KR20190079829A - Method for Registering Product Information by using Image - Google Patents

Abstract

The present invention relates to a method for automatically registering product information using an image, which confirms an easily searched keyword and automatically registers the same with a product to be registered. The method of the present invention comprises: a first step of registering i product images and storing the same in a designated storage medium; a second step of extracting N (N >= 1) product information keywords; a third step of confirming P (P >= 1) similar product detail pages; a fourth step of extracting Q (Q >= 1) product search keywords; a fifth step of confirming R (R >= 1) search keywords; a sixth step of selecting r (1 <= r <= R) search keywords; a seventh step of setting n (1 <= n <= (r ∩ N)) keywords as main keywords of products to be registered; and an eighth step of performing a procedure of automatically including the set n main keywords in product information on the product to be registered.

Description

Method for registering product information using image [

The present invention is characterized in that, based on an image analysis of a product image registered by a registrant, n (n ⊂ N) is set so that a commodity detail page of a commodity registered by a registrant is included first in a search result of a designated search engine, The present invention relates to an automatic product registration method for automatically selecting a main keyword and automatically including it in product information.

With the recent development of Internet information technology, various types of shopping malls have appeared and they sell a wide variety of products online.

In order to sell products through online shopping malls and the like, a shopping mall operator or a product seller registers information on the goods to be sold. In order to sell the goods more well, a product description of a unique product or a product image that stands out is used .

Korean Patent Laid-Open Publication No. 10-2015-0064768 (published on June 12, 2015) relates to a product information registration system and a method thereof, in which a person who sells a product in a shopping mall or the like sells a product image, A price, a product description, and the like.

Korean Patent Laid-Open Publication No. 10-2014-0087090 (published on Jul. 09, 2014) relates to a method and system for registering product information, in which product information is generated by inputting only the product name, And the product information is updated by using the comment list.

However, in the case of the above-mentioned prior patent, since the person who registers the goods generally decorates or creates and registers the goods information about the goods to be registered, the registration process of the goods information is cumbersome, and more people are universally And can not confirm whether it is the product information which can be accessed.

Accordingly, it is time to search for a new method of registering product information that can be more easily retrieved or exposed to the public on-line.

It is an object of the present invention to overcome the above-mentioned problems, and it is an object of the present invention to solve the above problem, There is provided a method of automatically registering an item information using an image to register a product detail page including a registered product image from an over-priced product detail page.

A method for automatically registering goods information using an image executed via an operation server communicating with a registrant terminal according to the present invention, comprising the steps of: registering i (i? 1) (N) corresponding to at least j (1? J? J) keyword categories among J (J? 1) specified keyword categories based on image analysis of the registered i product images, (M &gt; = 1) product information keywords extracted from the product details pages of M (M &gt; = 2) shopping malls that can be included in the search results through the specified S Page product detail information and at least some of the N product information keywords are compared and analyzed according to the j keyword categories to correspond to the i product images (P &gt; = 1) similar merchandise detail pages that sell the same or similar merchandise as the registered merchandise to be registered, A fourth step of extracting Q (Q? 1) product search keywords that can be used to include similar product detail pages in the search results of S search engines; A fifth step of checking R (R? 1) search keywords matched with Q items retrieval keywords extracted through page-by-page product detail information, Selecting one of r (1? R? R) search keywords within a predetermined rank by sorting in descending order; A step of setting n (1? N? (R? N) keywords matched with the information keywords as the main keywords of the registration target product corresponding to the registered i product images, And the eighth step of automatically including the set n main keywords in the product information of the registration target product corresponding to the registered keyword.

According to the present invention, the first step may further include the step of registering u (u? 1) registrant setting keywords from the registrant terminal. If the registrant setting keyword of u is registered, (K) of the N product information keywords extracted from the i product images and a k (1? K? (N?)) Product information keywords matching with the registered u regenerator set keywords The method comprising the steps of:

According to the present invention, in the second step, k items information keywords matching an N product information keywords extracted from the i product image images and matching the registered u regenerator setting keywords are added to k items information keywords And weighting the ratios.

According to the present invention, in the third step, the product detailed information for each product detail page and the k product information keywords collected from the product detail pages of the M shopping malls are compared and analyzed according to the j keyword categories, (P &gt; = 1) similar merchandise detail pages that sell the same or similar merchandise as the registration target merchandise corresponding to the merchandise image of the number of merchandise.

