analyzes a recommended item through keyword information. The first artificial neural network may be updated through a process of optimizing a policy for determining actions to be taken.

For example, when there is a problem with the selection of the first item as a recommendation item through the first keyword information, the learning device generates first learning data including information indicating that there is a problem with the selection of the first item. and applying the first training data to the first artificial neural network so that the first item is not selected as a recommended item through keyword information similar to the first keyword information, through a process of learning the first artificial neural network. Neural networks can be updated.

Meanwhile, the process of optimizing the policy may be performed through a process of estimating the maximum value of the expected value of the sum of rewards or the maximum value of the Q-function, or estimating the minimum value of the loss function of the Q-function. Estimation of the minimum value of the loss function may be performed through stochastic gradient descent (SGD) or the like. The process of optimizing the policy is not limited thereto, and various optimization algorithms used in reinforcement learning may be used.

The learning device may gradually update the first artificial neural network by repeating the learning process of the first artificial neural network as described above.

Specifically, in selecting a recommendation item through keyword information, the learning device considers keywords included in the keyword information, and recommends items that are expected to attract the most interest in a specific place during a specific period among items belonging to a specific category. A first artificial neural network that analyzes and outputs items may be trained.

Through this, the learning device may learn the first artificial neural network that, after receiving keyword information, analyzes and outputs recommended items that are expected to have the most interest in a specific place during a specific period among items belonging to a specific category. .

That is, when analyzing the recommended item through the keyword information, the learning device may reflect reinforcement learning through the first reward and the second reward and adjust the analysis criterion to train the first artificial neural network.

3 is a flowchart illustrating a process of generating first content according to an exemplary embodiment, and FIG. 4 is a flowchart illustrating a process of generating a field in a body part according to an exemplary embodiment.

Referring to FIG. 3 , first, in step S301, the device 200 may confirm that a recommended item expected to have the most interest in the first place during the first period is selected for each category.

That is, by repeatedly performing steps S203 to S205, a recommended item may be selected for each category, and the apparatus 200 may confirm that each recommended item is selected for each category.

In step S302, the device 200 may classify recommended items selected for each category into a first group.

For example, a recommended item expected to have the most interest in a first place during a first period among items belonging to the first category is selected as the first item, and among items belonging to the second category during the first period When a recommended item expected to have the most interest in the first place is selected as the second item, the device 200 may classify the first item and the second item into the first group.

In step S303, the device 200 may generate a first template for inputting information on recommended items classified into the first group.

For example, when a first item and a second item are classified into a first group, the device 200 provides a first field for inputting information of the first item and a second field for inputting information of the second item. A first template including fields may be created.

Specifically, when items worn on a first part of the body are classified into a first category and the first item is selected as a recommended item of the first category, the device 200 generates a first field in the first part. can In this case, the name of the first field may be set to the first item.

As shown in FIG. 4, the body can be divided into parts such as the head, upper body, lower body, feet, hands, etc., items worn on the head are classified as hats, and items worn on the upper body are classified as upper body. If it is classified as a category, items worn on the lower body part are classified as the bottom category, items worn on the foot part are classified as the shoe category, and items worn on the hand part are classified as the bag category, the device 200, if a cap is selected as a recommended item in the hat category, a field for inputting information on the cap can be created in the head area, and if a shirt is selected as a recommended item in the upper body category, a field for entering the cap hat information can be created in the upper body area. A field for inputting information can be created, when jeans are selected as a recommended item for the lower body category, a field for inputting jeans information for the lower body can be created, and when sneakers are selected as a recommended item for the foot category , A field for inputting sneaker information can be created in the foot area, and if a handbag is selected as a recommended item in the hand category, a field for inputting handbag information can be created in the hand area.

That is, when recommended items are selected for each category, the device 200 generates fields for inputting information of the selected recommended items, arranges the generated fields according to the worn part, and through this creates a template. can

In step S304 , the device 200 may provide a work page generated based on the first template to the operator terminal 100 .

