KR102301535B1 - Shopping mall marketing system for shopping mall customers and their acquaintances - Google Patents

Abstract

The present invention relates to a shopping mall marketing system for shopping mall visiting customers and acquaintances of customers. The shopping mall marketing system for shopping mall visiting customers and acquaintances of customers stores information on all actions regarding a product search, purchase, and return in an online shopping mall of a customer as product-of-interest information, then analyzes the stored information by a deep learning-based analysis engine to analyze which products with which features the customer mainly purchases and which products with which features the customer avoids purchasing recently, then transmits an advertisement message for a product corresponding to the preference of the customer to the customer, and accumulates a prescribed ratio of corresponding product sales for the existing customer who has transmitted the advertisement message as points when the customer receiving the product advertisement message uses a text message or SNS to transfer the product advertisement message to an acquaintance of the customer and then the acquaintance of the customer clicks on the link of the received product advertisement message to connect to the online shopping mall and then purchases the product.

Description

Shopping mall marketing system for shopping mall visitors and customers' acquaintances

The present invention relates to a shopping mall marketing system for a customer visiting a shopping mall and an acquaintance of the customer, and more particularly, by analyzing the customer's existing product purchase history with an analysis engine based on Deep Learning to advertise products of customer preference About the shopping mall marketing system for shopping mall visitors and customers' acquaintances, in which a message is sent, an advertisement message delivered by an existing customer is sent back to an acquaintance to promote the product, and when the acquaintance purchases the product, he or she receives the benefit of earning points will be.

With the development of multimedia technology and the popularization of Internet use, the number of shopping malls that not only provide various product information through the Internet but also sell various products online has increased. became possible

As the number of online shopping malls increases as described above, product purchasers have to select one of a large number of online shopping malls to purchase products.

Conventionally, a shopping mall is recommended based on price ratio through comparison of each shopping mall for the same product, or a shopping mall is recommended according to users' recommendation information and the number of views.

In addition, each information site or shopping mall on the Internet provides a variety of information, and customized services are provided that collect and analyze this information to select and provide a user's expected product.

In Korean Patent Laid-Open Patent No. 10-2011-0043215, "Product price comparison recommendation system and its operation method," Internet domain information, product information, related product information, and price information that match or relate to a keyword search word input during web activity of a user We propose a product price comparison and recommendation system that uses the Internet search more quickly and efficiently by searching, storing, and displaying them in a user-centered manner.

However, since such conventional systems simply provide products in the order of high and low prices, not only do they not properly reflect the product price difference, but they also fail to properly reflect the user's preferences and recommendations from people who use the product. Efficient shopping malls and product recommendations are not made.

In addition, according to such a conventional system, since a product is recommended based on simple user's recommendation information and the number of views for the same product, it is difficult to see it as a more accurate reflection of recommendations of people who have used the product.

Republic of Korea Patent Publication No. 10-2020-0103202 “Method of providing product recommendation service based on deep learning neural network” is a service that recommends customized products based on the body shape and purchase history of online shopping mall users using a deep learning neural network. to provide. Republic of Korea Patent Publication No. 10-2020-0103202 uses a deep learning neural network to generate data necessary for a product recommendation service of an online shopping mall, analyze and classify or cluster it through artificial intelligence, and provide user feedback information on recommended products. By creating and utilizing , customized recommendation products are provided through artificial intelligence judgment through a deep learning neural network.

However, Korean Patent Laid-Open Patent No. 10-2020-0103202 is only at the level of the prior art that recommends a product list to the user and provides it as a recommended product, and does not suggest a plan to continuously promote existing customers of the online shopping mall and attract new customers. can't After all, in the fiercely competitive online shopping mall business, there are limits to the promotion of shopping malls, customer acquisition, and sales increase.

Also, for products with low sales rates in online shopping malls, it is necessary to analyze the cause of the problem and provide a method to increase the sales rate of the products.

Republic of Korea Patent Publication No. 10-2019-0055957 “Method of providing social network-based online shopping mall promotion service” states that posts written by a subscriber directly on an SNS server, posts with positive responses, and posts with comments are registered in the subscriber’s friend list By utilizing the function of being broadcast to other subscribers, the positive reaction or comment function of the SNS server is given to the online shopping mall promotion screen or product promotion screen, so that the promotional screen is propagated through the subscriber's friend list.