According to the present invention, the third step is a step of comparing and analyzing the product detailed information of each product detail page collected from the product detail pages of the M shopping malls and the u registrant setting keywords according to the j keyword categories, (P &gt; = 1) similar merchandise detail pages that sell the same or similar merchandise as the registration target merchandise corresponding to the merchandise image of the merchandise.

According to the present invention, the N product information keywords may include at least one of a product type keyword, a product type keyword, a product color keyword, a model sex keyword, an image atmosphere keyword, an image background keyword And a product information keyword corresponding to j keyword categories that are designated in advance.

According to the present invention, the product detail information may include at least one of a product image, a product name, a product detail description, a product category, a brand, a manufacturer, and a country of origin.

According to the present invention, the method of automatically registering goods information using the image is characterized in that at least m (1? M? M) of M (M? 2) shopping malls that can be included in search results through designated S ) Of at least m '(1 &lt; = m &lt; = M) malls linked to screen scraping for each product detail page operated by the shopping malls, Based on the detailed information and the character recognition based on the character recognition of the image information included in the product detail page of each shopping mall, Collecting big data on the information, storing the big data in the designated database, and periodically updating the big data.

According to the present invention, the third step may include comparing and analyzing the product detail information for each product detail page stored in the designated database and the N product information keywords for each j keyword categories, And confirming P (P? 1) similar product detail pages that sell the same or similar products as the product.

According to the present invention, the S search engines may include a search engine of a search portal, a search engine of a price comparison system, and a search engine of a goods query system, which can include product detail pages of the M shopping malls in a search result using a designated search keyword Or &lt; / RTI &gt; search engines.

According to the present invention, the M shopping malls include a shopping mall operating a product detail page that can be included in search results of a designated search keyword through S designated search engines among domestic and overseas open markets, personal shopping malls, general malls, and brand malls .

According to the present invention, the seventh step may further include: n keywords corresponding to an intersection matching with the selected r search keywords and matching N product information keywords extracted from the i product images, And setting it as a main keyword.

According to the present invention, in the seventh step, k (1? K? (N?)) Number of items matching with u (u? 1) registrant set keywords registered through the registrants terminal among the N product information keywords Wherein when the product information keyword is weighted by a certain ratio, n (1? N? (R? K)) keywords matching the k product information keywords among the selected r And setting the keyword as the main keyword of the registration target product corresponding to the image.

According to the present invention, in the seventh step, k (1? K? (N?)) Number of items matching with u (u? 1) registrant set keywords registered through the registrants terminal among the N product information keywords When a weight of a certain percentage is given to the product information keyword, n keywords corresponding to intersection matching with the selected r search keywords and matching the weighted k product information keywords are stored in the main And setting it as a keyword.

According to the present invention, the eighth step may include a step of automatically including the n major keywords in at least one of a product name, a product description, and a meta tag of the registration target product.

According to the present invention, in the eighth step, when u (u? 1) registrant setting keywords are registered from the registrant terminal, the n main keywords are added to the u registrant setting keywords or the u registrant setting keywords And replacing at least a part of the keywords with the n main keywords.

According to the present invention, the eighth step may include the steps of: when character-based product information of the registration target product is registered from the registrant terminal, adding n major keywords to the registered product information, And replacing at least some of the words with the n major keywords.

According to the present invention, product information is extracted through image analysis of a product to be registered, and similar product information having a high frequency of exposure among products similar to the extracted product information is confirmed, and keywords (for example, Or a combination of keywords) and automatically registers the same with the product to be registered.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a functional block diagram of a system for automatically registering goods information using an image according to the present invention; FIG.
FIG. 2 is a view illustrating a process of confirming a similar product detail page corresponding to a registration target product to be registered based on an image according to an embodiment of the present invention.
3A is a diagram illustrating a process of confirming similar product detail pages corresponding to a registration target product to be registered based on an image according to another embodiment of the present invention.
FIG. 3B is a view illustrating a process of confirming a similar product detail page corresponding to a registration target product to be registered based on an image according to another embodiment of the present invention.
4 is a diagram illustrating a process of setting a main keyword for a registration target product according to an embodiment of the present invention and automatically including the keyword in the product information of the registration target product.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood, however, that the drawings and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the invention, and are not to be construed as limiting the present invention.