In step S305, when information on recommended products selected for each recommended item is input through the work page generated based on the first template, the device 200 selects a first location for a first period based on the input recommended product information. First content suggesting first attire, which is attire when visiting , may be generated. Here, the input information on recommended products may include information about product names, images, sales page links, and the like.

For example, when a first item is selected as a recommended item of a first category and a second item is selected as a recommended item of a second category, the device 200 selects the first product as a recommended product for the first item. When information is entered and information on a second product is entered as a recommended product for the second item, a first outfit can be configured with the first product and the second product, and when visiting the first place during the first period First content may be generated with content suggesting the first attire.

Specifically, when information on a first product, which is a recommended product of the first item, is input through the first field, the device 200 determines whether the first product corresponds to the first item based on the information on the first product. , and if the first product is identified as a product corresponding to the first item, processing may be performed so that the image of the first product is displayed on the first part.

For example, when a first field for inputting cap and hat information is generated on the head, the device 200, when information on a first product is input through the first field, based on the information on the first product , it is checked whether the first product corresponds to a cap, and if the first product is identified as a product corresponding to a cap, processing may be performed to display an image of the first product on the head.

That is, when information on recommended products is input through each field, the device 200 processes product images to be displayed on the worn part based on the input recommended product information, and through this, product images displayed on each part are displayed. It is possible to configure the first outfit.

When any one of the product images constituting the first outfit is selected, the device 200 may process access to a sales page selling the corresponding product through the selected product image, through which commerce and content are linked. It can be, and content associated with commerce can be provided.

5 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a first place during a second period from among items belonging to a first category according to an exemplary embodiment.

Referring to FIG. 5 , first, in step S501 , the device 200 may collect posting data for a first place among postings uploaded on users' SNS.

In step S502, the device 200 may check whether posting data for the first place has been collected during the second period. Here, the second period may be set to a period after the first period, and for example, when the first period is set to January, the second period may be set to February.

If it is confirmed in step S502 that the posted material is not collected during the second period, the apparatus 200 may continue to collect posted material for the first place, returning to step S501.

If it is confirmed in step S502 that the posted material is collected during the second period, in step S503, the device 200 sets keywords related to the first category to a second keyword based on the hash tag of the posted material collected during the second period. information can be extracted.

In step S504, the device 200 may apply the second keyword information to the first artificial neural network.

In step S505, the apparatus 200 may select a recommended item expected to have the most interest in the first place during the second period from among items belonging to the first category, based on the output of the first artificial neural network. .

6 is a flowchart illustrating a process of calculating a degree of agreement between recommended items selected for each period according to an embodiment and setting content according to the degree of agreement.

Referring to FIG. 6 , first, in step S601, the device 200 may confirm that a recommended item expected to have the most interest in the first place during the second period is selected for each category.

That is, by repeatedly performing steps S503 to S505, a recommended item may be selected for each category, and the apparatus 200 may confirm that each recommended item is selected for each category.

In step S602, the device 200 may classify recommended items selected for each category into a second group.

In step S603, the apparatus 200 may calculate a degree of agreement between the recommended items classified as the first group and the recommended items classified as the second group at a first ratio.

Specifically, the device 200 identifies the number of recommended items classified as the first group or the number of recommended items classified as the second group as a first numerical value, and divides the recommended items classified as the first group into the second group. When the number of matching items among the classified recommended items is confirmed as the second numerical value, the first ratio may be calculated as a value obtained by dividing the second numerical value by the first numerical value.

In step S604, the device 200 may check whether the first ratio is higher than the first reference ratio. Here, the first reference ratio may be set differently according to embodiments.

If it is determined in step S604 that the first ratio is lower than the first reference ratio because the first ratio is not higher than the first reference ratio, step S701 may be performed, and a detailed description related to this will be described later with reference to FIG. 7 .

If it is confirmed in step S604 that the first ratio is higher than the first reference ratio, in step S605, the device 200 proposes clothes for visiting the first place from the first period to the second period for the first content. It can be set as content.

