KR20200023705A - Method and electric apparatus for recommending jewelry product - Google Patents

Abstract

One embodiment of a method for recommending a jewelry product to a user by a device for recommending a jewelry product comprises the steps of: receiving interest data of a user from the user; applying the interest data to an artificial intelligence nerve network; and displaying data related to a recommended jewelry product outputted from the artificial intelligence nerve network. The artificial intelligence nerve network may be learned for obtaining user analysis data related to taste of the user from the interest data, selecting the recommended jewelry product adapting to the taste of the user based on the user analysis data, and outputting data related to the selected recommended jewelry product.

Description

Jewelry product recommendation method and jewelry product recommendation device {Method and electric apparatus for recommending jewelry product}

The present disclosure relates to a method for recommending jewelry products and an apparatus for recommending jewelry products.

With the development of internet technology, technology for providing a service for recommending products to users by analyzing user's taste is becoming common. For example, a service for recommending a product of interest to a user based on a cookie related to an internet homepage visited by the user is provided.

However, since jewelry products have different user preferences, it is very difficult to recommend an optimal product. In particular, since there are so many kinds of goods, it is more difficult for a user to easily select a desired product. In other words, it is very difficult to grasp the user's intention and match and recommend similar products to the jewelry items of interest to the user.

Accordingly, there is a need for a method and apparatus for easily grasping a user's intention and quickly matching and recommending a product similar to a jewelry item of interest to the user.

The present disclosure is to provide an embodiment of a method and apparatus for easily grasping a user's intention and quickly matching and recommending a product similar to a jewelry item of interest to the user. Technical problems to be achieved by the disclosed embodiments are not limited to the above-described technical problems, and other technical problems may be inferred.

As a technical means for achieving the above technical problem, a method for recommending jewelry products to a user by a jewelry product recommendation device, receiving the interest data of the user from the user, applying the interest data to the artificial intelligence neural network And displaying data related to recommended jewelry items output from the artificial intelligence neural network, wherein the artificial intelligence neural network obtains user analysis data related to the taste of the user from the interest data, and the user analysis data. Based on, it may be learned to select the recommended jewelery items that match the taste of the user, and output the data related to the selected recommended jewelery products.

As a technical means for achieving the above technical problem, the disclosed jewelry product recommendation apparatus, the user input unit for receiving the interest data of the user from the user, a processor for applying the interest data to the artificial intelligence neural network and outputs from the artificial intelligence neural network And a display unit configured to display data related to recommended jewelry items, wherein the artificial intelligence neural network obtains user analysis data related to the user's taste from the interest data, and based on the user analysis data, the user's taste. It may be learned to select the recommended jewelery items corresponding to the, and to output data related to the selected recommended jewelery products.

In addition, as a technical means for achieving the above-described technical problem, the computer-readable recording medium may be a program for executing at least one of the embodiments of the disclosed method in a computer.

In addition, as a technical means for achieving the above-described technical problem, the application stored in the recording medium may be for executing at least one function of the embodiments of the disclosed method.

1 is a diagram illustrating an example of a method of recommending a jewelry item, according to an exemplary embodiment.
2 is a flowchart of a method of recommending a jewelry item, according to an exemplary embodiment.
3-5 illustrate an example of a method of obtaining interest data, according to one embodiment.
6 to 10 are diagrams illustrating input data related to jewelry items, according to an exemplary embodiment.
11 is a flowchart of a method of generating metadata according to an embodiment.
12 is a flowchart of a method of obtaining user analysis data, according to an embodiment.
FIG. 13 illustrates displaying recommended jewelry items according to an embodiment. FIG.
14 is a diagram illustrating displaying attribute data of a recommended jewelry item, according to an exemplary embodiment.
FIG. 15 is a diagram illustrating displaying of text data included in attribute data and user analysis data of a recommended jewelry product, according to an exemplary embodiment.
16 is a block diagram of an apparatus for recommending jewelry items, according to an exemplary embodiment.
17 is a block diagram of a controller of an apparatus for recommending jewelry goods, according to an exemplary embodiment.
18 is a block diagram of a server interworking with an apparatus for recommending jewelry items, according to an exemplary embodiment.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Some embodiments of the present disclosure may be represented by functional block configurations and various processing steps. Some or all of these functional blocks may be implemented in various numbers of hardware and / or software configurations that perform particular functions. For example, the functional blocks of the present disclosure may be implemented by one or more microprocessors or by circuit configurations for a given function. In addition, for example, the functional blocks of the present disclosure may be implemented in various programming or scripting languages. The functional blocks may be implemented in algorithms running on one or more processors. In addition, the present disclosure may employ the prior art for electronic configuration, signal processing, and / or data processing. Terms such as "mechanism", "element", "means" and "configuration" may be used widely and are not limited to mechanical and physical configurations.

Throughout the specification, when a part is "connected" to another part, it includes not only "directly connected" but also "electrically connected" with another element in between. . In addition, when a part is said to "include" a certain component, this means that it may further include other components, except to exclude other components unless otherwise stated.

In addition, the connecting lines or connecting members between the components shown in the drawings are merely illustrative of functional connections and / or physical or circuit connections. In an actual device, the connections between components may be represented by various functional connections, physical connections, or circuit connections that are replaceable or added.

In addition, terms including ordinal numbers such as “first” or “second” as used herein may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

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

1 is a diagram illustrating an example of a method of recommending a jewelry item, according to an exemplary embodiment.

Referring to FIG. 1, in a method of recommending jewelry items, a device (referred to as an electronic device 10) for recommending jewelry items receives user's interest data from the user and meets the user's taste based on the interest data. The recommended jewelery product may be selected, and data related to the selected recommended jewelery product may be output.

Jewelry products represent a set of jewelry (eg rings, earrings, necklaces, bracelets, and anklets) used to embellish your body. For example, jewelry products include jewelry containing jewelry and metal containing jewelry, as well as a series of ornaments made of various materials used to embellish your body.

The interest data of the user refers to information related to the interest of the user. The interests of the user may include fashion items (eg, clothing, jewelry items, bags, hats, etc.) of interest to the user. The data of interest may include text data and image data. The interest data may include data input by the user to the electronic device 10 and data acquired by the user using the electronic device 10.

According to an embodiment, the image data included in the interest data may include a series of photo data and picture data related to a user's interest.

The photo data may include data in the form of photographs taken with respect to the user's interests (for example, fashion products of interest to the user). That is, the photo data may include data in which at least one of clothes, hats, bags, and jewelry items, which are fashion items of interest to the user, is photographed. For example, the photo data may include a product catalog photo, a photo of a model wearing the product, a frame of a program aired on a TV that includes a jewelry item of interest, or a photo containing a jewelry item of interest obtained through the Internet. have. However, the present invention is not limited thereto.

The picture data may include a picture of a user drawing directly on the electronic device 10 or a picture drawn on another medium such as paper using a camera. However, the present invention is not limited thereto.

The electronic device 10 may input the interest data as the training data into the artificial intelligence neural network in order to recommend the recommended jewelry product corresponding to the user's taste.

The artificial intelligence neural network may be trained to obtain user analysis data related to the user's taste from the interest data, select jewelry items suitable for the user's taste as recommended jewelry products, and output data related to the recommended jewelry products. .

The user analysis data refers to data representing a result of analyzing a user's taste. The user analysis data may include text information and image features related to the taste of the user.

According to an embodiment, the user analysis data may be generated using metadata generated based on at least one of text data and image data obtained from the interest data.

Meta data refers to data linked to information obtained from the data of interest in order to analyze the taste of the user. For example, the metadata may include data linking at least one of text information and image features obtained from the data of interest. That is, the metadata may include data in which text information and image features are linked in many-to-many fashion, data linked between text information, and data linked between image features. In addition, the metadata may include text information linked to each other and data regarding weights assigned to image features. In addition, the metadata may be updated by changing weights assigned to text information and image features.

The user analysis data may include text information and image information having a weight greater than or equal to a threshold value among text information and image features included in the metadata.

The artificial intelligence neural network may be trained to receive a plurality of text data and image data as learning data, and to determine a correlation between the text data and the image data. The training data may include text data and image data received from a user. In addition, the training data may include text data and image data acquired by the electronic device 10 through the Internet for learning. The electronic device may crawl text data and image data included in a web page for learning. The text data may include a tag.

The artificial intelligence neural network may receive image data and text data related to each of the plurality of jewelry items as learning data. The artificial intelligence neural network may acquire data related to the attributes of each of the plurality of jewelry items by using the data recognition model of the training data.

For example, the artificial intelligence neural network may acquire data related to attributes of each of a plurality of jewelry items through image analysis using a CNN (Convolution Neural Network) technology.

As another example, the artificial intelligence neural network may acquire data related to data and additional information related to attributes of each of a plurality of jewelry items by analyzing text data and voice data using Recurrent Neural Network (RNN) technology.

The artificial intelligence neural network may link the text data and the image data obtained from the data of interest based on a criterion for determining the relevance of the learned text data and the image data.

The artificial intelligence neural network may be included in the server 20. The electronic device 10 may transmit input data received from the user to the server 20 and receive data related to the recommended jewelry product output from the server.

Alternatively, the artificial intelligence neural network may be included in the electronic device 10. The electronic device 10 may store and use data constituting the artificial intelligence neural network in an internal memory. For example, the electronic device 10 may store and use the artificial intelligence neural network in an internal memory by receiving data constituting the artificial intelligence neural network from the server 20. In addition, the electronic device 10 may receive data for updating the artificial intelligence neural network from the server 20.