In other words, the following embodiments correspond to the preferred embodiment of the preferred embodiment of the present invention. In the following embodiments, a specific configuration (or step) is omitted, or a specific configuration (or step) (Or steps), or an embodiment that incorporates functions implemented in more than one configuration (or step) into any one configuration (or step), a particular configuration (or step) It will be apparent that the present invention is not limited to the embodiments described below. Therefore, it should be clearly stated that various embodiments corresponding to subsets or combinations based on the following embodiments can be subdivided based on the filing date of the present invention.

In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. The terms used below are defined in consideration of the functions of the present invention, and may be changed according to the intention of the user, the intention or the custom of the operator. Therefore, the definition should be based on the contents throughout the present invention.

As a result, the technical idea of the present invention is determined by the claims, and the following embodiments are merely means for effectively explaining the technical idea of the present invention to a person having ordinary skill in the art to which the present invention belongs Only.

1 is a functional block diagram of an automatic registration system for goods information using an image according to the present invention.

More specifically, FIG. 1 shows an image of a registered commodity and a registered commodity image from the other commodity details page for selling similar commodities, even though the commodity information and the commodity information keyword are not registered at the time of registering the commodity image 1 is a block diagram illustrating a functional configuration of a system for automatically registering product information such that a product detail page is searched for at a first step. Referring to FIG. 1 and / or modified by the person skilled in the art It is possible to deduce various implementations of the simple configuration of the product information automatic registration system using the image. However, the present invention includes all of the above-described embodiments, Features are not limited.

1, the system for automatically registering goods information using the image includes a registrant terminal 155 operated by a registrant registering product information on the basis of an image, An operation server 100 for analyzing a product image of the registered product and setting an optimal keyword for the registered product by analyzing the product image of the registered product, And one or more search engines 160 for providing the searched search results to the operation server 100.

The registrant terminal 155 includes a wired terminal or a wireless terminal operated by the registrant. The registrant connects to the operation server 100 through the registrant terminal 155 and uses the image of the registration target product You can request automatic registration of goods information.

The search engine may include at least one search engine among a search engine of a search portal, a search engine of a price comparison system, and a search engine of a goods query system, which can include product detail pages of the M shopping malls in a search result using a designated search keyword .

Although the operation server 100 and the search engine 160 are shown separately in FIG. 1, the operation server 100 may include one or more search engines 160 in the course of a person skilled in the art .

Referring to FIG. 1, the operation server 100 includes a product image registration unit 105, a product information keyword extraction unit 110, a big data collection unit 115, a keyword analysis unit 120, A keyword extracting unit 125, a search keyword checking unit 130, a search keyword selecting unit 135, a main keyword setting unit 140, and a keyword processing unit 145.

The product image registration unit 105 can register i (i, 1) number of product images for the registration target product that the registrant wants to register from the registrant terminal 155 and store the image in the designated storage medium 150. [

According to the method of the present invention, the product image registration unit 105 further registers u (u? 1) registrant setting keywords for the registration target product in the process of registering the product image from the registrant terminal 155 Can receive.

The registrant setting keyword may be a keyword consisting of numbers, letters, special symbols, or a combination thereof, set by the registrant so that the registration target product is searched for preferentially in comparison with other similar products, and may include a hash tag have.

The product information keyword extracting unit 110 extracts J (J &gt; = 1) designated product images based on the image analysis of the registered i product images when the i product images are registered through the product image registering unit 105 N (N? 1) product information keywords corresponding to at least j (1? J? J) keyword categories out of the keyword categories can be extracted.

Here, the N product information keywords may include a product type keyword, a product type keyword, a product color keyword, a model sex keyword, an image mood keyword, and an image background keyword, which are extracted based on image analysis of the i product images, And may include a product information keyword corresponding to the keyword categories.

According to the embodiment of the present invention, when the u (u? 1) registrant setting keywords are registered from the registrant terminal 155 through the product image registering unit 105, the product information keyword extracting unit 110 extracts, it is possible to assign a certain weight to the k (1? k? (N?)) product information keywords matching the registered u regenerator set keywords among the N product information keywords extracted from the i product images .