That is, when the first content is set as content suggesting attire for visiting the first place during the first period, the device 200 presents the first content when the first ratio is higher than the first reference ratio. Set the clothes to be worn when visiting the first place from the 1st period to the 2nd period as content suggesting that the clothes that can be seen on the 1st content can be worn at the 1st place from the 1st period to the 2nd period can do.

7 is a flowchart illustrating a process of generating second content according to an exemplary embodiment.

Referring to FIG. 7 , first, in step S701, if the first ratio is lower than the first reference ratio, the device 200 selects recommended items classified as the first group among recommendation items classified as the second group. Recommended items that do not match may be classified into a third group.

In step S702, the device 200 may generate a second template for inputting information on recommended items classified into the third group. In this case, the second template may include only some of the plurality of fields included in the first template.

In step S703 , the device 200 may provide a work page generated based on the second template to the worker terminal 100 .

In step S704, when information on recommended products selected for each recommended item is input through the work page generated based on the second template, the device 200 configures the first outfit based on the input recommended product information. Product information of items classified into the third group among the products may be updated to generate second content suggesting second attire when visiting the first place during the second period.

For example, when a first item is selected as a recommended item of a first category and a second item is selected as a recommended item of a second category based on published data on a first place collected during a first period, When information on a first product is input as a recommended product for a first item and information on a second product is input as a recommended product for a second item, the device 200 sets the first outfit as the first product and the second product. , and when visiting a first place during a first period, first content may be generated with content suggesting a first outfit.

Thereafter, the device 200 selects the first item as a recommendation item of the first category and selects the third item as a recommendation item of the second category based on the posting data for the first place collected during the second period. In this case, as a result of comparing the recommended items classified as the first group with the recommended items classified as the second group, it is confirmed that the third item exists only in the second group, and the third item can be classified as the third group. and, when the third product information is input as a recommended product for the third item, the second product information in the first outfit is updated with the third product information, and the second outfit is composed of the first product and the third product. When visiting the first place during the second period, second content may be generated with content suggesting a second outfit.

8 is a flowchart illustrating a process of selecting a recommended item expected to have the most interest in a second place during a first period from among items belonging to a first category according to an exemplary embodiment.

Referring to FIG. 8 , first, in step S801 , the device 200 may collect posting data for a second place among postings uploaded on users' SNS. Here, the second place may be set in various ways according to the operator's request, such as a specific country, city, tourist destination, or street.

In step S802, the device 200 may check whether posting data for the second place has been collected during the first period.

If it is determined in step S802 that posting material is not collected during the first period, returning to step S801 , the device 200 may continue to collect posting material for the second location.

If it is determined in step S802 that the posted material is collected during the first period, in step S803, the device 200 sets keywords related to the first category to a third keyword based on the hash tag of the posted material collected during the first period. information can be extracted.

In step S804, the device 200 may apply the third keyword information to the first artificial neural network.

In step S805, the apparatus 200 may select a recommendation item expected to have the most interest in the second place during the first period from among items belonging to the first category, based on the output of the first artificial neural network. .

9 is a flowchart illustrating a process of calculating a degree of agreement between recommended items selected for each location and setting content according to the degree of agreement according to an embodiment.

Referring to FIG. 9 , first, in step S901 , the device 200 may confirm that a recommendation item expected to have the most interest in the second place during the first period is selected for each category.

That is, by repeatedly performing steps S803 to S805, a recommended item may be selected for each category, and the apparatus 200 may confirm that each recommended item is selected for each category.

In step S902, the device 200 may classify recommended items selected for each category into a fourth group.

In step S903, the apparatus 200 may calculate a degree of agreement between the recommended items classified as the first group and the recommended items classified as the fourth group at a second ratio.

Specifically, the device 200 identifies the number of recommended items classified as the first group or the number of recommended items classified as the fourth group as a third numerical value, and divides the recommended items classified as the first group and the recommended items into the fourth group. If the number of matching items among the classified recommended items is confirmed as the fourth numerical value, the second ratio may be calculated as a value obtained by dividing the fourth numerical value by the third numerical value.

In step S904, the device 200 may check whether the second ratio is higher than the second reference ratio. Here, the second reference ratio may be set differently according to embodiments.