The artificial intelligence neural network may exist in the form of a processor. The processor may include at least one universally used processor (eg, a CPU, an application processor) and at least one processor designed to perform a function of an artificial intelligence neural network.

Alternatively, the artificial intelligence neural network may exist in the form of a software module. A processor designed to perform a function of a general purpose processor or an artificial intelligence neural network may use an artificial intelligence neural network in the form of a software module by executing instructions.

The electronic device 10 may display data related to the recommended jewelry product output from the artificial intelligence neural network.

The artificial intelligence neural network may output data related to the selected recommended jewelry product to the electronic device 10. For example, the artificial intelligence neural network may include data related to the appearance of a product, such as a whole photo, partial photo, enlarged photo, and 3D modeling photo of the selected recommended jewelry product, data related to the attributes of the recommended jewelry product, and additional information of the recommended jewelry product. The relevant data may be output to the electronic device 10.

The electronic device 10 may output at least one of data related to the appearance of the recommended jewelry product output from the artificial intelligence neural network, data related to attributes, and data related to additional information. For example, the electronic device 10 may output photo data related to the appearance of the recommended jewelry product by using the display unit. In addition, the electronic device 10 may display text information included in the user analysis data and attribute data of the recommended jewelry product.

According to the disclosed embodiment, the electronic device 10 may easily grasp a user's intention, and quickly and accurately match and recommend a product similar to a jewelry item of interest to the user.

The electronic device 10 may receive a user input for selecting at least one of the output recommended jewelry items. For example, the electronic device 10 may receive a user input for selecting at least one of recommended jewelry items using a touch screen.

The electronic device 10 may indicate that the selected recommended jewelry product is selected using a check box. The electronic device 10 may display the attribute data of the selected recommended jewelry item by overlaying it on the image of the selected recommended jewelry item.

The electronic device 10 may transmit data related to a production request for the selected recommended jewelry product to a terminal of at least one jewelry product manufacturer based on a user input. For example, the electronic device 10 may transmit data related to a production request to a terminal of a jewelry product manufacturer in response to a user input of selecting at least one of the recommended jewelry products. Alternatively, the electronic device 10 may receive a user's input for requesting the production of the recommended jewelry product, which becomes a check box, and transmit data related to the production request to a terminal of the jewelry product manufacturer.

The terminal of a jewelry product manufacturer may include a mobile device (eg, a smartphone, a tablet PC, etc.), a general purpose computer (PC, personal computer), a server (Server), etc., capable of transmitting and receiving data related to an order production request. have.

According to the disclosed embodiment, the electronic device 10 may quickly and accurately match and recommend a product similar to a jewelry item of interest to the user, and may request an order production.

2 is a flowchart of a method of recommending a jewelry item, according to an exemplary embodiment.

Referring to step S210, the electronic device 10 may receive interest data from a user.

According to an embodiment, the interest data may include data acquired by the user using the electronic device 10. For example, the interest data may include data acquired by the user as in the embodiments of FIGS. 3 to 5 using the electronic device 10. Since the user acquires information about the interest through the search for the interest and the visit of the web page, the electronic device 10 may obtain data related to the interest of the user based on the user's behavior.

According to an embodiment, the interest data may include data input by the user to the electronic device 10. For example, the interest data may include data about a jewelry item of interest, such as the embodiment of FIGS. 6 to 10, which the user inputs to the electronic device 10.

Referring to step S230, the electronic device 10 may apply data of interest to an artificial intelligence neural network.

According to one embodiment, the artificial intelligence neural network may be applied to the pre-processed data.

The preprocessing of data may be performed by the electronic device 10 or the server 20. For example, the electronic device 10 may separate and acquire text data and image data from interest data. In addition, the electronic device 10 may extract text data (for example, color, atmosphere, and product attribute) indicating taste from the text data. In addition, the electronic device 10 may extract image data including a fashion product from the image data. In addition, the electronic device 10 may extract each of fashion items from the image data.

As another example, the electronic device 10 may adjust the brightness of at least a portion of the image data. In addition, the electronic device 10 may adjust the sharpness of the image data. In addition, the electronic device 10 may extract data related to a basic form constituting at least one of a jewelry product and a fashion product from the image data. In addition, the electronic device 10 may extract text information from the image data.

As another example, the electronic device 10 may classify the name, article number, text related to the attribute of the jewelry product, and text for describing the jewelry product from the text data and apply the same to the artificial intelligence neural network.

As another example, the electronic device 10 may obtain voice data related to a jewelry product included in the voice data through voice recognition and apply the voice data to an artificial intelligence neural network. In addition, the electronic device 10 may convert the voice data into text data through voice recognition, and apply the converted text data to the artificial intelligence neural network.

According to an embodiment, the electronic device 10 may transmit data input from the user to the server 20. The server 20 may preprocess data received from the electronic device 10. The server 20 may perform preprocessing performed by the electronic device 10 described above. Duplicate content is omitted.

According to an embodiment, the artificial intelligence neural network may preprocess the input data. The artificial intelligence neural network may perform preprocessing performed by the electronic device 10 described above. Duplicate content is omitted.

The artificial intelligence neural network may obtain information related to the taste of the user from the interest data, generate user interest data, and select recommended jewelry items based on the user interest data.

According to an embodiment, the artificial intelligence neural network may be learning a criterion for acquiring and connecting information related to taste using learning data. For example, the artificial intelligence neural network may have learned a criterion for obtaining text data and image data from input data. In addition, the artificial intelligence neural network may have learned a criterion for obtaining text information related to taste from text data. In addition, the artificial intelligence neural network may have learned a criterion for obtaining image characteristics related to taste from image data. In addition, the artificial intelligence neural network may have learned a criterion for determining the relationship between text information and image features. In addition, the artificial intelligence neural network may have learned a criterion for linking text information and image features having relevance.

According to one embodiment, the training data may include data directly input by a user in order to learn a criterion for the artificial intelligence neural network to generate user interest data. The learning data may include data related to text information corresponding to the taste of the user and image data related to fashion goods and jewelry products according to the taste of the user.

According to an embodiment of the present disclosure, the learning data may include text data and image data obtained by crawling by the electronic device 10 and the server 20. For example, it may include image data disclosed in the SNS and text data disclosed (eg, a tag) along with the image data.

According to an embodiment, the artificial intelligence neural network may be trained to generate virtual jewelry items based on attribute data of the jewelry items stored in the database and user interest data. The virtual jewelery product refers to a virtual jewelery product generated by an artificial neural network to match a user's taste. The artificial intelligence neural network may generate a virtual jewelry product based on attribute data similar to the user analysis data. For example, the artificial intelligence neural network may generate a virtual jewelry product corresponding to attribute data similar to the user analysis data. A method of generating a virtual jewelry product will be described in detail below.

The artificial intelligence neural network may be learning data related to a plurality of jewelry products in order to select a recommended jewelry product among a plurality of jewelry products and to output data related to the selected recommended jewelry product to the electronic device 10.

According to an embodiment of the present disclosure, the artificial intelligence neural network may learn attribute data of a plurality of jewelry items and data related to features of appearance. The data related to the attribute data of the plurality of jewelry items and the features of the appearance may be data stored in a database.

According to an embodiment, the attribute data of jewelry items stored in the database may be attribute data of jewelry items obtained from terminals of a plurality of jewelry product manufacturers. For example, the attribute data of the jewelry products included in the database may store data related to the jewelry products manufactured by the company provided by a plurality of jewelry product manufacturers. Alternatively, the attribute data of the jewelry products stored in the database may be data stored by the jewelry product manufacturers, such as the Internet and catalogs, for promotion.

According to an embodiment, the attribute data of the plurality of jewelry items stored in the database may include attribute data distinguished by the artificial intelligence neural network based on the crawled data.

According to an embodiment, the artificial intelligence neural network may classify the attribute data of the jewelry product as shown in Table 1.

For example, the jewelry category may include data related to the type of jewelry product. The stone category may include, but is not limited to, subcategories of stone type, stone color, stone weight, stone shape, and stone setting. . The data related to the stone type may include data related to the type of jewelry, the type of semi-precious stones, the type of substitute jewelry, and the type of pearl. Jewelry may refer to minerals used in jewelry products. Semi-precious stones are similar to gemstones, but may mean that they are rare. The replacement gemstone may mean replacing a plurality of gemstones of a small size with a gemstone of a large size. Pearls may refer to beads mainly composed of calcium carbonate produced in shells. The data related to the stone color may include data related to the color of each mineral. The data related to the stone weight may include data related to the weight of each stone included in the jewel. The data relating to stone weight may also include data relating to the size of each stone included in the jewel. The data related to the stone shape may include data related to the shape of each stone included in the jewel. The data related to the stone setting may include data related to a form and a method of fixing each stone included in the jewel. The metal category may include, but is not limited to, subcategories of metal type (metal type), metal weight, and metal finish (metal finish) included in the jewelery product. The data related to the metal type may include data related to the type of metal included in the jewelry product. The data related to the weight of the metal may include data related to the weight of the metal included in the jewelry product. The data related to the metal finish may include data relating to the finishing technique applied to the surface of the metal included in the jewelry product. The data related to the design includes the clasp type, the length of the string, the chain type, the design type of the jewelry product, and the design shape of the jewelry product. , Data related to a function (e.g., whether the length of a ring, necklace, bracelet can be adjusted, and whether multiple rings can be worn at the same time).

According to an embodiment, the artificial intelligence neural network may store data related to attributes of each of the acquired jewelery items and data related to additional information in a database. The AI neural network can access the data stored in the database.

