KR20220162343A - Method for recommending merchandise keywords according to seller type - Google Patents

Abstract

The present invention relates to a method for recommending the keywords of merchandise to be sold in an open market by a seller by clustering the merchandise keywords of an open market, determining the type of a seller using a survey and matching the merchandise keywords depending on the type of the seller. The present invention provides a method for recommending merchandise keywords depending on a seller type, wherein a merchandise keyword recommending device yields the attributes of merchandise keywords for each of a plurality of attribute classification criteria, determines the type of the merchandise keywords by combining the yielded attributes of the merchandise keywords, yields the attributes of a seller for each of a plurality of attribute criteria depending on the result of a survey each time the survey is performed, determines the type of the seller by combining the yielded attributes of the seller, and refers to merchandise keywords having the same type as the type of the seller to recommend the same to the seller.

Description

How to recommend product keywords according to seller type { METHOD FOR RECOMMENDING MERCHANDISE KEYWORDS ACCORDING TO SELLER TYPE }

The present invention relates to a method for clustering product keywords in an open market, determining the type of seller through a survey, and then matching product keywords according to the type of seller to recommend keywords for products to be sold by the seller in the open market. will be.

An open market is a platform in which a number of sellers and consumers conduct electronic commerce on-line. An open market operator provides a platform, a seller registers product information to be sold in the open market, and a consumer can purchase product information after searching for product information in the open market.

Since conventional open market platforms do not recommend products to be sold to sellers according to their inclinations, a method of recommending keywords of products to be sold according to their types to sellers who have difficulty selecting products to sell is required.

An object of the present invention is to provide a method of clustering product keywords in an open market, determining the type of seller, and then recommending keywords of products to be sold to the seller.

In order to achieve the above object, the present invention, a product keyword recommendation device calculates attributes of product keywords for each of a plurality of attribute classification criteria, determines the type of product keywords by combining the attributes of the calculated product keywords, and Each time a survey is conducted, the seller's attributes are calculated for each of a plurality of attribute criteria according to the survey results, the seller's type is determined by combining the calculated seller's attributes, and product keywords having the same type as the seller's type are selected. It provides a product keyword recommendation method according to the seller type, which is searched and recommended to the seller.

In addition, before calculating product keyword properties, the product keyword recommendation device receives top product keyword list data including N product keywords with high sales volume or search volume rankings for each product category from the open market server computer, and top product keywords. A method of recommending product keywords according to seller types is provided, in which product keywords including brand names or product keywords including product model names are deleted from list data.

In addition, the product keyword recommendation device deletes the remaining product keywords from the top product keyword list data except for one product keyword with the highest sales volume or search volume ranking among product keywords included in the product keyword synonym list, or A method of recommending a product keyword according to a seller type is provided, in which, among two product keywords having similar content information and sales content information, a product keyword having a low ranking in sales volume or search volume is deleted from upper product keyword list data.

In addition, the product keyword recommendation device includes product keywords that search for products whose monthly sales are equal to or greater than the standard amount, or product keywords that search for products whose sales rank is within the standard rank in the product category, or product keywords whose search volume has increased during the standard period. to provide a product keyword recommendation method according to seller type, generating a recommended product keyword pool.

In addition, the product keyword recommendation device targets the product keywords included in the recommended product keyword pool, the market search range attribute classification criteria for distinguishing concentrated and distributed attributes, and the strategy establishment method attribute for distinguishing warlike attributes from peaceful attributes. A method for recommending product keywords according to seller types, which calculates product keyword attributes according to classification criteria, selection criteria for distinguishing risky attributes from stable attributes, and attribute classification criteria for distinguishing effect-type attributes from efficiency-type attributes provides

In addition, the product keyword recommender uses the survey results to determine the market search range attribute classification criteria for distinguishing centralized and distributed attributes, strategy establishment method attribute classification criteria for distinguishing warlike attributes from peaceful attributes, and risky attributes based on the survey results. Provides a method for recommending product keywords according to seller types, which calculates seller attributes according to selection criteria for distinguishing between stable and stable attributes and attribute criteria for distinguishing effect-type and efficiency-type attributes.