In addition, according to the embodiment of the present invention, the product information keyword extracting unit 110 matches the N product information keywords extracted from the i product images, and at the same time, matches the registered u regenerator set keywords It is possible to assign a weighting value of a certain ratio to the corresponding k product information keywords.

The keyword analyzing unit 120 extracts N product information keywords through the product information keyword extracting unit 110 and extracts M items that can be included in the search results through the designated S (S? 1) (M &gt; = 2) shopping malls and the product information keywords of at least some of the N product information keywords according to the j keyword categories, P (P? 1) similar product detail pages that sell the same or similar products as the registration target product corresponding to the product image can be confirmed.

Here, the product detail information may include at least one of a product image, a product name, a product detail description, a product category, a brand, a manufacturer, and a country of origin.

In addition, the M shopping malls may include a shopping mall that operates product detail pages that can be included in search results of designated search keywords through S search engines 160 specified among domestic and overseas open markets, personal shopping malls, general malls, and brand malls .

According to the embodiment of the present invention, the keyword analyzing unit 120 extracts, from the N product information keywords extracted from the i product images through the product information keyword extracting unit 110, the registered u regenerator set keyword When the matching k product information keywords are extracted (or confirmed), the detailed product information for each product detail page and the k product information keywords collected from the product detail pages of the M shopping malls are sorted according to the j keyword categories The P similar product detail pages for selling the same or similar products as the registration target products corresponding to the i product images can be confirmed and compared.

Also, according to the embodiment of the present invention, the keyword analyzing unit 120 analyzes the detailed product information of each product detail page collected from the product detail pages of the M shopping malls and the u registrant setting keywords into the j keyword categories , The P similar product detail pages for selling the same or similar products as the registration target products corresponding to the i product images can be confirmed.

The big data collecting unit 115 collects the m (M? 2) shopping malls that are included in the search results through the designated S (S? 1) Based on screen scraping of the product details page and interworking with at least m '(1 ≤ m' ≤ M) shopping malls affiliated with each other, detailed information of the string based product included in each product detail page operated by each shopping mall, Collecting big data about the detailed product information of each product detail page of each shopping mall through data mining of one or more product detail information of the character recognition based product detailed information recognized through character recognition on the image information included in the detailed page And stored in a specified database (or storage medium) and periodically updated.

According to the embodiment of the present invention, when the big data for the detailed product information for each product detail page of each shopping mall is collected and stored through the big data collecting unit 115, P (P &gt; = P &lt; / = P &gt;) which sells commodities identical or similar to the registration target goods corresponding to the i product images, by comparing and analyzing the commodity detail page- 1) You can check the similar product detail page.

The product search keyword extracting unit 125 extracts P (P? 1) similar product detail pages for selling the same or similar products as the registration target products corresponding to the i product images through the keyword analyzing unit 120 (Q &gt; = 1) product search keywords usable for incorporating the P similar product detail pages into the search results of the S search engines, Can be extracted.

When the Q items search keyword is extracted through the goods search keyword extracting unit 125, the search keyword check unit 130 searches the P similar item detail pages of the S search keywords for each of the S search engines 160 R (R &gt; = 1) search keywords matched with Q product search keywords extracted through the product detailed information can be confirmed.

When the R search keywords matched with the Q product search keywords are confirmed through the search keyword checking unit 130, the search keyword selector 135 searches for the latest search frequency and popularity rank of the R search keywords R (1? R? R) search keywords belonging to within a predetermined rank can be selected.

When the r retrieval keywords are selected through the retrieval keyword selector 135, the main keyword establishing unit 140 sets n (1? N (n)) among the r retrieval keywords matched with the N product information keywords ≤ (r∩N) keywords as the main keywords of the registration target product corresponding to the registered i product images.

According to the embodiment of the present invention, the main keyword setting unit 140 sets n keywords corresponding to the intersection matching with the selected r search keywords and matching N product information keywords extracted from the i product images, The keyword can be set as the main keyword of the registration target product.

According to an embodiment of the present invention, the main keyword setting unit 140 sets k (k) that matches u (u? 1) registrant set keywords registered through the registrant terminal 155 among the N product information keywords 1 &amp;le; n &amp;le; r &apos; (n &apos;)) matching the k product information keywords among the r selected search keywords, k) keywords as the main keywords of the registration target product corresponding to the registered i product images.