According to an embodiment, the artificial intelligence neural network may obtain at least one of text data and image data from the data of interest. The artificial intelligence neural network may obtain text information related to the taste of the user from the text data. The artificial intelligence neural network may obtain image information related to a user's taste from the image data. The AI neural network may acquire metadata based on the text information and the image information. A method of generating metadata is described in detail below with reference to FIG. 11. The AI neural network can store metadata in a database.

According to an embodiment, the artificial intelligence neural network may obtain user analysis data from metadata. A method of obtaining user analysis data will be described in detail below with reference to FIG. 12. The AI neural network may store user analysis data in a database.

The artificial intelligence neural network may select recommended jewelry items based on user interest data.

The artificial intelligence neural network may use a conventional personalized recommendation technique such as text data, meta tag information related to a user, and log information related to a user to select a jewelry product that matches a jewelry product of interest as a recommended jewelry product. Through this, the artificial intelligence neural network may provide a customized jewelry product recommendation service.

The artificial intelligence neural network may output data related to the selected recommended jewelry product to the electronic device 10. For example, the artificial intelligence neural network may include data related to the appearance of a product, such as a whole photo, partial photo, enlarged photo, and 3D modeling photo of the selected recommended jewelry product, data related to the attributes of the recommended jewelry product, and additional information of the recommended jewelry product. The relevant data may be output to the electronic device 10.

Referring to step S250, the electronic device 10 may output data related to recommended jewelry items output from the artificial intelligence neural network. The electronic device 10 may output at least one of data related to the appearance of the recommended jewelry product output from the artificial intelligence neural network, data related to attributes, and data related to additional information. For example, the electronic device 10 may include data related to the appearance of the product, such as a whole photo, a partial photo, an enlarged photo, and a 3D modeling photo of the recommended jewelry product, data related to attributes of the recommended jewelry product, and additional information of the recommended jewelry product The data related to may be output using the display unit.

According to an embodiment of the present disclosure, the electronic device 10 may display data related to the recommended jewelery product and text information included in the user analysis data. The electronic device 10 may display text information linked to the image feature included in the image data of the recommended jewelry product together with the data related to the recommended jewelry product. A description with reference to FIG. 15 is as follows.

3-5 illustrate an example of a method of obtaining interest data, according to one embodiment.

The interest data may include information identified by the user with interest. That is, the interest data may include keywords, sentences, paragraphs, phrases, and images searched by the user using the electronic device 10.

In addition, the interest data may include a search result of the user using the electronic device 10. For example, the interest data may include data included in a web page indicating a search result of the user using the electronic device 10. The interest data is a result of the user's search using the electronic device 10, and includes text data (eg, keywords, sentences, paragraphs and phrases), image data (eg, still images and dynamic images) and audio data. It may include at least one of.

In addition, the interest data may include data included in a web page visited by the user using the electronic device 10. In this case, the interest data may include at least one of text data, image data, and voice data included in a web page visited by the user.

In addition, the interest data may include data included in a web page linked to at least one of a web page indicating a search result of the user using the electronic device 10 and a web page visited by the user.

3 illustrates a web page 300 showing a search result of a user using the electronic device 10.

According to an embodiment, the interest data may include a keyword, a sentence, a paragraph, a phrase, and an image searched by the user using the electronic device 10. The user can use the search function to obtain information about the interest. Since the keywords, sentences, paragraph phrases and images input by the user reflect the user's interests, the electronic device 10 uses the keywords, sentences, paragraphs, phrases, and images input by the user as the interest data. Can be obtained. Referring to FIG. 3, for example, the electronic device 10 may obtain the “work daily accessory” input into the search box 310 as the interest data.

According to an embodiment of the present disclosure, the interest data may include data included in a web page indicating a search result of the user using the electronic device 10. Web pages that display user search results are likely to contain information about the user's interests.

For example, the electronic device 10 may obtain text data 330a, 330b, and 330c included in the web page 300 as interest data. The text data 330a, 330b, and 330c may include keywords, sentences, paragraphs, phrases, and text data indicating a web page linked to the web page 300. In addition, the text data 330a, 330b, and 330c may include a tag.

3, for example, the electronic device 10 may obtain image data 350 included in the web page 300 as interest data. The image data 350 may acquire still image data (eg, a picture or a picture) and dynamic image data as the interest data. For example, the electronic device 10 may obtain still image data including the jewelry product as the interest data. In addition, the electronic device 10 may obtain still image data in which a person wears jewelry items as interest data. In addition, the electronic device 10 may obtain still image data including the jewelry product together with the fashion product as the interest data.

4 illustrates a web page 400 visited by a user using the electronic device 10.

According to an embodiment, the interest data may include data included in a web page visited by the user using the electronic device 10. In addition, the interest data may include data included in a web page linked to a web page indicating a search result of the user using the electronic device 10. The user may visit the web page using the electronic device 10 and obtain information about the interest from the visited web page. The web page visited by the user is likely to contain information about the user's interests. The electronic device 10 may obtain, as interest data, data included in a web page visited by the user using the electronic device 10.

The web page 400 may include a web page visited by a user by inputting a web address into an address bar of a web browser. In addition, the web page 400 may include a web page visited by the user using a link included in the web page 300 indicating the search result. In addition, the web page 400 may include a web page visited by the user using a link included in the web page such as a banner advertisement.

For example, the electronic device 10 may obtain a plurality of text data 410a and 410b included in the web page 400 as the interest data. The text data 410a and 410b may include keywords, sentences, paragraphs, phrases, and text data indicating a web page linked to the web page 400. In addition, the text data 410a and 410b may include a tag.

In addition, the electronic device 10 may obtain a plurality of image data 430a and 430b included in the web page 400 as the interest data. Since the image data is similar to the image data described above with reference to FIG. 3, overlapping contents are omitted.

5 illustrates a web page 500 linked to a web page 400 visited by a user using the electronic device 10.

According to an embodiment, the interest data may include data included in a web page linked to a web page visited by the user using the electronic device 10. The web page linked to the web page visited by the user using the electronic device 10 is likely to include information similar to the web page visited by the user. That is, a web page linked to a web page visited by the user is likely to contain information about the user's interests. The electronic device 10 may obtain, as interest data, data included in a web page linked to a web page visited by the user using the electronic device 10.

The web page 500 may include a web page linked directly to a web page 400 visited by a user, a web page linked to a banner advertisement, and a web page linked to a web page linked to the web page 400.

Referring to FIG. 5, the electronic device 10 may obtain a plurality of text data 510a and 510b included in the web page 500 as interest data. Since the text data 510a and 510b are similar to the text data described above with reference to FIGS. 3 and 4, overlapping contents are omitted.

In addition, the electronic device 10 may obtain a plurality of image data 530 included in the web page 500 as interest data. Since the image data is similar to the image data described above with reference to FIGS. 3 and 4, overlapping contents are omitted.

6 to 10 are diagrams illustrating input data of a user related to a jewellery of interest according to an exemplary embodiment.

According to an embodiment of the present disclosure, the interest data may include a user's input regarding jewelry items of interest (hereinafter, referred to as jewelry items of interest). User input data related to the jewelery of interest may include at least one of image data, text data, and voice data. For example, the user's input data may include image data and text data at the same time. In another example, the user's input data may include image data and voice data at the same time.

According to an embodiment, the image data included in the user's input regarding the jewelery of interest may include a series of photo data and picture data related to the jewelery of interest.

The photographic data may include data in a series of photographic forms in which a jewelery product of interest is included, such as a product photograph in which only a jewelery product of interest is photographed and a wearing photo of wearing a jewelry product of interest. However, the present invention is not limited thereto.

For example, the photo data may include a product catalog photo, a photo of a model wearing the product, a frame of a program aired on a TV that includes a jewelry item of interest, or a photo containing a jewelry item of interest obtained through the Internet. have. However, the present invention is not limited thereto.

The picture data may include a picture of a user drawing directly on the electronic device 10 or a picture drawn on another medium such as paper using a camera. However, the present invention is not limited thereto.

Referring to FIG. 6, the electronic device 10 may receive user input data as image data. In detail, the electronic device 10 may receive catalog photo data 610 of the jewelry item of interest as input data of the user. The electronic device 10 may apply the catalog photo data 610 of the product to the artificial intelligence neural network.

According to an embodiment, the artificial intelligence neural network may acquire image features from the photo data 610 by analyzing the photo data 610. The artificial intelligence neural network may obtain attribute data of the jewelry product of interest from the image feature. Specifically, the artificial intelligence neural network may acquire attribute data related to the type of jewelry, the type of stone, the shape of the stone, the setting of the stone, the type of the metal, and the metal finish of the jewelry product of interest. The artificial intelligence neural network may obtain text information of the jewelery of interest from the attribute data of the jewelery of interest.

Referring to FIG. 7, the electronic device 10 may receive user input data as image data. In detail, the electronic device 10 may receive wearing photo data 710 of the jewelery of interest as input data of the user. The electronic device 10 may apply the wearable photo data 710 to the artificial intelligence neural network.

According to an embodiment, the electronic device 10 may perform preprocessing before applying the wearable photo data 710 to the artificial intelligence neural network. For example, the electronic device 10 may adjust the brightness of at least a part of the wearing picture data 710. In addition, the electronic device 10 may adjust the sharpness of the wearing picture data 710. In addition, the electronic device 10 may obtain data related to the type of jewelry item of interest from the wearable photo data 710.

According to an embodiment, the electronic device 10 may apply the wearable photo data 710 on which the preprocessing is performed, to the artificial intelligence neural network. For example, the electronic device 10 may apply only data related to the shape of the jewelry item of interest obtained from the wearable photo data 710 to the artificial intelligence neural network. As another example, the electronic device 10 may apply the processed data to the artificial intelligence neural network by adjusting the brightness, contrast, sharpness, etc. of the wearable photo data 710.