In the present invention, clustering can be performed by determining attributes of product keywords and determining types. And after determining the type of seller through a survey, product keywords that match the type of seller can be recommended. Accordingly, suitable sales product information may be provided according to the seller's type and marketing strategy, and the seller's convenience may be improved.

1 is a block diagram schematically illustrating a product keyword recommendation device and a device connected thereto for executing an embodiment of the present invention.
2A is a flowchart illustrating a step-by-step process of clustering product keywords according to an embodiment of the present invention.
2B is a table showing attributes of product keywords according to an embodiment of the present invention.
2c is a table showing product keywords and seller types according to an embodiment of the present invention.
3A is a flowchart illustrating a step-by-step process of matching product keywords according to seller types according to an embodiment of the present invention.
3B is a table showing survey questions according to an embodiment of the present invention.

The present invention can be variously modified and practiced without departing from the gist, and may have one or more embodiments. In addition, the embodiments described in the "specific contents for carrying out the invention" and "drawings" in the present invention are examples for specifically explaining the present invention, and do not limit or limit the scope of the present invention.

Therefore, what can be easily inferred from the “specific details for carrying out the invention” and “drawings” of the present invention by those skilled in the art to which the present invention belongs is construed as belonging to the scope of the present invention. can do.

In addition, the size and shape of each component shown in the drawings may be exaggerated for description of the embodiment, and does not limit the size and shape of the actual invention.

Terms used in the specification of the present invention may have the same meaning as commonly understood by a person of ordinary skill in the art to which the present invention belongs unless specifically defined.

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

1 is a block diagram schematically illustrating a product keyword recommendation device and a device connected thereto for executing an embodiment of the present invention.

The product keyword recommendation device 100 may be connected to the open market server computer 200 . Although the product keyword recommendation device 100 and the open market server computer 200 are shown as separate components, the product keyword recommendation device 100 may be a single component included in the open market server computer 200 . The open market server computer 200 may be a device that stores product information, sales information, seller information, and consumer information, and provides electronic commerce services to terminal devices used by sellers and consumers.

The product keyword recommendation device 100 may be connected to the seller terminal device 300 . The seller terminal device 300 may be any one of a personal computer, a laptop computer, a smart phone, and a tablet computer. While communicating with the open market server computer 200 using the seller terminal device 300, the seller may register product information to be sold, and check sold product information.

The product keyword recommendation device 100 may include a central processing unit 110 , a storage device 120 , an input/output device 130 , and a bus 140 .

The product keyword recommendation method according to the seller type according to an embodiment of the present invention may be implemented as a computer program that is a plurality of command sets readable by the product keyword recommendation device 100 and stored in a recording medium.

The central processing unit 110 may execute commands of a computer program in sequence or according to branch and repeat conditions.

The memory device 120 may include an auxiliary memory device 121 and a main memory device 122 . The auxiliary storage device 121 may semi-permanently store a computer program. The secondary storage device 121 includes a hard disk drive, an optical disk drive, a magnetic tape, a floppy disk, a memory card, and a solid SSD (SSD). State Drive) may include one or more. The main storage device 122 may load a computer program from the auxiliary storage device 121 and read it in the central processing unit 110 . The main memory device 122 may be RAM (Random Access Memory).

The input/output device 130 may transmit binary data to or receive binary data from the open market server computer 200 or the seller terminal device 300 . The input/output device 130 may include an external input/output terminal (I/O port). External I/O ports include serial port, parallel port, SCSI (Small Computer System Interface), USB (Universal Serial Bus), IEEE 1394, e-SATA (external Serial Advanced Technology Attachment) and Thunderbolt. The input/output device 130 may include a network interface controller (NIC). A network interface controller (NIC) is connected to a Local Area Network (LAN) based on Ethernet in a wired manner or to a Wireless Local Area Network (WLAN) based on Wi-Fi in a wireless manner. can be connected

The bus 140 may connect the central processing unit 110 , the memory device 120 , and the input/output device 130 . Components coupled to bus 140 may transmit or receive binary data over bus 140 .