According to an embodiment of the present invention, the main keyword setting unit 140 sets k (k) that matches u (u? 1) registrant set keywords registered through the registrant terminal 155 among the N product information keywords 1 &amp;le; k &amp;le; N &amp;bull; N &amp; circ &amp; N) product information keywords are matched with the selected r search keywords and matched with the weighted k product information keywords The corresponding n keywords can be set as the main keywords of the registration target product.

If the n keywords are set as the main keywords of the registration target product corresponding to the registered i product images through the main keyword setting unit 140, The process of automatically including the set n main keywords in the product information of the corresponding registration target product can be performed.

According to the embodiment of the present invention, the keyword processing unit 145 may perform a process of automatically including the n major keywords in at least one of a product name, a product description, and a meta tag of the registration target product.

According to the embodiment of the present invention, when u (u? 1) registrant setting keywords are registered from the registrant terminal 155, the keyword processing unit 145 searches the u registrant setting keywords for the n main A keyword may be added or at least some of the u registrant setting keywords may be replaced with the n main keywords.

According to the embodiment of the present invention, when the character-based product information on the registration target product is registered from the registrant terminal 155, the keyword processing unit 145 adds n main keywords Or at least a part of the words included in the product information may be replaced with the n main keywords.

FIG. 2 is a view illustrating a process of confirming a similar product detail page corresponding to a registration target product to be registered based on an image according to an embodiment of the present invention.

In more detail, FIG. 2 illustrates an operation of the operation server 100 shown in FIG. 1, in which an image of a registration target product requested by a registrant terminal 155 is analyzed to extract a product information keyword, And comparing and analyzing the product information for each product detail page collected from the product detail page of one or more shopping malls that can be included in the search result through the one or more search engines 160, If the person who has ordinary knowledge in the technical field to which the present invention belongs is referred to and / or modified by referring to FIG. 2 to confirm the similar item detail page It will be appreciated that various implementations of the process may be deduced, And the technical features thereof are not limited only by the method shown in FIG.

Referring to FIG. 2, in the process of confirming the similar product detail page, the operation server 100 receives i (i? 1) product images from the registrant terminal 155 and registers (200). &Lt; / RTI &gt;

When the product image is registered, the operation server 100 stores the product image in a designated storage medium 150 (205).

Then, the operation server 100 performs image analysis on the registered i product images, and determines at least j (1? J) among J (J? 1) (N &gt; = 1) pieces of goods information keywords corresponding to the keyword categories of (N &lt; = J)

When the N product information keywords are extracted, the operation server 100 extracts the product information from the product detail pages of M (M? 2) shopping malls that can be included in the search result through the designated S (S? 1) (220) comparing and analyzing at least a part of the product information keywords of the product details page and the N product information keywords according to the j keyword categories, P (P? 1) similar product detail pages selling similar goods are confirmed (225).

FIG. 3A is a diagram illustrating a process of confirming a similar product detail page corresponding to a registration target product to be registered based on an image according to another embodiment of the present invention.

In more detail, FIG. 3A illustrates an example of a case where an operation server 100 shown in FIG. 1 registers an image and a registrant setting keyword for a registration target product from the registrant terminal 155, analyzes the image of the registration target product, A product information keyword matching the registrant set keyword and product information for each product detail page collected from a product detail page of one or more shopping malls that can be included in a search result through one or more search engines 160 The method comprising the steps of confirming at least one similar product detail page that sells a product identical or similar to a registration target product corresponding to the product image. As a person skilled in the art, 3a is referred to and / or modified to confirm the similar product detail page Would be able to infer a variety of exemplary methods for the information, the present invention is made, including any exemplary way in which the inference, to which the technical feature that is not limited to the exemplary method shown in the figure 3a.

Referring to FIG. 3A, the process of confirming the similar product detail page shown in FIG. 3 is performed by the operation server 100, in which i (i, 1) product images and u (u &Gt; 1) of registrant setting keywords (300a).

When the product image and the registrant setting keyword are registered, the operating server 100 stores the keyword in the designated storage medium 150 (305a).

Then, the operation server 100 performs image analysis (310a) on the registered i product images, and calculates at least j (1? J (N &gt; = 1) pieces of goods information keywords corresponding to the number of keyword categories of &lt; = J) are extracted (315a).