However, Korean Patent Laid-Open Patent No. 10-2019-0055957 does not analyze the characteristics of products purchased by customers to advertise customized products, and it is not a method in which an existing customer directly transmits an advertisement message to his/her acquaintances through SNS. Therefore, not only does the customer's acquaintances lack the motivation to access the shopping mall and purchase the product, but also when the customer's acquaintance purchases the product from the shopping mall through the advertisement message, there is no benefit to the existing customer who sent the advertisement message, so the advertising ripple effect is reduced. There is a problem that is not high.

Patent Publication No. 10-2011-0043215 Patent Publication No. 10-2020-0103202 Patent Publication No. 10-2019-0055957

The present invention is to solve the problem of the prior art, and stores information on all actions from customer's online shopping mall to product search, purchase and return as interest product information, and then stores the stored information as deep learning (Deep Learning) Learning) based analysis engine to analyze what characteristics customers mainly purchase products with and which characteristics they avoid purchasing, and then send advertisement messages for products that meet customer preferences to customers. And, the customer who has received the product advertisement message delivers the product advertisement message to his/her acquaintance using a text message or SNS.

When an acquaintance of a customer clicks a link in a product advertisement message received by a customer to access an online shopping mall and purchases a product, a certain percentage of the product's sales are credited to existing customers who have sent the advertisement message as points. It aims to provide a shopping mall marketing system for

In order to achieve the above object, the present invention provides a customer shopping information collection unit that collects information on all actions of a customer from product search, purchase and return in an online shopping mall and stores it in a customer product shopping DB; a product feature analysis unit that analyzes customers' product purchase reviews and return reasons for each product stored in the product DB, and analyzes price, quality, functionality, preferred age, and preferred gender corresponding to the characteristics of the product; Using the product characteristic information analyzed by the product characteristic analysis unit, the product characteristics searched and clicked by the customer stored in the customer product shopping DB, the purchased product characteristics, and the returned product characteristics are based on deep learning a customer preference analysis unit that analyzes the characteristics of a preferred product by product category by customer by analyzing it with an analysis engine of ; a product advertisement transmission unit that searches the product DB for products having similar characteristics based on the analysis contents of the customer preference analysis unit and transmits a product advertisement message to the corresponding customer; and an acquaintance of the customer who received the product advertisement message sent to the customer by the product advertisement transmission unit clicks the link in the product advertisement message to access the online shopping mall It includes a point accumulating unit for accumulating points to customers.

When the product advertisement transmission unit transmits a product advertisement message, it analyzes the inventory quantity of the product in the inventory DB, and for products with a low sales rate, it may further include a product price discount unit that applies a discount to the product selling price of the product advertisement message by a certain percentage. have.

and a product category lifecycle analyzer configured to analyze the sales rate of products stored in the inventory DB and set a sales lifecycle for each product category, wherein the product advertisement transmission unit determines that the sales life of the corresponding product is maintained for a certain period through the product category lifecycle analysis unit. For a product that is determined to last, a product advertisement message may be transmitted to the corresponding customer.

In addition, the product category life cycle analysis unit analyzes the inventory quantity of the corresponding product in the inventory DB and applies a discount rate to the sales life cycle of the product category to which the product belongs for products with a low sales rate to calculate the life cycle of individual products. and a product life cycle calculation module, wherein the product advertisement transmission unit may transmit a product advertisement message to a corresponding customer for a product whose sales life of the corresponding product is determined to last for a certain period as a result of the calculation of the individual product life cycle calculation module.

And for a product with a low sales rate in the inventory DB, based on the analysis contents of the product characteristic analysis unit, one core problem with low sales of the product is extracted, and each problem cause category from a plurality of problem cause categories that cause the core problem It may further include a product problem analysis unit for subclassing the problem causes by category belonging to, and extracting one or more major problem causes from among the subdivided problem causes.

The type and number of the plurality of problem-causing categories may be predefined in advance according to the nature of a product, such as an industrial product or food.

The type and number of the plurality of problem-causing categories may be predefined in advance according to the nature of a commodity such as an industrial product or food and the nature of the extracted core problem.

According to the present invention having the configuration as described above, the following effects can be achieved.