A product keyword recommendation method according to seller types according to an embodiment of the present invention may include a process of clustering product keywords and a process of matching product keywords according to seller types.

The product keyword recommendation device 100 may calculate product keyword attributes for each of a plurality of attribute classification criteria, and may determine a type of product keyword by combining the calculated product keyword attributes. The product keyword recommendation apparatus 100 may calculate seller attributes for each of a plurality of attribute criteria according to a survey result, and may determine a seller type by combining the calculated seller attributes. The product keyword recommendation device 100 may search for product keywords having the same type as the seller's type and recommend the product keyword to the seller.

2A is a flowchart illustrating a step-by-step process of clustering product keywords according to an embodiment of the present invention. 2B is a table showing attributes of product keywords according to an embodiment of the present invention. 2c is a table showing product keywords and seller types according to an embodiment of the present invention.

A first step (S101) is a step of receiving a list of top product keywords for each product category. The top product keyword list may include N product keywords having high rankings for each product category in the open market. For example, the top product keyword list may include the top 500 ranked product keywords for each product category, and the ranking is specified in the open market, and is arranged in descending order according to product sales volume or product keyword search volume. Alternatively, it may be determined by combining other sales information.

The product keyword recommendation device 100 may transmit a top product keyword list request message M1 to the open market server computer 200 and receive top product keyword list data D1 from the open market server computer 200. . The top product keyword list data D1 may include N product keywords having the highest sales volume for each product category.

The second step (S102) is a step of filtering product keywords. Product keywords that do not affect attribute determination can be deleted in advance through filtering, thereby shortening the clustering time and improving quality by recommending product keywords that are more suitable for the type of seller.

For example, the product keyword recommendation device 100 may store a list of brand names in advance in an auxiliary storage device ( 121 in FIG. 1 ). The product keyword recommendation device 100 may search the top product keyword list data D1 for product keywords including brand names, and then delete the searched product keywords from the top product keyword list data D1. Alternatively, the product keyword recommendation device 100 may predefine the product model name as a regular expression. The product keyword recommendation device 100 searches whether there is a product keyword matching a regular expression or including a product model name in the top product keyword list data D1, and then searches the searched product keyword in the top product keyword list data D1. can be deleted

The third step (S103) is a step of receiving a ranking page searched for with product keywords. The ranking page may include product content information and sales content information of products searched according to product keywords in the open market. For example, the ranking page may include product content information and sales content information such as product name, product category, sales volume, sales price, number of consumer reviews, and product registration date.

The product keyword recommendation device 100 transmits a ranking page request message M2 including product keywords remaining in the top product keyword list data D1 in a second step S102 to the open market server computer 200, , Ranking page data D2 corresponding to product keywords may be received from the open market server computer 200 . The ranking page data D2 may include product content information and sales content information for each product searched for by product keywords remaining in the upper product keyword list data D1 in the second step S102.

A fourth step (S104) is a step of filtering product keywords again. Clustering time can be shortened by deleting all but one of a plurality of product keywords that are duplicated or have similar meanings, and quality can be improved by recommending product keywords that are more suitable for the type of seller.

The product keyword recommendation device 100 stores the product keyword synonym list in the auxiliary storage device (121 in FIG. 1), and then synonyms among the product keywords remaining in the upper product keyword list data D1 in a second step S102. When there are two or more product keywords included in the list, all but one can be deleted. For example, except for one product keyword having the highest rank in the top product keyword list data D1, all remaining product keywords may be deleted.