If it is confirmed in step S904 that the second ratio is lower than the second reference ratio because the second ratio is not higher than the second reference ratio, step S1001 may be performed, and a detailed description related to this will be described later with reference to FIG. 10 .

If it is determined in step S904 that the second ratio is higher than the second reference ratio, in step S905, the device 200 proposes attire for visiting the first and second places for the first content during the first period. It can be set as content.

That is, when the first content is set as content suggesting attire for visiting the first place during the first period, and the second ratio is higher than the second reference ratio, the device 200 presents the first content. Set the clothes for visiting the 1st and 2nd places during the 1st period as content suggesting that the clothes that can be seen on the 1st content can be worn not only at the 1st place but also at the 2nd place during the 1st period can do.

10 is a flowchart illustrating a process of generating third content according to an exemplary embodiment.

Referring to FIG. 10 , first, in step S1001, if it is determined that the second ratio is lower than the second reference ratio, the device 200 generates a third template for inputting information on recommendation items classified into the fourth group. can do.

In step S1002 , the device 200 may provide a work page generated based on the third template to the worker terminal 100 .

In step S1003, when information on recommended products selected for each recommended item is input through the work page generated based on the third template, the device 200 selects a second location for a first period based on the input recommended product information. It is possible to generate third content suggesting third attire, which is attire when visiting .

In step S1004, the device 200 may check whether the second ratio is higher than the third reference ratio. Here, the third reference ratio may be set differently according to embodiments.

If it is confirmed in step S1004 that the second ratio is higher than the third reference ratio, in step S1005, the device 200 may set the third content as recommended content associated with the first content.

That is, if it is confirmed that the second ratio is higher than the second reference ratio, the device 200 does not create new content, and the device 200 wears clothes for visiting the first place and the second place for the first period of time. If it is determined that the second ratio is lower than the second reference ratio but higher than the third reference ratio, third content is generated and the third content is set as recommended content linked to the first content. can

11 is a flowchart illustrating a process of determining a marketing cost of a product according to an embodiment.

Referring to FIG. 11 , first, in step S1101 , the apparatus 200 may receive first product information from the worker terminal 100 . In this case, the first product information may include, but is not limited to, product name, sales destination, item, category, sales price, number of sales, and marketing price of the first product. In this case, the marketing amount may refer to an amount of money desired to be spent in marketing the product.

Specifically, the device 200 generates first content in which an image of the first product is displayed in a first part by inputting information on a first product, which is a recommended product of the first item, through a first field, and then generating product information. An input page for input may be generated and transmitted to the worker terminal 100 , and when first product information is input through the product information input page, the first product information may be received from the worker terminal 100 .

In step S1102, the device 200 may collect customer activity information from a sales site that sells the first product. In this case, the customer activity information may include a search history of a first item to which the first product belongs, a review of the first product, and an inquiry about the first product.

For example, when the first product is a jeans product, the first item may be jeans, and the device 200 collects a search history of jeans from a sales site, collects reviews of the first product, and selects the first review You can collect inquiries about and create customer activity information through the collected information. To this end, the device 200 may be connected to a server providing a sales site or implemented by including a server providing a sales site.

In step S1103, when the first product is being sold by the first company, the device 200 may extract competitor information of the first company from the first product information.

For example, the device 200 considers the item and seller of the first product identified through the first product information, and sets a seller that sells an item matching the item of the first product as competitor information of the first company. can be extracted. At this time, the competitor information may include the name of a competitor's company, the name of a product sold by a competitor, an item of a product sold by a competitor, a sales volume of a product sold by a competitor, a sales period, and the like.

In step S1104, the device 200 determines that a competitor of the first company is a second company based on the competitor information, and that the second company sells a second product as a product corresponding to the first item. , marketing information of the second product may be extracted from competitor information. In this case, the marketing information may include the number of marketing of the second product sold by the second company, marketing period, marketing time, marketing cost, marketing name, and the like.

In step S1105, the device 200 may generate positive keyword information by extracting positive keywords for the first product from the customer activity information.