By collecting product searches, purchases, and returns from customers' online shopping malls, and analyzing the characteristics of those products with an analysis engine based on deep learning, it is possible to identify which characteristics of products each customer prefers, and then determine customer preferences. You can search for a product and send a product advertisement message.

As described above, as it is possible to advertise products customized to customers with the deep learning-based analysis engine, it is natural that there is an advertisement effect for existing customers. Existing customers send advertisement messages to their acquaintances, and acquaintances If you click on an advertisement message to access the shopping mall and purchase a product, a certain percentage of sales will be credited to the existing customer who sent the advertisement message, which not only improves the shopping mall operator’s sales, but also benefits existing customers. of voluntary advertising and customer loyalty can be increased.

When the above advertising method is applied using a micro-influencer, not only the customer expansion power is further increased, but also the sales of the shopping mall company can be dramatically increased in a short time.

1 is a functional configuration diagram of a shopping mall marketing system for a customer visiting a shopping mall and an acquaintance of the customer according to the present invention;
2 is a flowchart illustrating each step of the shopping mall marketing system for a customer visiting a shopping mall and an acquaintance of the customer according to the present invention;
3 is a detailed flowchart of a process of transmitting a product advertisement message;
4 is an example in which, when a product sold in a shopping mall is an industrial product, a core problem is extracted for a product with a low sales rate, and a plurality of problem cause categories that cause the core problem are determined
5 is an example of sub-categorizing problem causes by category in a plurality of problem cause categories determined in FIG.
6 is an example in which one or more major problem causes are extracted from the sub-classified problem causes after sub-categorizing the problem causes by category in FIG.
7 is another example of extracting a core problem for a product with a low sales rate, determining a plurality of problem cause categories that cause the core problem, and then subdividing the problem causes by category

Advantages and features of the present invention, and methods for achieving them, will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings.

However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms.

In the present specification, the present embodiment is provided to complete the disclosure of the present invention, and to fully inform those of ordinary skill in the art to which the present invention pertains to the scope of the present invention.

And the invention is only defined by the scope of the claims.

Accordingly, in some embodiments, well-known components, well-known operations, and well-known techniques have not been specifically described to avoid obscuring the present invention.

In addition, like reference numerals refer to like elements throughout the specification, and terms used (referred to) herein are for the purpose of describing the embodiments and are not intended to limit the present invention.

In this specification, the singular also includes the plural unless specifically stated otherwise in the phrase, and elements and operations referred to as 'comprising (or having)' do not exclude the presence or addition of one or more other elements and operations. .

Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used with the meaning commonly understood by those of ordinary skill in the art to which the present invention belongs.

Also, terms defined in commonly used dictionary are not to be interpreted ideally or excessively unless defined.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

1 to 3 , the shopping mall marketing system 100 for a customer visiting a shopping mall and an acquaintance of the customer includes a customer shopping information collection unit 110 , a product characteristic analysis unit 120 , a customer preference analysis unit 130 , It includes a product advertisement transmission unit 140 and a point accumulation unit 150 , and may further include a product price discount unit 160 , a product category lifecycle analysis unit 170 , and a product problem analysis unit 180 . In addition, the database includes a customer product shopping DB 101 , a product DB 102 , and a stock DB 103 .

The customer shopping information collection unit 110 collects information on all customer behavior from the online shopping mall to the customer's product search, purchase, and return, and stores it in the customer product shopping DB 101 for each customer ID ( S210 ) .

The product characteristic analysis unit 120 analyzes the customer's product purchase reviews and return reasons for each product stored in the product DB 102 to analyze the price, quality, functionality, preferred age, and preferred gender corresponding to the characteristics of the product. .

At this time, the product characteristic analysis unit 120 collects product purchase reviews and return reasons using a web crawling function, and then performs a syntax analysis to determine that customers have reasonable prices for various types of products. Do you think that it is cheap, do you think it is expensive, do you evaluate the quality as medium or high, what is the problem among the quality (defective operation, durability, poor assembly, poor soldering, etc.), is the delivery fast or the delivery is late? Analyze whether there is a problem and whether there is a problem of damage to the delivered product.

In addition, it is possible to analyze how the functionalities of the product are evaluated, for example, the moisturizing function, wrinkle improvement function, whitening function of cosmetics, and the age and gender of purchasing the product. Such analysis is performed on a product-by-product basis.