According to an embodiment, the artificial intelligence neural network may perform preprocessing on the applied wearing photo data 710. Since the pre-processing performed by the artificial neural network to the wearing picture data 710 is similar to the pre-processing performed by the electronic device 10 described above, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may acquire image features from the photo data 710 by image analyzing the photo data 710. The artificial intelligence neural network may obtain attribute data of the jewelry product of interest from the image feature. Specifically, the artificial intelligence neural network may acquire attribute data related to the type of jewelry, the type of stone, the shape of the stone, the setting of the stone, the type of the metal, and the metal finish of the jewelry product of interest. The artificial intelligence neural network may obtain text information of the jewelery of interest from the attribute data of the jewelery of interest.

According to an embodiment, the artificial intelligence neural network may acquire attribute data related to the price of the jewelery of interest. For example, the artificial intelligence neural network may obtain data related to a price of a jewelry product including an attribute similar to the jewelry product of interest based on data related to the jewelry product of interest obtained through image analysis.

In detail, the artificial intelligence neural network may acquire data related to a price of a jewelry product including an appearance similar to the jewelry product of interest based on the image feature acquired from the photo data 710. In addition, the artificial intelligence neural network may acquire attribute data related to the price of the jewelry product by searching for the jewelry product including an attribute similar to that of the jewelry product of interest. In another example, the artificial intelligence neural network may be applied to the price of a jewelery product of interest by applying the attribute data of the jewelery product of interest based on the information related to the price of the stone and the information related to the price of the metal (eg, gold and silver). Related attribute data can be obtained. Specifically, the artificial intelligence neural network is interested in a jewelry item of interest, which is a ring of “1.47g” consisting of “14k gold”, by obtaining a price per unit weight of 14k gold using the Internet and applying it to the weight of the jewelry item of interest. It is possible to obtain attribute data related to the price of the jewelry product.

The electronic device 10 may receive user input data as image data and text data. The text data may include, but is not limited to, a jewelry product name that is a name for a product, a product number of a jewelry product, a manufacturer of a jewelry product, text related to an attribute of a jewelry product, and text for describing a jewelry product.

Text related to an attribute of a jewelry item refers to data received as a series of texts representing an attribute of a jewelry item. The text related to an attribute of a jewelry item may include information about an attribute of at least one jewelry item. As a specific example, the text indicating the attributes of a jewelry product may include a 14K Rose gold ring containing a 1 carat diamond in a crown.

The text for describing the jewelry product refers to data received as a series of text including additional information related to the jewelry product. The additional information refers to the name or part number of the jewelry product or does not indicate the attributes of the jewelry product, but refers to information related to the jewelry product. For example, the additional information may include information such as jewelry worn by a specific celebrity, jewelry worn by a specific character in a specific drama, wedding rings of a specific celebrity, and jewelry worn by a specific celebrity at a specific event.

According to an embodiment of the present disclosure, the voice data included in the user's input regarding the jewelery of interest refers to data composed of voices directly input by the electronic device 10 from the user. The voice data may include all of a series of data input by voice related to a jewelry product. For example, the electronic device 10 may include data that receives a voice related to a jewelry item from a user by using the voice input unit. The voice data may include, but is not limited to, a jewelry product name that is a name for a product, a product number of a jewelry product, a manufacturer of a jewelry product, text related to an attribute of the jewelry product, and voice data for describing the jewelry product.

The electronic device 10 may obtain information related to the jewelry product included in the voice data through voice recognition of the voice data received from the user. For example, the electronic device 10 may convert voice data into text data and obtain information related to a jewelry product based on the converted text data.

Referring to FIG. 8, the electronic device 10 may receive, as input data of the user, portrait photo data 810 of jewelry items of interest. In addition, the electronic device 10 may receive text data 830 including additional information related to a jewelry item of interest as input data of a user. The text data 830 may be text data for describing the portrait photo data 810. For example, the text data 830 may be text data such as "the wedding rings of celebrities A and B."

According to an embodiment, the electronic device 10 may apply the portrait photo data 810 and the text data 830 to the artificial intelligence neural network.

According to an embodiment, the electronic device 10 may perform preprocessing before applying the portrait photo data 810 to the artificial intelligence neural network. Since the preprocessing performed by the electronic device 10 on the portrait photo data 810 is similar to the preprocessing described above with reference to FIG. 7, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may perform preprocessing on the portrait photo data 810. Since the pre-processing performed by the artificial intelligence neural network to the portrait photograph data 810 is similar to the pre-processing performed by the electronic device 10 described above, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may acquire an image feature from the portrait photo data 810 by analyzing the portrait photo data 810. For example, the artificial intelligence neural network may acquire image characteristics of each of the plurality of subjects included in the portrait photo data 810. Here, the plurality of subjects may include at least one person, at least one fashion item, and at least one jewelry item. The artificial intelligence neural network may distinguish each subject from other subjects based on the acquired image characteristic. The artificial intelligence neural network may acquire image features related to attributes of each of the subjects. The AI neural network may acquire attribute data of each subject based on the image feature. For example, the artificial intelligence neural network may recognize the jewelery of interest based on the type of the jewelery of interest included in the portrait photography data 810. For another example, the artificial intelligence neural network may recognize a jewelry item of interest based on the position of the jewelry item of interest included in the portrait photography data 810.

According to an embodiment, the artificial intelligence neural network may obtain attribute data of a jewelery product of interest from an image feature. Specifically, the artificial intelligence neural network may acquire attribute data related to the type of jewelry, the type of stone, the shape of the stone, the setting of the stone, the type of the metal, and the metal finish of the jewelry product of interest. The artificial intelligence neural network may obtain text information of the jewelery of interest from the attribute data of the jewelery of interest.

According to an embodiment, the artificial intelligence neural network may obtain text information from the text data 830. The artificial intelligence neural network may obtain text information related to an attribute of the jewelry product of interest from the text data 830. In addition, the artificial intelligence neural network may include additional information of the jewelry item of interest from the text data 830. Specifically, the text information `` celebrities A and B '', "marriage" and "ring" can be obtained from the text data 830 of "the wedding rings of celebrities A and B".

According to an embodiment of the present disclosure, the text data included in the user's input regarding the jewelry item of interest refers to data composed of text directly input by the electronic device 10 from the user. The text data may include a series of data input as text related to a jewelry product. For example, the electronic device 10 may receive an input from a user in a typing method or a handwritten method. The electronic device 10 may recognize the text data input by a handwritten method.

According to an embodiment, the image data included in the user's input regarding the jewelry item of interest may include text information. For example, the text information included in the image data may include text information written on the image by hand, text information described through typing on the image, and text information included in the metadata of the image. The metadata of the image refers to data including separate information, not data constituting the image itself.

Referring to FIG. 9, input data of a user may be received as image data. For example, the electronic device 10 may receive as picture data 910 depicting a jewelry item of interest. In detail, the electronic device 10 may receive data related to a picture described by the user as the electronic device 10 as the picture data 910. The electronic device 10 may apply the picture data 910 to the artificial intelligence neural network.

According to an embodiment, the electronic device 10 may perform preprocessing before applying the picture data 910 to the artificial intelligence neural network. Since the preprocessing performed by the electronic device 10 on the picture data 910 is similar to the preprocessing described above with reference to FIG. 7, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may perform preprocessing on the picture data 910. Since the pre-processing performed by the artificial intelligence neural network on the picture data 910 is similar to the pre-processing performed by the electronic device 10 described above, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may acquire image features from the image data 910 by image analysis of the image data 910. The artificial intelligence neural network may obtain attribute data of the jewelry product of interest from the image feature. Specifically, the artificial intelligence neural network may acquire attribute data related to the type of jewelry, the type of stone, the shape of the stone, the setting of the stone, the type of the metal, and the metal finish of the jewelry product of interest.

Specifically, the artificial intelligence neural network sets the attribute data related to the type of jewelry of the jewelry product of interest as “ring”, the attribute data related to the type of stone is “none”, and the attribute data related to the type of metal is “14k rose gold”, Attribute data related to metal finish can be obtained as “tangled” or “polish”. The artificial intelligence neural network may obtain text information of the jewelery of interest from the attribute data of the jewelery of interest.

Referring to FIG. 10, the electronic device 10 may receive user input data as image data and text data. In detail, the electronic device 10 may receive, as input data of the user, picture data 1010 depicting a jewelry item of interest. In addition, the electronic device 10 may receive text data 1030 including data related to an attribute of the jewelery of interest as input data of the user. For example, the text data 1030 may include information that the type of stone of the jewelery product of interest is “topaz”.

According to an embodiment, the electronic device 10 may apply the picture data 1010 and the text data 1030 to the artificial intelligence neural network.

According to an embodiment, the electronic device 10 may perform preprocessing before applying the picture data 1010 to the artificial intelligence neural network. Since the preprocessing performed by the electronic device 10 on the picture data 1010 is similar to the preprocessing described above with reference to FIG. 7, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may perform preprocessing on the picture data 1010. Since the pre-processing performed by the artificial intelligence neural network to the picture data 1010 is similar to the pre-processing performed by the electronic device 10 described above, overlapping contents are omitted.

According to an embodiment, the artificial intelligence neural network may acquire image features from the image data 1010 by analyzing the image data 1010. The artificial intelligence neural network may obtain attribute data of the jewelry product of interest from the image feature. The method of acquiring the attribute data of the jewelery of interest by using the image feature of the picture data 1010 is similar to the method of acquiring the attribute data of the jewelery of interest by using the image feature of the photo data described above. Omit.