When the N product information keywords are extracted, the operation server 100 determines k (1? K? (N?)) Items matching the registered u regenerator set keywords among the extracted N product information keywords The information keyword is confirmed (320a).

When the k product information keywords are confirmed, the operation server 100 gives a predetermined weight to the identified k product information keywords (325a).

The operation server 100 stores detailed product information of each product detail page collected from product detail pages of M (M? 2) shopping malls that can be included in the search result through the designated S (S? 1) Comparing the k product information keywords according to the j keyword categories (330a), and comparing P (P &gt; = 1) similar product details for selling the same or similar products as the registration target products corresponding to the i product images The page is confirmed (335a).

FIG. 3B is a view illustrating a process of confirming a similar product detail page corresponding to a registration target product to be registered based on an image according to still another embodiment of the present invention.

In more detail, FIG. 3B illustrates an example of the operation of the operation server 100 shown in FIG. 1, in which an image and a registrant setting keyword for a registration target product are registered from the registrant terminal 155 and an image of the registration target product is analyzed The goods information keyword matching the product information keyword and the registrant setting keyword is matched and the product information keyword is collected from the product detail page of one or more shopping malls that can be included in the search result through the one or more search engines 160 And comparing the analyzed product information of each product detail page to identify one or more similar product detail pages that sell the same or similar products as the registration target product corresponding to the product image. Those skilled in the art will be able to refer to Figure 3b and / or It is possible to deduce various methods for checking the similar item detail page. However, the present invention includes all of the above-described methods of implementation, Features are not limited.

Referring to FIG. 3B, the process of confirming the similar item detail page shown in FIG. 3 is performed by the operation server 100, in which i (i? 1) product images and u (u &Gt; 1) of registrant setting keywords (300b).

When the product image and the registrant setup keyword are registered, the operation server 100 stores the keyword in the designated storage medium 150 (305b).

Then, the operation server 100 performs image analysis on the registered i product images (310b). Based on the image analysis, at least j (1? J (N &gt; = 1) pieces of goods information keywords corresponding to the number of keyword categories of &lt; = J) are extracted (315b).

When the N product information keywords are extracted, the operation server 100 identifies k product information keywords corresponding to the intersection matching with the extracted N product information keywords and matching the u regenerator set keywords ( 320b).

When k product information keywords corresponding to the intersection are identified, the operation server 100 gives a weight of a certain percentage of the k product information keywords identified (325b).

The operation server 100 stores detailed product information of each product detail page collected from product detail pages of M (M? 2) shopping malls that can be included in the search result through the designated S (S? 1) Comparing the u registrant set keywords according to the j keyword categories (330b), and comparing P (P &gt; = 1) similar product details for selling the same or similar items as the registration target goods corresponding to the i product images The page is confirmed (335b).

FIG. 4 is a diagram illustrating a process of setting a main keyword for a registration target product according to an embodiment of the present invention and automatically including the keyword in the product information of the registration target product.

More specifically, FIG. 4 is a flowchart illustrating an operation of the operation server 100 after the similar product detail page for selling the same or similar product as the registration target product is confirmed through the process of FIG. 2, FIG. 3a or FIG. 3b, A product search keyword having a possibility to be used for inclusion in the search result of the search engine by using the item detail information of the detailed page, a search keyword of each search engine, a popular search keyword among the search keywords matched with the product search keyword, A key keyword matched with a popular search keyword frequently used in actual search in each search engine and matched with a product information keyword extracted from the product image is automatically included in the product information of the registration target product. Those of ordinary skill in the art will appreciate that, It is possible to deduce various methods for the process of automatically including the main keyword for the registration target product in the product information of the registration target product by referring to and / or modifying the FIG. 4. However, All the methods of implementation are included, and the technical features thereof are not limited only by the method shown in FIG.

Referring to FIG. 4, the process of setting the main keyword for the registration target product shown in FIG. 4 and automatically including the registered keyword in the product information of the registration target product is performed by the operation server 100 in the process of FIG. 2, FIG. Extracting Q (Q &gt; = 1) product search keywords that can be used to read P similar product detail page-by-page product details information identified through the search results of the P similar product detail pages into S search engine search results (400).