The customer preference analysis unit 130 uses the product characteristic information analyzed by the product characteristic analysis unit 120 to perform deep learning on the characteristics of products searched, clicked, purchased, and returned by the customer stored in the customer product shopping DB 101 . (Deep Learning) based analysis engine to analyze the characteristics of the preferred product for each product category for each customer ( S220 ). Using the characteristics of the product analyzed by the product characteristic analysis unit 120, each customer analyzes the purchased, returned, or clicked product, and analyzes the product with which characteristics they prefer by using a deep learning-based analysis engine to infer. have.

Here, the product category includes a variety of body/hair products, companion animal products, automobile supplies, household goods, kitchen supplies, cosmetics/perfume, etc. Among body/hair products, hair loss shampoo, beauty soap, foot cream, etc. This applies to various types of products. Products such as cheese gum, green grape gum, and cat hair clippers may correspond to the pet category. Cosmetics/perfume can be subdivided into children's moisturizing products, children's natural products, women's moisturizing products, women's wrinkle improvement products, and women's whitening products.

Even in the female moisturizing product category, the price level, quality level, and efficacy level vary depending on the product, and the deep learning algorithm analyzes which category the customer prefers, and what level of price, quality, etc., is preferred within the product category. can be analyzed.

The product advertisement transmission unit 140 searches the product DB 102 for products having similar characteristics based on the analysis contents of the customer preference analysis unit 130 and transmits a product advertisement message to the corresponding customer ( S230 ). The product advertisement message may be transmitted using text messages or SNS such as KakaoTalk.

The point accumulation unit 150 receives the product advertisement message sent by the product advertisement transmission unit 140 to the customer, and when an acquaintance of the customer clicks the link in the product advertisement message to access the online shopping mall and purchases the product, the A certain percentage is accumulated as points to the existing customer who has transmitted the advertisement message ( S240 ).

For example, when a customer receives advertisement contents including an advertisement link from the shopping mall server, the customer may transmit the contents to his or her acquaintances through SNS. In the advertisement message, the customer's own identification number or phone number is stored as an identifier. By sending an advertisement message to an acquaintance, when the acquaintance clicks the advertisement link and accesses the shopping mall and purchases the product or clicks the link and accesses the shopping mall to purchase another product, the advertisement message is sent for about 3-5% of product sales Earn points for one existing customer. Because the identifier of the existing customer is stored in the advertisement message, when an acquaintance clicks on the advertisement message and accesses the shopping mall, it is possible to know which advertisement message the acquaintance accessed through the advertisement message. .

The product price discount unit 160 analyzes the inventory quantity of the product in the inventory DB 103 when the product advertisement transmission unit 140 transmits the product advertisement message, and for products with a low sales rate, the product selling price of the product advertisement message. A certain percentage discount is applied ( S231 ).

By monitoring the inventory quantity of products in the inventory DB 103 at regular intervals, it is possible to analyze which products are selling well and which products have a low sales rate. The degree to which a product's sales rate is low depends on the level of sales compared to the period. If the sales rate is very low, the product selling price is discounted a lot, and if the sales rate is not too low, a small discount is used. The discount rate is determined between a minimum of 5 and a maximum of 30%, and the discount rate varies depending on the type of product and the level of sales.

The product category life cycle analysis unit 170 analyzes the sales rate of products stored in the inventory DB 103 and sets the sales life cycle for each product category.

The product category includes various items such as body/hair products, companion animal products, automobile products, household products, kitchen products, cosmetics/perfume, and the like. If the average is obtained by comprehensively analyzing the sales rate of products stored in the inventory DB 103, the sales life cycle can be calculated for each product category.

If the sales life cycle is calculated by product category, it may be different from one year to six months for companion animal products, one year for cosmetics/perfume, and three years for automobile products.

The product advertisement transmission unit 140 transmits a product advertisement message to the corresponding customer for a product whose sales life is determined to last for a certain period through the product category lifecycle analysis unit 170 ( S232 ).

If the sales life cycle of cosmetics is 6 months, if company A's B cosmetics were recently released and sales started only two months ago, it can be determined that the sales life of cosmetics B will continue for a certain period of time. For such a product, a product advertisement message is transmitted to a customer who has a taste for the product.