The artificial intelligence neural network may obtain text information from the text data 1030. For example, the artificial intelligence neural network may obtain attribute data related to the type of stone as “topaz”.

11 is a flowchart of a method of generating metadata according to an embodiment. The artificial intelligence neural network may perform only at least one of steps S1130 and S1150.

Referring to step S1110, the artificial intelligence neural network may obtain at least one of text data and image data from the input interest data.

For example, the artificial intelligence neural network may obtain at least one of a keyword, a sentence, a paragraph, and a phrase input to the search box 310 of FIG. 3 as text data. The artificial intelligence neural network may acquire text data 330a, 330b, and 330c included in the web page 300 representing the search result as the interest data. The artificial intelligence neural network may acquire text data 410a and 410b included in the web page 400 visited by the user. The artificial intelligence neural network may acquire text data 510a and 510b included in the web page 500 linked to the web pages 300 and 400. The artificial intelligence neural network may obtain text data 830 and 1030 from the interest data received from the user as shown in FIGS. 8 and 10.

As another example, the artificial intelligence neural network may acquire an image input to the search box 310 of FIG. 3 as image data. The artificial intelligence neural network may acquire image data 350 included in the web page 300 representing the search result of the user. The artificial intelligence neural network may acquire image data 430a and 430b included in the web page 400 visited by the user. The artificial intelligence neural network may acquire image data 530 included in the web page 500 linked to the web pages 300 and 400. The artificial intelligence neural network may acquire image data 610, 710, 810, 910, and 1010 received from a user as illustrated in FIGS. 6 to 10.

Referring to step S1130, the artificial intelligence neural network may obtain text information from the text data.

In this case, the text information means information in a series of texts related to taste. For example, the text information may include a noun representing a name of a product related to a taste, a proper noun representing a brand, an adjective and an adverb representing a taste. In addition, the text information may include a tag assigned for searching.

For example, the AI neural network may be “worker,“ daily ”,“ accessory ”,“ simple ”,“ basic ”,“ gold ”,“ 14k ”,“ gold ”from text data (311, 330a, 330b, 330c). You can get text information such as "earing", "no clutter", and "clean". In addition, the artificial intelligence neural network may obtain text information such as “moon” and “smile” from the text data 410a and 410b. In addition, the artificial intelligence neural network may obtain text information of “rose gold”, “gold”, “silver”, and “earring” from the text data 510a and 510b. In addition, the artificial intelligence neural network may obtain text information of “celebrity A”, “celebrity B”, “marriage”, and “ring” from the text data 830 about the jewelry item of interest. In addition, the artificial intelligence neural network may obtain text information regarding the stone type of the jewelry item called “topaz” from the text data 1030 about the jewelry item of interest.

Referring to step S1150, the artificial intelligence neural network may obtain image information from the image data.

According to an embodiment, the artificial intelligence neural network may obtain image features related to taste from image data. An image feature refers to an image form consisting of at least one color, a point, a line, a face, and a combination thereof related to taste. For example, image features may comprise points, lines, faces, and combinations thereof and at least one color (e.g., polygon, circle, star, sea, moon, etc.) that form a particular shape (e.g., Red).

The artificial intelligence neural network may acquire image features of each of the plurality of objects included in the image data. The plurality of subjects may include at least one person, at least one fashion item, and at least one jewelry item. The AI neural network can acquire image features related to the properties of each of the objects. In detail, the artificial intelligence neural network may obtain at least one of an image feature of a jewelry product, an image feature of a person wearing the jewelry product, and an image feature of a fashion product worn with the jewelry product from the image data.

The image feature of the jewelry product may include a feature related to the appearance of the jewelry product, a feature related to a color of the jewelry product, and a stone, metal, and design related to the jewelry product.

The image features of the person who wears the jewelry item may include a feature related to the appearance and size of the person who wears the jewelry item, a feature related to the position where the jewelry item is worn, and a feature related to the fashion item worn with the jewelry item. have.

The image feature of the fashion product may include a feature related to the appearance of the fashion product, a feature related to the color of the fashion product, and a feature related to a material forming the fashion product.

Referring to step S1170, the artificial intelligence neural network may generate metadata by linking text information and image features.

According to an embodiment, the artificial intelligence neural network may generate metadata by linking the text information obtained in operation S1130 with the image feature obtained in operation S1150. The AI neural network weights a plurality of linked image features and text information, thereby increasing the accuracy of the link between the image feature and the text information.

For example, the AI neural network may be “worker,“ daily ”,“ accessory ”,“ simple ”,“ basic ”,“ gold ”,“ 14k ”,“ obtained from text data 311, 330a, 330b, 330c ”. along with text information such as gold, “earing”, “no clutter”, and “clean” and image features about jewelry products obtained from image data 350, image features about people wearing jewelry products, and jewelry products Many-to-many image features about worn fashion items can be linked.

In another example, the artificial intelligence neural network may include text information of the "rose gold", "gold", "silver", and "earring" obtained from the text data 510a and 510b, and image features obtained from the image data 530. You can link.

In another example, the artificial intelligence neural network may be formed from text information and image data 810 of "celebrity A", "celebrity B", "marriage", and "ring" obtained from text data 830 of a jewelry item of interest. The acquired image features may be linked.

As another example, the artificial intelligence neural network may link text information and image features obtained from the attribute data of the jewelry product. The attribute data of the jewelry item may be obtained from an image feature of the jewelry item.

According to an embodiment, the artificial intelligence neural network may generate and update metadata by weighting text information and image features. The artificial intelligence neural network may weight the text information when text information similar to the text information included in the metadata is obtained from the data of interest. The artificial intelligence neural network may weight the image feature when an image feature similar to the image feature included in the metadata is obtained from the data of interest. The artificial intelligence neural network may weight the linked text information and the image feature when a link similar to the link between the text information and the image feature included in the metadata is obtained from the data of interest.

According to an embodiment, the artificial intelligence neural network may store the generated metadata in a database. The AI neural network may update the metadata stored in the database by storing the generated metadata in the database.

According to an embodiment, the artificial intelligence neural network may generate the second metadata by linking the text information and the image feature based on the first metadata stored in the database. The artificial intelligence neural network may link the text information and the image feature acquired from the data of interest based on the link between the text information and the image feature included in the first metadata. For example, an AI neural network may be of interest based on first metadata including textual information of "daily" and "simple", image characteristics of "silver" and image features of "polish". The second metadata may be generated by linking image features related to "daily" and "polish" which are text information obtained from the data. The artificial intelligence neural network may update the first metadata by weighting an image feature regarding “daily” and “polish” of the first metadata based on the second metadata.

According to an embodiment, the artificial intelligence neural network may obtain attribute data of the jewelry product from the image feature. The artificial intelligence neural network may obtain text information of the jewelry product from the acquired attribute data. The AI neural network can generate metadata by linking image features and textual information.

According to an embodiment, the artificial intelligence neural network may link text information and image features related to jewelry items and fashion items to each other. For example, the artificial intelligence neural network may obtain a first image feature related to the jewelry product and a second image feature related to the fashion product from image data including at least one jewelry product and at least one fashion product. Also, the artificial intelligence neural network may obtain first text information related to the jewelry product and second text information related to the fashion product from the at least one text data. The artificial intelligence neural network may link the first image feature and the first text information, and may link the second image feature and the second text information. The artificial intelligence neural network may link the first image feature and the second image feature, and link the first text information and the second text information. That is, the artificial intelligence neural network may link a plurality of text information and image features in many to many.

The AI neural network can store metadata in a database.

The AI neural network can update data stored in the database. For example, the artificial intelligence neural network may update metadata stored in a database. The artificial intelligence neural network may update metadata stored in a database by changing weights of text information and image features included in the metadata. In addition, the artificial intelligence neural network may update data and additional information related to attributes of each of the plurality of jewelry items stored in the database.

12 is a flowchart of a method of obtaining user analysis data, according to an embodiment.

Referring to step S1210, the artificial intelligence neural network may update the weight of at least one of the text information and the image feature included in the metadata based on the interest data (S1210). Since the method for updating the weight of the artificial intelligence neural network has been described above with reference to FIG. 11, redundant descriptions thereof will be omitted.

Referring to step S1230, the artificial intelligence neural network may acquire at least one of text information and image features having a weight greater than or equal to a threshold value as user analysis data.

For example, the artificial intelligence neural network may acquire text information “daily” having a weight of five or more times as user analysis data. In addition, the artificial intelligence neural network may acquire image features linked with the text information "daily" as user analysis data. In addition, the artificial intelligence neural network may acquire the text information “simple” linked to the text information “daily” as user analysis data.

According to an embodiment, when the weight of each of the linked text information and the image feature is greater than or equal to a threshold value, the artificial intelligence neural network may acquire the linked text information and the image feature as user analysis data.

According to an embodiment, when the weight of the link between the text information and the image feature is greater than or equal to a threshold, the artificial intelligence neural network may acquire the linked text information and the image feature as user analysis data.

The artificial intelligence neural network may select recommended jewelry items based on user interest data.

According to an embodiment, the artificial intelligence neural network selects attribute data similar to the user analysis data and recommends a jewelry product matching the selected attribute data by comparing the similarity between the attribute data of the jewelry items stored in the database with the user analysis data. You can choose as a jewelry item.

According to an embodiment, the artificial intelligence neural network compares the similarity between the data related to the attribute of the jewelry item of interest of the user obtained from the user and the attribute data of the plurality of jewelry items included in the database, thereby at least matching the jewelry item of interest. One jewelry item can be selected as the recommended jewelry item.