Specifically, the device 200 extracts positive keywords having a positive meaning from customer activity information including search history of the first item, reviews of the first product, and inquiries about the first product, and the extracted positive keyword Positive keywords can be extracted from

In this case, the positive keyword information may include, but is not limited to, extracted positive keywords, the number of positive keywords, a ratio of positive keywords to the extracted keywords, and the like.

In step S1106, the device 200 may apply the customer activity information, marketing information, and positive keyword information to the second artificial neural network.

Specifically, the device 200 may generate a first input signal based on customer activity information, marketing information, and positive keyword information. Here, the first input signal may include a first value corresponding to customer activity information, a second value corresponding to marketing information, and a third value corresponding to positive keyword information.

The first input layer of the second artificial neural network may include input nodes respectively applied to the first value, the second value, and the third value.

According to an embodiment, the device 200 may generate a higher first value as the number of searches for an item to which the product belongs increases, the number of reviews for the product increases, and the number of inquiries about the product increases. The first value may be generated as low as the number of searches for the belonging item decreases, as the number of product reviews decreases, and as the number of inquiries about the product decreases. In this case, the first value may be a real value between 0 and 1.

According to an embodiment, the device 200 may generate a higher second value as the competitor's marketing number of products sold, the longer the marketing period, the longer the marketing time, and the higher the marketing cost. The lower the second value may be generated as the number of times of marketing of a product on sale is reduced, the shorter the marketing period, the shorter the marketing time, and the lower the marketing cost. In this case, the second value may be a real value between 0 and 1.

According to an embodiment, the apparatus 200 may generate a higher third value as the number of extracted positive keywords increases, and may generate a lower third value as the number of extracted positive keywords decreases. In this case, the third value may be a real value between 0 and 1.

In step S1107, the device 200 may set an activation index of the first product based on the output of the second artificial neural network.

Specifically, the device 200 may apply the first input signal to the second artificial neural network and obtain the first output signal as an output of the second artificial neural network. Here, the first output signal may include an output value that is a real number between 0 and 1. The first output layer of the second artificial neural network may include an output node generating an output value.

Thereafter, the device 200 may set an activation index of the first product based on the first output signal.

In step S1108, the device 200 may determine a first weight based on the activation index of the first product. In this case, the first weight may be determined as a value proportional to the activation index of the first product. That is, the first weight may be set higher as the activation index of the first product increases, and may be set lower as the activation index of the first product decreases.

In step S1109 , the device 200 may calculate the average sales volume of the second product, that is, the average sales volume of the competitor based on the competitor information.

The apparatus 200 may obtain the sales volume of the second product sold by the competitor during a specific period based on the competitor information, and may calculate the average sales volume of the competitor, which is an average sales volume of the sales volume during the specific period. In this case, the specific period may be set differently according to embodiments.

In step S1110, the device 200 may calculate the expected sales volume of the first product by applying a first weight to the average sales volume of competitors.

In step S1111, the device 200 may determine a second weight based on the expected sales volume of the first product. In this case, the second weight may be generated as a value proportional to the expected sales volume of the first product. That is, the second weight may be set higher as the expected sales volume of the first product increases, and set lower as the predicted sales volume of the first product decreases.

In step S1112, the device 200 may check the marketing price of the first product based on the first product information.

In step S1113, the device 200 may determine the marketing cost of the first product by applying a second weight to the marketing cost of the first product.

12 is a flowchart illustrating a process of providing a product marketing strategy according to an embodiment.

Referring to FIG. 12 , first, in step S1201, the device 200 may extract search time, review time, and inquiry time from customer activity information.

In this case, the search time may mean time spent searching for the first item, review time may mean time spent registering a review of the first product, and inquiry time may mean time spent searching for the first product. It can mean the time required to do

The device 200 may extract search time, review time, and inquiry time from customer activity information.

In step S1202, the device 200 may apply the search time, review time, and inquiry time to the third artificial neural network.

In step S1203, the device 200 may set a marketing time for the first product based on the output of the third artificial neural network. In this case, the marketing time may mean a time during which product marketing is performed.