Alternatively, the product keyword recommendation device 100 compares the ranking page data (D2) between the two product keywords remaining in the top product keyword list data (D1) in the second step (S102), and then, product content information and sales When content information is similar to each other, a product keyword having a low rank may be deleted from the top product keyword list data D1. For example, the product keyword recommendation device 100 compares product names in the ranking page data D2 retrieved with the i-th product keyword and the ranking page data D2 retrieved with the j-th product keyword. Then, in a similar case, the j-th product keyword having a lower rank may be deleted from the top product keyword list data D1. (i < j)

A fifth step (S105) is a step of generating a recommended product keyword pool. By selecting some of the product keywords filtered in the fourth step ( S104 ), the number of detail pages to be received by the product keyword recommendation device 100 may be reduced, and product keywords with low sales may be removed in advance.

Product keyword recommendation device 100, in the fourth step (S104), among the product keywords remaining in the top product keyword list data D1, products whose monthly sales are more than a standard amount or products whose sales rank is within the standard rank in the product category It is possible to create a pool of recommended product keywords by including product keywords that are searched for. For example, the standard amount may be 10 million won, and the standard ranking may be the top 10%.

The product keyword recommendation device 100 generates a recommended product keyword pool by including product keywords whose search volume has increased during the reference period among product keywords remaining in the top product keyword list data D1 in a fourth step S104. can For example, the reference period may be a period from now to 6 months ago.

A sixth step (S106) is a step of receiving a detailed page searched with product keywords. The detail page may include information not included in the ranking page among product content information and sale content information of products searched according to product keywords in the open market. For example, the detailed page may include information such as manufacturer, country of manufacture, proxy purchasing ratio, review content, and inquiry content.

The product keyword recommendation device 100 transmits a detailed page request message M3 including product keywords to the open market server computer 200, and provides detailed page data D3 corresponding to the product keyword from the open market server computer 200. can receive Product keywords may be included in a pool of recommended product keywords. The detail page data D3 may include product content information and sales content information not included in the ranking page data D2 for each product searched by product keyword.

A seventh step (S107) is a step of calculating attributes of product keywords. The attribute classification criterion may be a criterion for classifying product keywords and sellers' properties, and a plurality of attribute classification criteria may be used.

For example, as shown in FIG. 2B, four attribute classification criteria such as market search range, strategy establishment method, selection tendency, and implementation method can be used. The criteria for classifying the market search range attribute are the concentrated attribute, which indicates "tendency to prefer markets with limited product categories or consumers," and the distributed attribute, which indicates "tendency to prefer markets with various product categories or consumers." can include The criterion for classifying attributes of a strategy establishment method may include a warlike attribute representing “a propensity to win through competition” and a peaceful attribute representing “a tendency to prefer places with less competition”. The criterion for classifying selection propensity attributes may include an adventurous attribute representing “a propensity to try even when uncertain” and a stable attribute representing “a propensity to hope for a steady market with little change in market size”. The criteria for classifying performance method attributes are effect-type attributes that indicate “the propensity to take costs, efforts, and risks to achieve sales” and efficiency-type attributes that indicate “the propensity to prioritize efficiency in the process of achieving sales” can include

The product keyword recommendation device 100 acquires ranking page data D2 of product keywords obtained in the third step S103 and the product keywords included in the recommended product keyword pool in the sixth step S106. From the product content information and sales content information respectively included in the detailed page data D3 of one product keyword, an index related to an attribute classification criterion may be extracted, and attributes may be calculated according to the content. For example, the product keyword recommendation device 100 uses indicators such as product name, product category, sales volume, sales price, number of consumer reviews, product registration date, manufacturer, country of manufacture, purchase agent ratio, review content, and inquiry content. , generate a function that maps to concentrated and distributed attributes, warlike and peaceful attributes, adventurous and stable attributes, and effective and efficient attributes, and then rank page data (D2) of product keywords included in the recommended product keyword pool, and Attributes may be calculated by inputting product content information and sales content information respectively included in the detailed page data D3.

An eighth step (S108) is a step of determining the type of product keyword. In a seventh step ( S107 ), the type of product keyword may be determined by combining one or more attributes calculated for each product keyword. Since the product keyword type is determined by combining a limited number of attributes, the clustering time can be shortened.