The artificial intelligence neural network may update the similarity between the attributes of the plurality of jewelry items stored in the database and the text information included in the user analysis data. In addition, the artificial intelligence neural network may update the similarity between the image feature included in the image data of the plurality of jewelry items stored in the database and the image feature included in the user analysis data.

According to an embodiment, the artificial intelligence neural network may generate a virtual jewelry product based on the user analysis data, and select the virtual jewelry product as the recommended jewelry product. The artificial intelligence neural network may select a virtual jewelry product corresponding to attribute data similar to the user analysis data among the virtual jewelry products as the recommended jewelry product.

The artificial intelligence neural network may generate a virtual jewelry product in which at least one of a category of attribute data similar to the user analysis data (eg, jewelry, stone, metal, design), or at least one of a sub category is similar. Alternatively, the artificial intelligence neural network may generate a virtual jewel product in which at least one of categories of attribute data similar to the user analysis data is dissimilar.

The artificial intelligence neural network may generate a virtual jewelry item by changing at least one of attribute data of jewelry items stored in a database. For example, the artificial intelligence neural network may generate a virtual jewelry product by changing a stone of one of ring products stored in a database. Specifically, the artificial intelligence neural network may generate a virtual jewelery product in which the stone of the ring product is changed from emerald to sapphire. As another example, the artificial intelligence neural network may generate a virtual jewelery product in which the metal of the ring product is changed from silver to 14K Rose Gold. In another example, an artificial neural network can create a virtual jewelery product that changes the metal finish of a ring product from polished to milgrain. Here, an example of generating a virtual jewelry product by changing one of the attribute data of the jewelry product is clear, but it is obvious that the virtual jewelry product can be generated by changing two or more attribute data.

The artificial intelligence neural network may extract at least one feature from the attribute data of the jewelry items stored in the database, and generate a virtual jewelry item based on the extracted feature. For example, the artificial intelligence neural network may generate virtual jewelery products based on external features of jewelery products stored in a database using a GAN (Generative Adversarial Network) technology.

Specifically, the artificial intelligence neural network includes the type of stone jewelry, stone color, stone weight and size, stone shape, stone setting, metal type, metal finish, clasp type, string length and string type of jewelry products stored in the database. The external features of the same jewelry product may be learned, and a virtual jewelry product similar to the image feature included in the user analysis data may be generated.

In addition, the artificial intelligence neural network may weight each of at least one data included in the attribute data of the jewelry items stored in the database, and generate a virtual jewelry item based on the weight. For example, an AI neural network may assign different weights to each of the stones, metals, and designs. In addition, the AI neural network may assign different weights to stone shapes and stone settings. Specifically, the artificial intelligence neural network can give a high weight to the stone shape of the ring product. In addition, the artificial intelligence neural network can give a high weight to the metal type of the ring product. The artificial intelligence neural network can learn the external features of the jewelry product based on the assigned weights, and generate the virtual jewelry product.

The artificial intelligence neural network can learn by matching the appearance characteristics and attribute data of jewelry products stored in the database. For example, an artificial intelligence neural network can learn by matching attribute data for which the metal finish of the ring product is polished with the external characteristics of the metal part of the ring product. The artificial intelligence neural network may also generate attribute data corresponding to the external characteristics of the virtual jewelry item when the virtual jewelry item is generated.

The artificial intelligence neural network may generate a virtual jewelry item similar to the jewelry item of interest based on an external feature of the jewelry item of interest. The artificial intelligence neural network may generate a virtual jewelry product having a similar appearance to the jewelry product of interest based on the result of learning the external features of the jewelry products stored in the database. The artificial intelligence neural network may generate a virtual ring product in which some of the external features of the jewelry product of interest are changed based on the learning result.

For example, the artificial intelligence neural network can generate a virtual ring product by changing the shape of the stone of the ring, the jewelry item of interest, from the heart shape to the Pear shape. In another example, the AI neural network can create a virtual ring product by changing the stone setting of a ring of jewelry items of interest from Prong Set to Bezel Set.

AI neural networks have a similar appearance to the jewelry products of interest, but can produce virtual jewelry products made of less expensive materials. For example, the artificial intelligence neural network can generate a virtual ring product by changing a stone of a ring of interest jewelry product into diamonds and cubic. In another example, an artificial neural network can create a virtual ring product that changes the metal of a jewelery ring of interest from 14K Platinum to Silver.

The artificial intelligence neural network may select attribute data similar to the user analysis data among the attribute data of the jewelry items stored in the database.

According to an embodiment, the artificial intelligence neural network may select attribute data similar to the user analysis data by comparing the image feature included in the user analysis data with the image feature of the jewelry items stored in the database. The artificial intelligence neural network may weight the image features included in the user analysis data. The AI neural network can give high priority to the high weight image features and compare the high priority image features.

According to an embodiment, the artificial intelligence neural network may search for an image feature linked to predetermined text information and output image data including the retrieved image feature. For example, the AI neural network may obtain text information of "modern" "14K" "daily" "rose gold" "ring" from text data of "modern 14K daily rose gold ring". The AI neural network features image features linked to “rings”, image features linked to “Rose Gold”, image features linked to “Daily”, image features linked to “14K” and image features linked to “Modern”. The image data of the jewelry item including the retrieved image feature may be selected. At this time, the artificial intelligence neural network may select image data corresponding to the intersection of the retrieved image features. The artificial intelligence neural network may output data related to an attribute of a jewelry product related to the selected image data.

According to an embodiment, the artificial intelligence neural network obtains attribute data of a jewelery product of interest from an image feature included in the user analysis data, and compares the attribute data of the jewelery product of interest with the attribute data of the jewelry items stored in the database. You can select attribute data similar to the analytical data. The artificial intelligence neural network may apply weights to the attribute data of the acquired jewelery of interest.

The artificial intelligence neural network may weight each attribute data based on the input data. For example, the AI neural network can give high weight to the shape of the stone, the setting of the stone, and the metal finish. As another example, when the input data is a product catalog photograph, the artificial intelligence neural network may weight the stone type, the shape of the stone, and the setting of the stone among the attribute data. In another example, an artificial intelligence neural network may have text data such as “a ring of 14K Rose gold with a 1 carat diamond in a crown”. Among the attribute data, the type of stone, the weight of the stone, the stone setting, The metal type can be weighted.

The artificial intelligence neural network may further weight the at least one of attribute data and additional information obtained from the text data when the image data and the text data are input together.

The AI neural network may compare the similarity based on the attribute data of the jewelery product of interest to which the weight is applied. The AI neural network may give priority to high weighted attribute data. AI neural networks can compare high priority attribute data.

According to an embodiment, the artificial intelligence neural network may compare the similarity based on the data included in the additional information. For example, an artificial intelligence neural network compares the information on a wedding ring of a particular celebrity with the additional information of each of a plurality of jewelery stored in a database based on the additional information of the wedding ring of a specific celebrity. Jewelry products that match to may be selected as recommended jewelry products.

The artificial intelligence neural network may select a jewelry item corresponding to attribute data similar to the user analysis data as the recommended jewelry item. The artificial intelligence neural network may output data related to the recommended jewelry product to the electronic device 10. For example, an artificial intelligence neural network may be related to data related to the appearance of a product, such as a whole photo, partial photo, enlarged photo, or 3D modeled photo of a recommended jewelry product, data related to an attribute of the recommended jewelry product, and additional information of the recommended jewelry product. Data may be output to the electronic device 10.

FIG. 13 is a diagram illustrating that the electronic device 10 displays recommended jewelry items according to an embodiment.

The electronic device 10 may output data related to the recommended jewelry product output from the artificial intelligence neural network.

The electronic device 10 may display data related to the recommended jewelry product in a pop-up window. The electronic device 10 may display data related to the recommended jewelry product in a push manner.

In addition, referring to FIG. 13, the electronic device 10 may display data 1310a, 1310b, 1310c, and 1310d related to the recommended jewelry product on at least a portion of the web page 1300 visited by the user. For example, the data 1310a, 1310b, 1310c, and 1310d related to the recommended jewelry item may include at least a portion of image data including the appearance of the jewelry item, attribute data of the recommended jewelry item, and additional information of the recommended jewelry item. It may include. The electronic device 10 may display text information of the user analysis data similar to the attribute data of the recommended jewelry product.

According to an embodiment of the present disclosure, the electronic device 10 may display the recommended jewelry products in ascending order of the attribute data similar to the user analysis data.

For example, the electronic device 10 may display the recommended jewelry products in ascending order of the attribute data values similar to the user analysis data. For another example, the electronic device 10 may display the recommended jewelry products in the order in which the appearance of the image features having high weights are similar to the appearances.

According to an embodiment of the present disclosure, the electronic device 10 may display at least one of a whole photo, a partial photo, an enlarged photo, and a 3D modeling photo of the recommended jewelry product.

14 is a diagram illustrating displaying attribute data of a recommended jewelry item, according to an exemplary embodiment.

According to an embodiment of the present disclosure, the electronic device 10 may output data related to the selected recommended jewelry item based on a user input for selecting one of the displayed plurality of recommended jewelry items. The electronic device 10 may output data related to the appearance of the recommended jewelry product and data related to the attributes of the recommended jewelry product. In addition, the electronic device 10 may also display data related to additional information of the recommended jewelry product.