According to an embodiment, the third artificial neural network may be an algorithm that receives a search time, a review time, and an inquiry time as inputs and outputs a marketing time when the marketing effect of a product is expected to appear the most.

The learning device in which the third artificial neural network is learned may be the same device as the device 200 or may be a separate device. Hereinafter, a process of learning the third artificial neural network will be described.

First, the learning device may generate an input based on search time, review time, and inquiry time.

Specifically, the learning device may perform a process of pre-processing search time, review time, and inquiry time. The preprocessed matching result may be used as an input of the third artificial neural network, or an input may be generated through a normal process of removing unnecessary information.

Next, the learning device may apply the input to the third artificial neural network. The third artificial neural network may be a third artificial neural network trained according to reinforcement learning. The third artificial neural network may be a Q-Network, a Depp Q-Network (DQN), or a relational network (RL) structure suitable for outputting abstract inference through reinforcement learning.

The third artificial neural network learned through reinforcement learning may be updated and optimized by reflecting evaluations on various rewards. For example, the third reward may increase the reward value by predicting the marketing time in which the marketing effect of the product is expected to appear the most in consideration of the search time, and the fourth reward may consider the review time, and the marketing of the product If you predict the marketing time when the most effect is expected to appear, the reward value can be increased, and the fifth reward will increase the reward value if you predict the marketing time when the marketing effect of the product is expected to appear the most, considering the inquiry time. can

Next, the learning device may obtain an output from the third artificial neural network. The output of the third artificial neural network may include information about a marketing time when a marketing effect of a product is expected to appear the most. That is, the output of the third artificial neural network may be a result of predicting a marketing time in which the marketing effect of a product is expected to appear the most through search time, review time, and inquiry time.

The third artificial neural network may select and output a marketing time when the marketing effect of the product is expected to appear the most through search time, review time, and inquiry time. At this time, the third artificial neural network may predict and output the marketing time when the marketing effect of the product is expected to appear the most in consideration of the search time, review time, and inquiry time.

Next, the learning device may evaluate the output of the third artificial neural network and provide a reward. At this time, the evaluation of the output may be divided into third compensation, fourth compensation, fifth compensation, and the like.

Specifically, when the learning device accurately predicts the marketing time when the marketing effect of the product is expected to appear the most in consideration of the search time, a large number of third rewards are awarded, and considering the review time, the product's marketing effect is expected to appear the most. If the marketing time expected to occur is accurately predicted, a lot of fourth rewards may be awarded, and if the marketing time when the marketing effect of the product is expected to appear the most is predicted in consideration of the inquiry time, a lot of fifth rewards may be awarded.

Next, the learning device may update the third artificial neural network based on the evaluation.

Specifically, the learning device is a reward value (reward value) in an environment in which the third artificial neural network selects and extracts the marketing time in which the marketing effect of the product is expected to appear the most through search time, review time, and inquiry time. ), the third artificial neural network may be updated through a process of optimizing a policy for determining actions to be taken in specific states so as to maximize the expectation of the sum of the states.

For example, if the marketing time in which the marketing effect of the product is expected to appear the most is selected in the first time, the learning device indicates that there is no problem with the analysis of the marketing time if there is no problem with the selected result. , and apply the first learning data to the third artificial neural network to select the same marketing time when the search time, review time, and inquiry time are the same or similar. Through the process of learning, the third artificial neural network may be updated.

Meanwhile, the process of optimizing the policy may be performed through a process of estimating the maximum value of the expected value of the sum of rewards or the maximum value of the Q-function, or estimating the minimum value of the loss function of the Q-function. Estimation of the minimum value of the loss function may be performed through stochastic gradient descent (SGD) or the like. The process of optimizing the policy is not limited thereto, and various optimization algorithms used in reinforcement learning may be used.

The learning device may gradually update the third artificial neural network by repeating the above-described learning process of the third artificial neural network. Through this, the learning device can train the third artificial neural network that selects and outputs the marketing time when the marketing effect of the product is expected to appear the most through the search time, review time, and inquiry time.