The individual product life cycle calculation module 171 analyzes the inventory quantity of the corresponding product in the inventory DB 103, and applies a discount rate to the sales life cycle of the product category to which the product belongs to a product with a low sales rate. to calculate

The product advertisement transmission unit 140 transmits a product advertisement message to the corresponding customer for a product whose sales life is determined to last for a certain period as a result of the calculation of the individual product life cycle calculation module 171 ( S233 ).

As a result of analyzing the inventory quantity of Cosmetics B of Company A, the sales rate may be low. In this case, a discount rate according to the sales rate of the product is applied individually to one year of the sales life cycle of the cosmetic category.

As described above, the discount rate is determined between a minimum of 5 and a maximum of 30%, and the discount rate varies depending on the type of product and the level of sales volume.

If a discount rate of 20% is applied to one year of the sales life cycle because the sales rate of B cosmetics is very low, the life cycle of individual products = 12 months × (1 - 0.2) = 9.6 months. In other words, it is predicted that cosmetic B will no longer be sold after 9.6 months from the initial release date. 9.6 months equals 9 months and 18 days.

If 6 months have passed since the launch of cosmetic B or the start of shopping mall sales, the remaining life cycle is 3.6 months (3 months and 18 days). Therefore, it is possible to transmit a product B advertisement message to a customer who prefers cosmetic B (customer who likes cosmetic B).

However, even if the sales life of cosmetic B is 3 months and 18 days, if the sales outlook of cosmetic B is not good due to cosmetic C because the functions of cosmetics C of other companies similar to cosmetic B are superior, recommendations and products for cosmetic B The transmission of the advertisement message may be withheld, and a product advertisement message for cosmetics C of another company may be transmitted.

Here, the condition is that there must be another cosmetic product similar to cosmetic B, and only when it is judged that the function is superior to cosmetic B, a product advertisement message for other competing cosmetic products with superior function is transmitted. In other words, if the function of cosmetic B is not similar or superior to that of cosmetic B, the product advertisement message for cosmetic B is transmitted because the sales life of cosmetic B is 3 months and 18 days.

The product problem analysis unit 180 extracts one or more major problem causes based on the analysis contents of the product characteristic analysis unit 120 for a product having a low sales rate in the inventory DB 103 ( S250 ). As mentioned above, the cause of the problem may be various things such as malfunction, poor durability, assembly defect, soldering defect, processing defect, delivery delay, etc., if the product is an industrial product. In the case of food, product damage, freshness, delivery delay, etc. , packaging defects, etc.

The product problem analysis unit 180 extracts one core problem of poor sales of the product based on the analysis contents of the product characteristic analysis unit 120 for a product with a low sales rate in the inventory DB 103, and causes the core problem From the plurality of problem cause categories, the problem causes for each category belonging to each problem cause category are subdivided, and one or more major problem causes are extracted from among the subdivided problem causes ( S250 ).

Here, the type and number of the plurality of problem-causing categories may be predefined in advance according to the nature of the product, such as an industrial product or food.

Alternatively, the type and number of the plurality of problem-causing categories may be predefined in advance according to characteristics of products such as industrial products or food and the characteristics of the extracted core problems.

Referring to FIG. 4 , if a product with a low sales rate in the inventory DB 103 is a humidifier, one core problem of low sales of the humidifier is extracted based on the analysis contents of the product characteristic analysis unit 120 . In this case, the extraction method collects product reviews and return reasons, then extracts the most frequent words through syntax analysis, and analyzes customer reviews and return reasons to analyze and extract the main problems.

In the example of FIG. 4, an example in which 'operational failure' of the humidifier is extracted as a core problem is shown. After extracting the core problem, a plurality of problem cause categories that cause the core problem, 'malfunction', are determined. The example of FIG. 4 shows a case in which the categories 'workers', 'parts', and 'delivery', which are three problem causes that cause 'operational malfunction' of the humidifier, are determined. The problem cause categories may be three as in the example of FIG. 4 , four or more.

The problem cause category refers to the main causes that cause the core problem.

The three categories of 'worker', 'parts', and 'delivery' are determined in advance according to the quality of products such as industrial products or food. That is, in the example of FIG. 4 , in the case of a humidifier, which is an industrial product, three problem-causing categories such as 'worker', 'part', and 'delivery' may be predefined. In the case of food such as fish or vegetables, the types and number of categories that cause a plurality of problems may be different. That is, in the case of food, four categories such as 'worker', 'sanitation', 'delivery', and 'material' may be used.