Referring to FIG. 14, the electronic device 10 may display image data 1410 related to the appearance of the selected recommended jewelry product and data 1411 related to the attribute of the product. Specifically, the electronic device 10 together with the image data 1410, the product name of the recommended jewelery product, the type of jewelry "ring", the type of stone "pearl", the color of the stone "white", the appearance of the stone "round", Setting type “Prong set”, Metal type “14K white gold”, Metal color “white”, Metal finish “Polish”, Jewelry weight “1 / 5ct. t.w ”, type of pearl“ south sea ”, size of pearl“ 11.0 ~ 12.0mm ”, color of pearl“ white ”can be displayed together with data 1411 related to the attributes of recommended jewelry products.

According to an embodiment of the present disclosure, the electronic device 10 may display an indication that a changeable attribute exists among the attributes of the recommended jewelry product. In addition, the electronic device 10 may also display a changeable property. The electronic device 10 may display the attribute data related to the recommended jewelry product in which the changed attribute is reflected, based on a user's input for changing the attribute. The electronic device 10 may display image data in which the changed property is reflected.

FIG. 15 is a diagram illustrating displaying of text data included in attribute data and user analysis data of a recommended jewelry product, according to an exemplary embodiment.

Referring to FIG. 15, the electronic device 10 may display image data related to the appearance of the recommended jewelry items 1511, 1513, and 1515 and text data 1521, 1523, and 1525 related to the recommended jewelry item.

The text data related to the recommended jewelry product may include text data included in attribute data and user analysis data of the recommended jewelry product. For example, the text data 1521 related to the recommended jewelry product 1511 may include attribute data of products such as "silver", "cubic" and "heart" and text information such as "modern" and "daily". Can be. As another example, the text data 1523 related to the recommended jewelry item 1513 may include attribute data such as "gold", "cubic", "crown", and text information such as "splendor" and "daily". Can be. As another example, the text data 1525 related to the recommended jewelry item 1515 includes attribute data such as "rose gold", "daisy", "cubic" and text information such as "chain" and "daily". can do.

The electronic device 10 may make a production order for the selected recommended jewelry item based on a user's input for selecting one of the output recommended jewelry items.

The electronic device 10 may receive a user input for selecting one of the recommended jewelry items output in step S250. For example, the electronic device 10 may receive a user input for selecting at least one of recommended jewelry items using a touch screen. The electronic device 10 may indicate that the selected recommended jewelry product is selected using a check box. The electronic device 10 may display the attribute data of the selected recommended jewelry item by overlaying it on the image of the selected recommended jewelry item.

The electronic device 10 may transmit data related to a request for manufacturing a selected jewelry item based on a user input to a terminal of at least one jewelry product manufacturer. For example, the electronic device 10 may transmit data related to a production request to a terminal of a jewelry product manufacturer in response to a user input of selecting at least one of the recommended jewelry products. Alternatively, the electronic device 10 may receive a user's input for requesting the production of the recommended jewelry product, which becomes a check box, and transmit data related to the production request to a terminal of the jewelry product manufacturer.

The electronic device 10 may transmit data related to a production request for the selected recommended jewelry product to a terminal of a manufacturer of the selected recommended jewelry product. Alternatively, the electronic device 10 may transmit data related to a production request for a selected jewelry item selected by terminals of a plurality of jewelry product manufacturers.

The data related to the production request for the jewelry product may include attribute data of the selected jewelry product. In addition, the data related to the production request for jewelry products may include data for requesting a quotation.

The electronic device 10 may transmit data for requesting a quotation request relating to a production request for a virtual jewelry product to terminals of a plurality of jewelry product manufacturers. Specifically, data regarding whether the manufacturer can manufacture the selected jewelry product, how much is the order production cost, and whether there is a changeable option among the attributes of the jewelry product may be included in the data relating to the production request for the jewelry product.

Alternatively, the data related to the production request for the jewelry product may include data related to the order and payment. For example, when the user selects a ready-made product, the electronic device 10 may transmit data related to the order and payment to the terminal of the jewelry product manufacturer so that the order and payment may be performed without request for quotation request.

The electronic device 10 may transmit data related to the selected jewelry product to a terminal of the jewelry product manufacturer. For example, the electronic device 10 may transmit attribute data of the selected jewelry product to a terminal of the jewelry product manufacturer. In addition, the electronic device 10 may transmit data for requesting a quotation.

The electronic device 10 may receive quotation data from a terminal of a jewelry product manufacturer. For example, the electronic device 10 may provide quotation data including at least one of data on whether the manufacturer can manufacture the selected jewelry product, how much is the order production cost, and whether there is a changeable option among the attributes of the jewelry product. Can be received.

The electronic device 10 may output the received quotation data and data related to the jewelry product manufacturer that transmitted the quotation data. For example, the electronic device 10 may display data related to a changeable option among a name of a jewelry product manufacturer who sent the quotation data, an order production cost, and an attribute of the jewelry product. In addition, the electronic device 10 may display data related to the user's reviews and the reliability of the jewelry product manufacturer who have made an order to the jewelry product manufacturer. In addition, the electronic device 10 may display data related to a plurality of jewelry product manufacturers so that a user can easily compare the data.

The electronic device 10 may receive a user input for selecting one of the at least one quotation data. In this case, the received user's input may include a user's input for requesting a manufacturing request to the manufacturer who transmitted the selected quotation data.

The electronic device 10 may generate a contract related to the production of jewelry products corresponding to the selected quotation data. In addition, the electronic device 10 may display a contract. In addition, the electronic device 10 may transmit the contract to the terminal of the jewelry product manufacturer that transmits the selected quotation data. In addition, the electronic device 10 may store the generated contract in a memory or a database. In addition, the electronic device 10 may store the generated contract in a database of a server connected to the electronic device.

The electronic device 10 may receive a user input for changing an attribute of the recommended jewelry product from the user. For example, the electronic device 10 may receive a user input for changing an attribute of the output recommended jewelry product. As another example, the electronic device 10 may receive a user input for changing an attribute of the recommended jewelry product included in the quotation data.

The electronic device 10 may receive a user's input for changing an attribute as shown in Table 1 of the recommended jewelry product from the user. For example, the electronic device 10 may receive a user input of changing a stone type from emerald to sapphire. As another example, the electronic device 10 may receive a user input of changing a stone shape from a heart shape to a heart shape. As another example, the electronic device 10 may receive a user input of changing a metal type from platinum to silver.

The electronic device 10 may transmit data related to the recommended jewelry product having changed attributes to the terminal of the jewelry product manufacturer. Since it is similar to step S1231, overlapping contents are omitted.

The electronic device 10 may receive quotation data of a recommended jewelry product having a changed attribute from a terminal of the jewelry product manufacturer. Since it is similar to step S1233, overlapping contents are omitted.

According to the disclosed embodiment, a user can quickly and cheaply purchase a product similar to a jewelery of interest. In addition, a jewelry product manufacturer may easily receive a manufacturing request related to a jewelry product of interest to a user.

16 is a block diagram of an apparatus for recommending jewelry items, according to an exemplary embodiment.

Referring to FIG. 16, an electronic device 10 according to some embodiments may include a user input unit 11, an output unit 12, a processor 13, a communication unit 15, and a memory 17. However, not all components illustrated in FIG. 16 are essential components of the electronic device 10. The electronic device 10 may be implemented by more components than those illustrated in FIG. 16, or the electronic device 10 may be implemented by fewer components than the components illustrated in FIG. 16.

The user input unit 11 means a means for a user to input data for controlling the electronic device 10. For example, the user input unit 11 includes a key pad, a dome switch, a touch pad (contact capacitive type, pressure resistive layer type, infrared sensing type, surface ultrasonic conduction type, and integral type). Tension measurement method, piezo effect method), a touch screen, a jog wheel, a jog switch, and the like, but are not limited thereto.

The user input unit 11 may receive input data of a user related to a jewelery of interest. The user input unit 11 may receive input data of a user who selects at least one of recommended jewelry items. The user input unit 11 may receive user input data for changing an attribute of the recommended jewelry product. Since the input data of the user has been described above, overlapping contents are omitted.

The output unit 12 displays and outputs information processed by the electronic device 10. The output unit 12 may include a display unit 12-1 displaying data related to recommended jewelry items. For example, the display unit 12-1 may include data related to appearance such as a whole photo, partial photo, enlarged photo, and 3D modeling photo of recommended jewelry products, data related to attributes of recommended jewelry products, and additional information of recommended jewelry products. It may display at least one of the data related to. The display unit 12-1 may display a user interface for receiving user input data related to a jewelery of interest. In addition, the display unit 12-1 may display at least one quotation data and data related to the manufacturer who transmitted the quotation data.

The processor 13 typically controls the overall operation of the electronic device 10. For example, the processor 13 may control the user input unit 11, the output unit 12, the communication unit 15, the memory 17, and the like by executing programs stored in the memory 1700. . For example, the processor 13 may control at least one of the user input unit 11 and the communication unit 15 to receive data of interest of the user. The processor 13 may control the display unit 12-1 to output data related to the recommended jewelry product output from the artificial intelligence neural network.

The processor 13 may perform the function of the artificial intelligence neural network described above with reference to FIGS. 1 to 15. The processor may include at least one processor designed to perform a function of an artificial intelligence neural network. Alternatively, the processor may be at least one processor used universally. The processor 13 may perform a function of an artificial intelligence neural network existing in the form of a software module by executing a series of instructions stored in the memory 17.

The processor 13 may control the communication unit 15 to perform a function of the artificial intelligence neural network stored in the server 20. The processor 13 may control the communicator 15 to transmit the data of interest to the server 20 in order to perform the function of the artificial intelligence neural network. The processor 13 may control the communication unit 15 to receive data output from the artificial intelligence neural network stored in the server 20.