For example, as shown in FIG. 2C, one type may be determined by combining attributes determined according to four attribute classification criteria, such as market search range, strategy establishment method, selection tendency, and execution method. The first type (T1) is a “brave combatant” type, which is determined when the attribute is “focused/warlike/adventurous/effective”. The second type (T2) is an “excellent tactician” type, which is determined when the attribute is “focused/warlike/adventurous/efficient”. The third type (T3) is a “visionary leader” type, which is determined when the attributes are “focused/warlike/stable/effective”. The fourth type (T4) is a “persistent boxer” type, which is determined when the attributes are “concentrated/warlike/stable/efficient”. The 5th type (T5) is the “eagle looking for an opportunity” type, which is determined when the attribute is “focused/peaceful/adventurous/effective”. The 6th type (T6) is a “Meticulous Analyst” type, which is determined when the attribute is “Concentrated/Peaceful/Adventurous/Efficient”. The 7th type (T7) is the "seer with insight" type, which is determined when the attribute is "focused/peaceful/stable/effective". The eighth type (T8) is a “peaceful Mother Teresa” type, which is determined when the attribute is “focused/peaceful/stable/efficient”. The ninth type (T9) is a “curious explorer” type, which is determined when the attribute is “dispersive/warlike/adventurous/effective”. The 10th type (T10) is a “manipulative businessman” type, which is determined when the attributes are “distributed/warlike/adventurous/efficient”. The 11th type (T11) is the “Versatile Leader” type, which is determined when the attribute is “Distributed / Aggressive / Stable / Effective”. The twelfth type (T12) is a “flexible multi-tasker” type, which is determined when the attributes are “distributed/aggressive/stable/efficient”. The thirteenth type (T13) is a “sensible trend leader” type, which is determined when the attributes are “distributed/peaceful/adventurous/effective”. The 14th type (T14) is a “popular idea bank” type, which is determined when the attributes are “distributed/peaceful/adventurous/efficient”. The 15th type (T15) is a “meticulous jeweler” type, which is determined when the attribute is “dispersive/peaceful/stable/effective”. The sixteenth type (T16) is a “real exhaustion seeker” type, which is determined when the attribute is “distributed/peaceful/stable/efficient”.

The product keyword recommendation device 100 may determine the type of product keyword by combining one or more attributes calculated for each product keyword in a seventh step ( S107 ). Also, the product keyword recommendation device 100 may generate type data D4 for each product keyword including type information for each of a plurality of product keywords, and store it in an auxiliary storage device ( 121 in FIG. 1 ).

3A is a flowchart illustrating a step-by-step process of matching product keywords according to seller types according to an embodiment of the present invention. 3B is a table showing survey questions according to an embodiment of the present invention.

A ninth step (S109) is a step of executing a survey to the seller. The type of seller can be judged according to the view selected by the seller for each item in the survey. The survey may be conducted more than once, and each time the survey is conducted, the type of seller may vary according to the survey results.

For example, as shown in FIG. 3B, one or more alternative questions and short-answer questions may be included. The result of the seller's response to each question can be used to determine the attribute presented in the seventh step (S107 in FIG. 2A).

The seller terminal device 300 may transmit a product keyword recommendation request message M4 to the product keyword recommendation device 100 and receive survey form data D5. The survey form data D5 may be an HTML document, and the seller terminal device 300 may display the survey form data D5 using a web browser. The seller may select one option from the alternative questions displayed on the seller terminal device 300 and input an answer from the short-answer question. The seller terminal device 300 may generate survey result data D6 including response contents for each item input by the seller and transmit the result data D6 to the product keyword recommendation device 100 .

A tenth step (S110) is a step of determining the type of seller. The product keyword recommendation device 100 may calculate attributes of the seller according to response contents for each question included in the survey result data D6.

For example, the product keyword recommendation device 100 selects one of concentrated or distributed attributes, one of warlike or peaceful attributes, one of adventurous or stable attributes, and one of effective or efficient attributes shown in FIG. 2B. You can choose one. At this time, after assigning a weight to the response content for each question, the properties of the seller may be calculated.