That is, when the learning device predicts the marketing time when the marketing effect of the product is expected to appear the most through the search time, review time, and inquiry time, reinforcement learning is performed through the third reward, the fourth reward, and the fifth reward. In reflection, the third artificial neural network may be trained by adjusting the analysis criteria.

In step S1204 , the device 200 may confirm that the marketing time of the first product is determined as the first time and the marketing cost of the first product is determined as the first cost.

In step S1205, the device 200 may calculate the marketing cost per day as the second cost by applying the marketing cost per hour to the first time. In this case, the marketing cost per hour may mean the cost required to market a product for one hour, and the marketing cost per hour may be stored in the database of the device 200 for each product.

For example, if the first hour is from 1:00 pm to 4:00 pm every day and the marketing cost per hour is 100,000 won, the device 200 applies the marketing cost per hour to the marketing time to set the marketing cost per day to 300,000 won. can be calculated

In step S1206, the device 200 may determine the marketing period of the first product as the first period by dividing the first cost by the second cost.

For example, when the first cost is 30 million won and the second cost is 300 thousand won, the device 200 may determine 100 days calculated through “3000 / 30” as the first period.

In step S1207 , the device 200 may generate a marketing strategy for the first product including a first cost, a first time, and a first period. In this case, the marketing strategy may include marketing cost, marketing time, and marketing period, but is not limited thereto.

In step S1208 , the device 200 may provide a marketing strategy for the first product to the worker terminal 100 .

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

Device 200 according to an embodiment includes a processor 210 and a memory 220 . The processor 210 may include at least one device described above with reference to FIGS. 1 to 12 or may perform at least one method described above with reference to FIGS. 1 to 12 . A person or organization using the device 200 may provide services related to some or all of the methods described above with reference to FIGS. 1 to 12 .

The memory 220 may store information related to the methods described above or may store a program in which the methods described below are implemented. Memory 220 may be volatile memory or non-volatile memory.

The processor 210 may execute a program and control the device 200 . Program codes executed by the processor 210 may be stored in the memory 220 . The device 200 may be connected to an external device (eg, a personal computer or network) through an input/output device (not shown) and exchange data through wired/wireless communication.

Apparatus 200 may be used to train an artificial neural network or to use a trained artificial neural network. The memory 220 may include a learning or learned artificial neural network. The processor 210 may learn or execute an artificial neural network algorithm stored in the memory 220 . The learning device for learning the artificial neural network and the device 200 using the learned artificial neural network may be the same or separate.

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. 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)