Alternatively, the type and number of categories of a plurality of problem causes may be predefined in advance in consideration of the nature of industrial products such as humidifiers or food such as fish and the extracted core problems. That is, even for the same industrial product, the type and number of categories that cause a problem may vary depending on what the core problem of the industrial product is.

After all, the types and numbers of a plurality of problem-causing categories should be stored in advance in the database of the system. When a core problem of poor sales of a product is extracted based on the analysis contents of the product characteristic analysis unit 120, a plurality of problem cause categories matching the core problem are retrieved from the database of the system. As shown in FIG. 4 , when the problem cause category is determined to be three of 'worker', 'part', and 'delivery', the problem cause is subdivided into categories belonging to each problem cause category.

Referring to FIG. 5, when the 'worker' problem cause category is subdivided, 'function' and 'quality awareness' are extracted, and when the 'part' problem cause category is subdivided, 'type', 'lifetime', and 'low cost parts' are extracted If the category of the cause of the 'delivery' problem is subdivided, it is subdivided into 'damage during delivery', 'delivery method', and 'delivery person'. As described above, the subdivided result of each problem cause category is performed by analyzing the customer's purchase reviews and return reviews, and extracting words that appear frequently along with each problem cause category.

For example, if you analyze the words that appear along with 'part' in the purchase review, such as 'part' is strange, wears out quickly, or uses cheap parts, the part is strange in the system's database => kind, Cheap parts => cheap parts, parts wear out quickly => You can subcategorize the categories that cause problems with 'parts' by searching for lifespan. 'Delivery' and 'Worker' are performed in the same way. After subcategorizing each problem cause category, one or more major problem causes should be extracted from the subdivided problem causes.

Referring to Figure 6, among the subcategories of the 'worker' problem cause category, 'quality awareness' is derived as the main problem cause, and among the subcategories of the 'parts' problem cause category, 'lifetime' is derived as the main problem cause, and ' Among the subcategories of the 'delivery' problem cause category, the example in which 'damage during delivery' was derived as the main cause of the problem is shown. Purchase reviews and return reviews are analyzed and extracted by the frequency of words used by customers, and two main problem causes may be derived as a result of subdivision of each problem cause category. 6 shows a case in which one major problem cause is derived from each problem cause category as a result of subclassification of each problem cause category.

In summary, the core problem of humidifiers is 'operational failure'. analysis can be provided. By automatically analyzing and providing major problem causes, the shopping mall operator notifies the manufacturer of the problem, and the manufacturer can receive a lot of help in improving the problem.

Fig. 7 shows another example, and is a case in which 'delivery delay' of fruit, which is a food, is extracted as a key problem. The multiple problem cause categories that cause 'delivery delay' are 'system', 'worker' and 'material'. It is extracted as a cause, and as a result of subclassification of the 'worker' problem cause category, 'lack of manpower' is extracted as the main cause of the problem, and 'manual work of packaging' is extracted as the main cause of the subclassification of the 'material' problem cause category.

As such, if 'online receipt confirmation process', 'lack of personnel', and 'manual packaging' are extracted as the main causes of 'delivery delay' of fruits, fruit vendors simplify the online application process and increase the number of workers. And by automating the packaging, the 'delivery delay' problem can be solved.

According to the present invention, when a customer of an online shopping mall transmits a product advertisement message to an acquaintance, and the acquaintance clicks a link in the message to access the shopping mall and purchases a product, a certain portion of the sales can be accumulated as their points. can be This product advertisement and sales promotion system can bring benefits to both existing customers and online shopping mall companies.

In addition, by analyzing customers' product reviews and returned reviews, it is possible to analyze the cause of various problems such as whether the core problem of the product is malfunction or delay in delivery, so it is an intelligent solution that can improve problems with existing products. An online shopping mall may be provided.

Although the present invention has been described in detail with reference to preferred embodiments so far, those skilled in the art to which the present invention pertains can implement the present invention in other specific forms without changing the technical spirit or essential features thereof. The embodiments are to be understood in all respects as illustrative and not restrictive.