Also, the product keyword recommendation device 100 may determine the type of seller by combining attributes of the seller. For example, the product keyword recommendation device 100 may determine the type of seller by substituting attributes of the seller in the table shown in FIG. 2C.

An eleventh step (S111) is a step of matching the type of seller and the type of product keyword. After searching for product keywords with the same type as the seller's type, you can recommend it to the seller.

The product keyword recommendation device 100 searches for product keywords having the same type as the seller's type for each product category, and then generates the recommended product keyword list data D7 by including them. The product keyword recommendation device 100 may transmit recommended product keyword list data D7 to the seller terminal device 300, and the seller terminal device 300 may transmit product keywords included in the recommended product keyword list data D7. can be displayed The seller may select a product to be sold according to the product keyword displayed on the seller terminal device 300 .

Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and may vary within the scope of the detailed description of the invention and the accompanying drawings, as long as the spirit and effect of the present invention are not impaired. It can be implemented by making changes. It goes without saying that such embodiments fall within the scope of the present invention.

100: Product keyword recommendation device
110: central processing unit
120: storage device
121: auxiliary storage unit 122: main storage unit
130: I/O device
140: bus
200: open market server computer
300: seller terminal device
S101 ~ S108: Process of clustering product keywords
S109 ~ S111: Process of matching product keywords according to the type of seller

Claims (6)

A product keyword recommendation device,
Product keyword attributes are calculated for each of a plurality of attribute classification criteria, and the product keyword type is determined by combining the calculated product keyword attributes;
Whenever a survey is conducted, the seller's attributes are calculated for each of a plurality of attribute criteria according to the survey results, and the seller's type is determined by combining the calculated seller's attributes;
Search for product keywords with the same type as the seller's type and recommend them to the seller.
How to recommend product keywords by seller type.
The method of claim 1,
The product keyword recommendation device before calculating the attributes of product keywords,
Receive top product keyword list data, including N product keywords with high sales volume or search volume rankings for each product category, from an open market server computer;
Deleting product keywords including brand names or product keywords including product model names from the top product keyword list data;
How to recommend product keywords by seller type.
The method of claim 2,
A product keyword recommendation device,
Among the product keywords included in the product keyword synonym list, except for one product keyword ranked highest in sales volume or search volume, delete the remaining product keywords from the top product keyword list data;
Among two product keywords with similar product content information and sales content information on the ranking page, product keywords with low sales volume or search volume ranking are deleted from the top product keyword list data.
How to recommend product keywords by seller type.
The method of claim 3,
A product keyword recommendation device,
Create a pool of recommended product keywords by including product keywords that search for products with monthly sales greater than or equal to a standard amount, or product keywords that search for products whose sales rank is within the standard rank in a product category, or product keywords that have increased search volume during the standard period doing,
How to recommend product keywords by seller type.
The method of claim 4,
The product keyword recommendation device targets product keywords included in the recommended product keyword pool,
The criteria for distinguishing the market search scope property that distinguishes the concentrated property from the distributed property and
Strategies for distinguishing between warlike and peaceful attributes Criteria for distinguishing attributes
criteria for classifying selection propensity attributes that distinguish adventurous attributes from stable attributes;
Calculate the attributes of product keywords according to the attribute classification criteria that distinguishes effect-type attributes from efficiency-type attributes,
How to recommend product keywords by seller type.
The method of claim 1,
The product keyword recommendation device targets the survey results,
The criteria for distinguishing the market search scope property that distinguishes the concentrated property from the distributed property and
Strategies for distinguishing between warlike and peaceful attributes Criteria for distinguishing attributes
criteria for classifying selection propensity attributes that distinguish adventurous attributes from stable attributes;
Calculate the seller's attributes according to the attribute classification criteria that distinguishes the effective attribute from the efficiency attribute.
How to recommend product keywords by seller type.

Images