When the Q product search keywords are extracted, the operation server 100 searches the Q detailed product search keywords extracted from the S detailed search results of the S similar search results, (R &gt; = 1) search keywords (405).

The operating server 100 selects r (1? R? R) search keywords within a predetermined rank (410) by sorting in descending order based on the latest search frequency or popularity ranking of the R search keywords, .

When the r retrieval keywords are selected, the operating server 100 transmits n (1? N? (R? N) keywords matching the N product information keywords among the selected r retrieval keywords to the registered (415) as the main keyword of the registration target product corresponding to the i product images.

Here, the operation server 100 may compare n keywords corresponding to intersections matched with the selected r search keywords and the N product information keywords extracted from the i product images to the main keywords of the registration target product Can be set.

3A, k (1? K? (N?)) Items matching with u (u? 1) registrant set keywords registered through the registrant terminal 155 among the N product information keywords as shown in FIG. When the information keyword is weighted at a certain rate, the operation server 100 stores n (1? N? (R? K) keywords matching the k product information keywords among the selected r search keywords It can be set as the main keyword of the registration target product corresponding to the registered i product images.

As shown in FIG. 3B, k (1? K? (N?)) Items matching with u (u? 1) regenerator set keywords registered through the registrant terminal 155 among the N product information keywords If the information keywords are weighted at a certain rate, the operation server 100 matches the selected r search keywords with n keywords corresponding to the intersection matching with the weighted k product information keywords Can be set as the main keyword of the registration target product.

Meanwhile, when the main keyword of the registration target product corresponding to the i product images is set, the operation server 100 stores the set of the n main keywords set in the product information of the registration target product corresponding to the registered i product images (420). &Lt; / RTI &gt;

Here, the operating server 100 may automatically include the n major keywords in at least one of a product name, a product description, and a meta tag of the registration target product.

When the u (u? 1) registrant setting keywords are registered from the registrant terminal 155 as shown in FIG. 3A or FIG. 3B, It is possible to add a keyword or replace at least a part of the u registrant setting keywords with the n main keywords.

In addition, when the character-based product information on the registration target product is registered from the registrant terminal 155, the operation server 100 adds n major keywords to the registered product information or includes n main keywords in the product information It is possible to replace at least a part of the words with the n main keywords.

100: Operation server 105: Product image registration part
110: goods information keyword extracting unit 115: big data collecting unit
120: Keyword analysis unit 125: Product search keyword extracting unit
130: Search keyword verification unit 135: Search keyword selection unit
140: Keyword setting unit 145: Keyword processing unit
150: storage medium 155: registrant terminal
160: Search engine

Claims (16)