A method for automating keyword extraction and content template generation for a commerce-linked content providing platform service, performed by an apparatus,
Collecting posting data for a first place among postings uploaded on users'SNS;
When the posted material for the first place is collected during the first period, keywords related to the first category, which is any one of the fashion categories, are classified as first keyword information based on the hash tag of the posted material collected during the first period. extracting;
The first keyword information is applied to a first artificial neural network, and based on the output of the first artificial neural network, items belonging to the first category will have the most interest in the first place during the first period. Selecting an expected recommendation item;
classifying the recommended items selected for each category into a first group when recommended items expected to have the most interest in the first place during the first period are selected for each category;
generating a first template for inputting information on recommended items classified into the first group;
providing a work page generated based on the first template to a worker terminal; and
When information on recommended products selected for each recommended item is input through the work page generated based on the first template, based on the input recommended product information, when visiting the first place during the first period Generating first content suggesting a first attire that is attire;
When posting materials for the first place are collected during a second period after the first period, keywords related to the first category are classified as second keyword information based on hash tags of posting materials collected during the second period. Extracting with;
The second keyword information is applied to the first artificial neural network, and items belonging to the first category, based on the output of the first artificial neural network, are of the most interest in the first place during the second period. Selecting recommended items that are expected to be recommended;
classifying the recommended items selected for each category into a second group when recommended items expected to have the most interest in the first place during the second period are selected for each category;
calculating a degree of agreement between the recommended items classified as the first group and the recommended items classified as the second group at a first ratio;
setting the first content as content suggesting attire for visiting the first place from the first period to the second period, if it is confirmed that the first ratio is higher than a preset first reference ratio; ;
When it is determined that the first ratio is lower than the first reference ratio, classifying recommended items that do not match the recommended items classified as the first group among the recommended items classified as the second group into a third group. step;
generating a second template for inputting information on recommended items classified into the third group;
providing a work page generated based on the second template to the worker terminal; and
When information on recommended products selected for each recommended item is input through the work page created based on the second template, the third item among the products constituting the first outfit is based on the information on the input recommended products. Updating product information of items classified into groups to generate second content suggesting second attire that is the attire when visiting the first place during the second period;
Collecting posting data for a second place among postings uploaded on users'SNS;
extracting keywords related to the first category as third keyword information based on hash tags of the posted materials collected during the first period, when posting materials for the second place are collected during the first period;
The third keyword information is applied to the first artificial neural network and, based on the output of the first artificial neural network, among items belonging to the first category, items of most interest in the second place during the first period Selecting recommended items that are expected to be recommended;
classifying the recommended items selected for each category into a fourth group when recommended items expected to have the most interest in the second place during the first period are selected for each category;
calculating a degree of agreement between the recommended items classified as the first group and the recommended items classified as the fourth group at a second ratio;
setting the first content as content suggesting attire for visiting the first place and the second place during the first period when it is determined that the second ratio is higher than a preset second reference ratio; ;
generating a third template for inputting information on recommendation items classified into the fourth group when it is determined that the second ratio is lower than the second reference ratio;
providing a work page generated based on the third template to the worker terminal;
When information on recommended products selected for each recommended item is input through the work page generated based on the third template, based on the input recommended product information, when visiting the second place during the first period generating third content suggesting a third outfit that is the outfit; and
If it is confirmed that the second ratio is higher than a preset third reference ratio, further comprising setting the third content as recommended content associated with the first content,
Generating the first template,
If items worn on a first part of the body are classified into the first category and the first item is selected as a recommended item of the first category, generating a first field in the first part;
Generating the first content,
When information on a first product, which is a recommended product of the first item, is input through the first field, based on the information on the first product, whether the first product corresponds to the first item is checked, , When the first product is identified as a product corresponding to the first item, processing an image of the first product to be displayed on the first part,
Receiving first product information including a product name, a seller, an item, a category, a selling price, a sales number, and a marketing amount of the first product from the worker terminal;
collecting customer activity information including a search history of the first item, a review of the first product, and an inquiry about the first product from a sales site that sells the first product;
extracting competitor information of the first company from the first product information when the first product is sold by the first company;
Based on the competitor information, when a competitor of the first company is identified as a second company and it is confirmed that the second company sells a second product as a product corresponding to the first item, the competitor information extracting marketing information of the second product;
generating positive keyword information by extracting positive keywords for the first product from the customer activity information;
setting an activation index of the first product based on an output of the second artificial neural network by applying the customer activity information, the marketing information, and the positive keyword information to a second artificial neural network;
determining a first weight based on an activation index of the first product;
calculating a competitor's average sales volume, which is the average sales volume of the second product, based on the competitor information;
calculating an estimated sales volume of the first product by applying the first weight to the average sales volume of the competitor;
determining a second weight based on the expected sales volume of the first product;
checking a marketing price of the first product based on the first product information;
determining a marketing cost of the first product by applying the second weight to a marketing cost of the first product;
extracting a search time required to search for the first item, a review time required to register a review of the first product, and an inquiry time required to inquire about the first product from the customer activity information;
setting a marketing time for the first product based on an output of the third artificial neural network by applying the search time, the review time, and the inquiry time to a third artificial neural network;
When the marketing time of the first product is set to the first time and the marketing cost of the first product is determined to be the first cost, the marketing cost per hour is applied to the first time and the marketing cost per day is determined as the second cost. , and determining a marketing period of the first product as a first period by dividing the first cost by the second cost;
generating a marketing strategy for the first product including the first cost, the first time, and the first period; and
Further comprising providing a marketing strategy of the first product to the worker terminal,
Keyword extraction and content template generation automation method for commerce-linked content providing platform service. delete delete

Images