The processor 13 may control the communication unit 15 to receive data for updating the artificial intelligence neural network from the server 20.

The processor 13 may execute a series of instructions for the artificial intelligence neural network to learn new training data.

The communication unit 15 may include one or more components that allow the electronic device 10 to communicate with other devices (not shown) and the server 20. The other device (not shown) may be a computing device such as the electronic device 10, but is not limited thereto.

In addition, the communication unit 15 may transmit / receive information necessary for performing an operation of selecting a jewelry product matching the jewelry item of interest as the recommended jewelry item with another device (not shown) and the server 20.

The memory 17 may store a program for processing and controlling the processor 13, and may store data input to or output from the electronic device 10.

The memory 17 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (eg SD or XD memory, etc.), RAM (RAM, Random Access Memory) Static Random Access Memory (SRAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Programmable Read-Only Memory (PROM), Magnetic Memory, Magnetic Disk It may include at least one type of storage medium of the optical disk.

17 is a block diagram of a controller of an apparatus for recommending jewelry goods, according to an exemplary embodiment.

Referring to FIG. 17, the processor 13 may include a data learner 13-1 and a data recognizer 13-2.

The data learner 13-1 may learn a criterion for acquiring and classifying attribute data of a jewelry product. The data learner 13-1 may acquire the text information and the image feature from the user's interest data, and learn the criteria for generating the meta data by linking with each other. The data learner 13-1 may learn a criterion for assigning weights to text information and image features included in the metadata and links thereof. The data learner 13-1 may learn a criterion for obtaining user analysis data from metadata based on the assigned weight. The data learner 13-1 may select attribute data similar to the user analysis data and learn a criterion for selecting a jewelry product corresponding to the selected attribute data as the recommended jewelry product.

The data recognizer 13-2 may acquire and classify attribute data of the jewelry product by using the learned data recognition model. The data recognizer 13-2 may obtain text information and image features from the user's interest data, and generate meta data by linking each other. The data recognizer 13-2 may assign weights to text information, image features, and links thereof included in the meta data. The data recognizer 13-2 may obtain user analysis data from metadata based on the assigned weight. The data recognizing unit 13-2 may select attribute data similar to the user analysis data, and select a jewelry product corresponding to the selected attribute data as the recommended jewelry product. The data recognizing unit 13-2 may select a jewelry product that matches the jewelry product of interest and recommends the jewelry product. The data recognizing unit 13-2 may acquire predetermined data according to a preset reference by learning, and use the data recognition model by using the acquired data as an input value. Also, a result value output by the data recognition model using the obtained data as an input value may be used to update the data recognition model.

At least one of the data learner 13-1 and the data recognizer 13-2 may be manufactured in the form of at least one hardware chip and mounted on the electronic device. For example, at least one of the data learner 13-1 and the data recognizer 13-2 may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or a conventional general purpose processor. (Eg, a CPU or an application processor) or a graphics-only processor (eg, a GPU) and may be mounted on the aforementioned various electronic devices.

In this case, the data learner 13-1 and the data recognizer 13-2 may be mounted on one electronic device or may be mounted on separate electronic devices, respectively. For example, one of the data learner 13-1 and the data recognizer 13-2 may be included in the electronic device, and the other may be included in the server. In addition, the data learning unit 13-1 and the data recognizing unit 13-2 provide the model information constructed by the data learning unit 13-1 to the data recognizing unit 13-2 via wired or wireless. The data inputted to the data recognizing unit 13-2 may be provided to the data learning unit 13-1 as additional learning data.

Meanwhile, at least one of the data learner 13-1 and the data recognizer 13-2 may be implemented as a software module. When at least one of the data learning unit 13-1 and the data recognizing unit 13-2 is implemented as a software module (or a program module including an instruction), the software module may be read by a computer. It may be stored in a readable non-transitory computer readable media. In this case, at least one software module may be provided by an operating system (OS) or by a predetermined application. Alternatively, some of the at least one software module may be provided by an operating system (OS), and others may be provided by a predetermined application.

18 is a block diagram of a server interworking with an apparatus for recommending jewelry items, according to an exemplary embodiment.

Referring to FIG. 18, the server 20 according to some embodiments may include a communication unit 25, a DB 27, and a processor 23.

The communication unit 25 may include one or more components for communicating with the electronic device 10.

The DB 27 may store data and programs for selecting recommended jewelry items that match user input data related to the jewelry item of interest.

The processor 23 typically controls the overall operation of the server 20. For example, the processor 23 may control the DB 27, the communication unit 25, and the like overall by executing programs stored in the DB 27 of the server 20. The processor 23 may execute a part of the operation of the electronic device 10 of FIGS. 1 to 15 by executing programs stored in the DB 27.

Meanwhile, the electronic device 10 and the server 20 may effectively distribute and perform a task for learning the data recognition model and data recognition, thereby processing data in order to provide a service corresponding to the user's intention. It can perform efficiently and protect user's privacy effectively.

Some embodiments may also be embodied in the form of a recording medium containing instructions executable by a computer, such as program modules executed by the computer. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. In addition, computer readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules.

In addition, in this specification, “unit” may be a hardware component such as a processor or a circuit, and / or a software component executed by a hardware component such as a processor.

The above description of the present invention is intended for illustration, and it will be understood by those skilled in the art that the present invention may be easily modified in other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

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

Claims (15)

In the jewelry recommendation device recommends a jewelry product to the user,
Receiving interest data of the user from the user;
Applying the interest data to an artificial intelligence neural network; And
Displaying data related to the recommended jewelery product output from the artificial intelligence neural network;
The artificial intelligence neural network,
Obtaining user analysis data related to the taste of the user from the interest data,
Based on the user analysis data, select the recommended jewelry products that match the taste of the user,
Learned to output data related to the selected recommended jewelry item.
According to claim 1,
The interest data,
The user uses the jewelry product recommendation device,
A web page representing the search results,
The webpages you visited, and
Comprising data contained in a webpage linked to the webpages
According to claim 1,
The artificial intelligence neural network
Obtaining text data and image data from the interest data,
Generating metadata related to the taste of the user based on the text data and the image data,
Obtaining the user analytics data from the metadata
The method of claim 3, wherein
The artificial intelligence neural network
Obtaining text information related to the taste of the user from the text data,
Obtaining an image characteristic related to the taste of the user from the image data,
Learned to generate the metadata by linking the text information and the image feature.
The method of claim 4, wherein
The artificial intelligence neural network
Trained to generate the metadata by weighting the linked text information and the image feature.
The method of claim 4, wherein
The artificial intelligence neural network
Based on the first metadata stored in the database, the second metadata is generated by linking the text information and the image feature obtained from the data of interest,
Update the first metadata using the second metadata.
The method of claim 4, wherein
The image data includes image data including jewelry products and fashion products,
The artificial intelligence neural network,
Obtaining a first image feature related to the jewelry item and a second image feature related to a fashion item from the image data,
Obtaining first text information related to the jewelry item and second text information related to the fashion item from the text data,
Linking the first image feature and the first text information and linking the second image feature and the second text information to generate the metadata;
The method of claim 7, wherein
The artificial intelligence neural network,
Link the first image feature and the second image feature,
Generating the metadata by linking the first text information and the second text information.
According to claim 1,
The artificial intelligence neural network,
Obtaining image data from the interest data,
Obtaining image characteristics of the jewelry product from the image data,
Obtaining text information related to an attribute of the jewelry product based on an image characteristic of the jewelry product;
By linking the image feature and the text information of the jewelry product, generating metadata related to the taste of the user,
Obtaining the user analytics data from the metadata
The method of claim 4, wherein
The artificial intelligence neural network,
Based on the data of interest, linked to each other to update weights of text information and image features included in the metadata,
Obtaining text information and image features having a weight greater than or equal to a threshold value among the text information and image features included in the metadata as the user analysis data,
Selecting attribute data similar to the user analysis data among the attribute data of the jewelry items stored in the database,
And selecting the jewelry item corresponding to the selected attribute data as the recommended jewelry item.
The method of claim 10,
The artificial intelligence neural network,
Selecting similar attribute data to the user analysis data by weighting each of the at least one data included in the attribute data of the jewelry items stored in the database.
The method of claim 11, wherein
The artificial intelligence neural network,
And weight each of the at least one data included in the attribute data of the jewelry items stored in the database based on the text information and the image feature obtained from the interest data.

The method of claim 11, wherein
Displaying data related to the recommended jewelry product,
And similar to the attribute data of the recommended jewelery item, and displaying together the text information included in the user analysis data and the attribute data of the recommended jewelery item.
An application stored on a recording medium for causing a mobile terminal to perform the following functions for recommending a jewelery item to a user, wherein the functions are:
Receiving interest data of the user from the user;
Applying the interest data to an artificial intelligence neural network; And
And a function of displaying data related to recommended jewelry items outputted from the artificial intelligence neural network.
The artificial intelligence neural network
Obtaining user analysis data related to the taste of the user from the interest data,
Based on the user analysis data, select the recommended jewelry products that match the taste of the user,
And learn to output data related to the selected recommended jewelry product.
A user input unit configured to receive interest data of the user from a user;
A processor for applying the interest data to an artificial intelligence neural network; And
A display unit configured to display data related to recommended jewelry items output from the artificial intelligence neural network;
The artificial intelligence neural network,
Obtaining user analysis data related to the taste of the user from the interest data,
Based on the user analysis data, select the recommended jewelry products that match the taste of the user,
A jewelry product recommendation device learned to output data related to the selected recommended jewelry product.

Images