A method executed via an operating server in communication with a registrar terminal,
A first step of registering i (i? 1) product images for a registration target product from a registrant terminal and storing the registered image in a designated storage medium;
(N &gt; = 1) product information keywords corresponding to at least j (1? J? J) keyword categories out of J (J? 1) specified keyword categories based on image analysis of the registered i product images A second step of extracting a second image;
(M &gt; = 2) shopping malls that can be included in search results through designated S (S &gt; = 1) search engines and at least some of the N product information keywords (P &gt; = 1) similar merchandise detail pages for selling the same or similar merchandise as the registration target merchandise corresponding to the i merchandise images by comparing and analyzing the merchandise information keyword of the step;
(Q &gt; = 1) product search keywords that can be used to read the P similar product detail page-by-page product detailed information and to include the P similar product detail pages in the search results of S search engines Step 4;
A fifth step of confirming R (R? 1) search keywords matched with Q product search keywords extracted through the S detailed search results of each of the P similar product detail pages;
A sixth step of sorting r (1? R? R) search keywords belonging to within a predetermined rank by sorting in descending order based on the latest search frequency or popularity ranking of the R search keywords;
And sets n (1? N? (R? N) keywords matched with the N product information keywords among the selected r search keywords as the main keywords of the registration target goods corresponding to the registered i product images Step 7 and
And a step of automatically including the set n key words in the product information of the registration target product corresponding to the registered i product images.
The method according to claim 1,
The first step may further include registering u (u? 1) registrant setting keywords from the registrant terminal,
When the registrant setting keyword of u is registered,
The second step comprises:
A step of assigning weights of k (1? K? (N∩u)) commodity information keywords matched with the registered u registrant set keywords among the N commodity information keywords extracted from the i commodity images, The method of claim 1, further comprising:
3. The method according to claim 2,
And a step of assigning weights of a certain ratio to k items of product information keywords matched with N product information keywords extracted from the i product images and corresponding to an intersection matched with the registered u regenerator setting keywords Wherein the method comprises the steps of:
3. The method according to claim 2,
Comparing the product detailed information of each product detail page collected from the product detail pages of the M shopping malls and the k product information keywords according to the j keyword categories and comparing the product detailed information of each product detail page with the registration target product corresponding to the i product images (P &gt; = 1) similar merchandise detail pages for selling a similar merchandise or selling a similar merchandise.
3. The method according to claim 2,
Comparing the product detailed information of each product detail page collected from the product detail pages of the M shopping malls and the u registrant setting keywords in accordance with the j keyword categories and comparing them with the registration target products corresponding to the i product images (P &gt; = 1) similar merchandise detail pages for selling a similar merchandise or selling a similar merchandise.
The method of claim 1, wherein the N product information keywords include:
A product information keyword corresponding to j keyword categories preliminarily specified among a product category keyword, a product form keyword, a product color keyword, a model sex keyword, an image mood keyword, and an image background keyword extracted based on image analysis of the i product images And automatically registering the product information using the image.
The information processing apparatus according to claim 1,
Wherein the product information includes at least one of a product image, a product name, a product description, a product category, a brand, a manufacturer, and a country of origin.
The method according to claim 1,
Screen scraping or affiliation for each product detail page operated by at least m (1? M? M) malls among M (M? 2) shopping malls that can be included in search results through designated S (S? 1) Based on at least m '(1? M'? M) shopping malls, at least m '(1? M'? M) shopping malls, and the image information included in the product detail page of each shopping mall Big data on the detailed product information of each product detail page of each shopping mall is collected through data mining on one or more item detail information of the character recognition based on the character recognition based on the character recognition and stored in the designated database and periodically updated Further comprising the steps of:
The third step is a step of comparing and analyzing the product detail information for each product detail page stored in the designated database and the N product information keywords for each of the j keyword categories, And confirming P (P? 1) similar product detail pages for selling the product.
9. The search engine according to claim 1 or 8,
A search engine of a search portal, a search engine of a price comparison system, and a search engine of a product search system, which can include a product detail page of the M shopping malls in a search result using a designated search keyword, How to automatically register product information using images.
9. The shopping mall according to claim 1 or 8,
And a shopping mall that operates product detail pages that can be included in search results of a designated search keyword through S designated search engines among domestic and overseas open markets, personal shopping malls, general malls, and brand malls. Way.
The method as claimed in claim 1,
And setting n keywords corresponding to intersection matching with the selected r search keywords and matching N product information keywords extracted from the i product images as the main keywords of the registration target product Automatic registration of product information using feature images.
3. The method as claimed in claim 1 or 2,
(K) of the product information keywords of k (1? K? (N?)) That are matched with u (u? 1) registrant setting keywords registered through the registrant terminal among the N product information keywords Occation,
And sets n (1? N? (R? K) keywords matched with the k product information keywords among the selected r search keywords as the main keywords of the registration target goods corresponding to the registered i product images The method of claim 1, further comprising:
3. The method as claimed in claim 1 or 2,
(K) of the product information keywords of k (1? K? (N?)) That are matched with u (u? 1) registrant setting keywords registered through the registrant terminal among the N product information keywords Occation,
And setting n keywords corresponding to an intersection matched with the selected r search keywords and matching the weighted k article information keywords as the main keywords of the registration target product Automatic registration of product information using images.
The method as claimed in claim 1,
Wherein the step of automatically including the n major keywords in at least one of a product name, a product description, and a meta tag of the registration target product is performed.
The method as claimed in claim 1,
When u (u? 1) registrant setting keywords are registered from the registrant terminal,
And adding the n main keywords to the u registrant setting keywords or replacing at least a part of the u registrant setting keywords with the n main keywords. .
The method as claimed in claim 1,
When the character-based product information of the registration target product is registered from the registrant terminal,
And adding n main keywords to the registered product information or replacing at least a part of words included in the product information with the n main keywords.

Images