And, the scope of the present invention is to be specified by the claims to be described later rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts are included in the scope of the present invention. should be interpreted as

100: shopping mall marketing system for shopping mall visitors and customers' acquaintances
110: customer shopping information collection unit
120: product characteristic analysis unit
130: customer preference analysis unit
140: product advertisement transmission unit
150: Points
160: product price discount unit
170: Product category life cycle analysis unit
171: Individual product life cycle calculation module
180: Product problem analysis unit
101: Customer product shopping DB
102: product DB
103: Inventory DB

Claims (7)

a customer shopping information collection unit that collects information on all actions of a customer from an online shopping mall to product search, purchase, and return and stores it in a customer product shopping DB;
a product feature analysis unit that analyzes customers' product purchase reviews and return reasons for each product stored in the product DB, and analyzes price, quality, functionality, preferred age, and preferred gender corresponding to the characteristics of the product;
Using the product characteristic information analyzed by the product characteristic analysis unit, the product characteristics searched and clicked by the customer stored in the customer product shopping DB, the purchased product characteristics, and the returned product characteristics are based on deep learning a customer preference analysis unit that analyzes the characteristics of a preferred product by product category by customer by analyzing it with an analysis engine of ;
a product advertisement transmission unit that searches the product DB for products having similar characteristics based on the analysis contents of the customer preference analysis unit and transmits a product advertisement message to the corresponding customer; and
An acquaintance of the customer who received the product advertisement message sent to the customer by the product advertisement transmission unit clicks the link in the product advertisement message to access the online shopping mall and then purchases the product. Including a point accumulator that accumulates points to
Further comprising a product category life cycle analysis unit that analyzes the sales rate of products stored in the inventory DB and sets the sales life cycle for each product category,
The product advertisement transmission unit transmits a product advertisement message to the customer for a product whose sales life is determined to last for a certain period through the product category lifecycle analysis unit,
The product category life cycle analysis unit analyzes the inventory quantity of the product in the inventory DB, and applies a discount rate to the sales life cycle of the product category to which the product belongs for a product with a low sales rate to calculate the life cycle of each product. Includes a cycle calculation module,
The individual product life cycle is obtained by the following formula (1),
Life cycle of individual product = Sales life cycle of the product category to which the product belongs × (1 - discount rate according to the sales rate of the product) ㆍㆍㆍㆍㆍㆍ Formula (1)
The product advertisement transmission unit transmits a product advertisement message to the customer for a product whose sales life is determined to last for a certain period as a result of the calculation of the individual product life cycle calculation module,
For a product with a low sales rate in the inventory DB, one core problem with poor sales of the product is extracted based on the analysis of the product characteristic analysis unit, and from a plurality of problem cause categories that cause the core problem, to each problem cause category It further includes a product problem analysis unit that subdivides the problem causes by category to which it belongs, and extracts one or more major problem causes from among the subdivided problem causes,
The type and number of the plurality of problem-causing categories are predefined in advance according to the nature of products such as industrial products or food and the nature of the extracted core problems,
The product problem analysis unit extracts the most frequent words through syntax analysis after collecting product reviews and return reasons, and extracts one core problem by analyzing customer reviews and return reasons,
The type and number of the plurality of problem cause categories vary depending on what the core problem is, and even for the same industrial product, the type and number of problem cause categories vary depending on what the core problem of the industrial product is, and a plurality of problem causes The type and number of categories are stored in advance in the database of the system, and if one core problem with poor sales is extracted based on the analysis of the product characteristic analysis unit, a plurality of problem-causing categories matching the core problem are searched for in the database of the system. bring,
By analyzing customers' purchase reviews and return reviews, and extracting words that appear frequently along with each problem cause category, the cause of problems by category belonging to each problem cause category is subdivided, and after analyzing customers' purchase reviews and return reviews, the customer's Customers and customers visiting a shopping mall, characterized in that one or more major problem causes are derived from each problem cause category as a result of subclassification of each problem cause category by extracting one or more major problem causes from among the sub-classified problem causes using the frequency of words Shopping mall marketing system for acquaintances of the.
The method according to claim 1,
When the product advertisement transmission unit transmits a product advertisement message, it analyzes the inventory quantity of the corresponding product in the inventory DB, and for products with a low sales rate, further comprising a product price discount unit that applies a discount to the product selling price of the product advertisement message by a certain percentage A shopping mall marketing system for shopping mall visitors and customers' acquaintances.
delete delete delete delete delete

Images