KR102753469B1 - Gift recommendation system chosen by recipients - Google Patents

Abstract

The present invention relates to a gift recommendation system selected by a recipient, and more specifically, to a gift recommendation system selected by a recipient, which generates a first recommendation gift list selected through information on a recipient, i.e., a gift recipient input by a purchaser, and a second recommendation gift list comprehensively considering open data on preferred products by gender, age, and occupation, and transmits the generated second recommendation gift list to the recipient, thereby allowing the recipient to directly select a desired gift, thereby maximizing the satisfaction of both the purchaser and the recipient.

Description

{Gift recommendation system chosen by recipients}

The present invention relates to a gift recommendation system selected by a recipient, and more specifically, to a gift recommendation system selected by a recipient, which generates a first recommendation gift list selected through information on a recipient, i.e., a gift recipient input by a purchaser, and a second recommendation gift list comprehensively considering open data on preferred products by gender, age, and occupation, and transmits the generated second recommendation gift list to the recipient, thereby allowing the recipient to directly select a desired gift, thereby maximizing the satisfaction of both the purchaser and the recipient.

Portable digital communication devices have become an essential element for many people living in the modern era. Consumers want to receive various high-quality services they want anytime and anywhere using portable digital communication devices.

Among the various high-quality services, the gift provision service is a service that provides information for receiving a gift to the gift recipient's portable digital communication device at the request of the gift giver, i.e. the purchaser, so that the gift is provided to the gift recipient.

Portable digital communication devices store various information for using the gift service and control communication circuits to transmit and receive various information for using the gift service with other devices.

In these gift-giving services, it is important to provide the best gift to the gift recipient, and recently, systems or methods are being developed to provide a gift recommendation list that reflects the recipient's tastes and preferences.

For example, Korean Patent Registration No. 10-2240260 and Korean Patent Publication No. 10-2019-0079582 disclose a fashion product gift recommendation method, a gift recommendation service method, and a gift recommendation server for executing the same, respectively. Their main technical configuration is characterized by providing a list of recommended gifts to the gift purchaser by analyzing information or preferences of the person receiving the gift, thereby improving the satisfaction of the recipient of the gift.

However, the above conventional technologies have limitations in providing the optimal gift to the gift recipient because the person who ultimately selects the gift from the recommended gift list is the gift purchaser.

In addition, usually, the gift information that is finally selected can be used when recommending gifts to other buyers. However, as mentioned above, since the final selection of a gift is made by the buyer, not the recipient, even if information or preferences about the recipient are used when creating a list of recommended gifts, the buyer's opinion inevitably intervenes, so there is a problem that the accuracy of creating a list of recommended gifts may be reduced.

1. Republic of Korea Patent Publication No. 10-2240260 (registered on April 8, 2021) 2. Republic of Korea Patent Publication No. 10-2019-0079582 (published on July 5, 2019)

The present invention has been made to solve the problems of the prior art as described above, and an object of the present invention is to provide a gift recommendation system selected by the recipient, which generates a second recommended gift list by comprehensively considering a first recommended gift list selected through recipient information and open data on preferred products by gender, age, and occupation, and transmits the generated second recommended gift list to the recipient so that the recipient can directly make the final selection of the gift, thereby improving the satisfaction of both the purchaser and the recipient.

In addition, the present invention has another purpose of providing a gift recommendation system selected by a recipient, which allows more accurate recommendations of gifts desired by the recipient by additionally generating a list of recommended gifts that takes into account the behavioral pattern of the recipient using the recipient terminal and providing the list to the recipient when the recipient fails to select a gift from the recommended list of gifts.

In order to achieve the above purposes, a gift recommendation system selected by a recipient according to the present invention is provided.
The present invention comprises a wireless communication capable buyer terminal and a recipient terminal, each of which is possessed by a buyer who wishes to purchase a gift and a recipient who is to receive the gift; a product index DB storing open data indices for products collectable from an open market; and a product DB storing a list of products that the buyers can purchase; and a gift recommendation server accessible via the buyer terminal and the recipient terminal, wherein the gift recommendation server generates a list of recommended gifts to be provided to the recipient using recipient information input via the buyer terminal and open data indices for products collectable from the open market, and transmits the list to the buyer terminal and the recipient terminal, and a gift selected by the recipient from the recommended gift list transmitted to the recipient terminal can be registered by the recipient to the gift recommendation server.

delete

At this time, the gift recommendation server includes a first data collection unit that receives purchaser information and recipient information input through a purchaser terminal, a second data collection unit that collects open data indicators for products from an open market, and a gift recommendation module that generates a recommended gift list using the recipient information collected through the first data collection unit and the open data indicators collected through the second data collection unit, and the gift recommendation module is characterized in that it includes a first extraction unit that primarily extracts a gift list to be recommended from the product DB using the recipient information collected through the first data collection unit, and a second extraction unit that secondarily extracts a gift list to be recommended to the recipient by matching the gift list extracted by the first extraction unit with the open data indicators collected by the second data collection unit using a CF algorithm.

In addition, the gift recommendation server further includes a third data collection unit that collects a recipient's terminal usage pattern for selecting a gift to be received from a list of recommended gifts, and the gift recommendation module is characterized in that it further includes a third extraction unit that uses an LF algorithm to match the analysis results of the recipient's terminal usage pattern collected by the third data collection unit with new products from a product DB to thirdly extract a list of gifts to be recommended to the recipient.

In addition, the gift recommendation server is characterized by further including an AI learning unit for learning the CF algorithm and LF algorithm to be used in the gift recommendation module.

delete

delete

delete

delete

According to the present invention, a first recommendation gift list selected through recipient information and a second recommendation gift list are generated by comprehensively considering open data on preferred products by gender, age and occupation, and the generated second recommendation gift list is transmitted to the recipient so that the recipient can directly make the final selection of the gift, thereby having an excellent effect of improving the satisfaction of both the purchaser and the recipient.

In addition, according to the present invention, if the recipient is unable to select a gift from the recommended gift list, a recommended gift list that takes into account the recipient's behavioral pattern using the recipient's terminal is additionally generated and provided to the recipient, thereby enabling more precise recommendations of gifts desired by the recipient, and at the same time, having the additional effect of recommending a more accurate gift list through artificial intelligence (AI) learning using various pieces of information that can identify the recipient's tendencies and preferences.

Figure 1 is a conceptual diagram illustrating a gift recommendation system selected by a recipient according to the present invention.
Figure 2 is a flow chart sequentially showing a gift recommendation method selected by a recipient according to the present invention.
FIG. 3 is a diagram conceptually illustrating an algorithm used in the first and second gift extraction steps of the present invention shown in FIG. 2.
FIG. 4 is a diagram conceptually illustrating an algorithm used in the third gift extraction step of the present invention shown in FIG. 2.
FIG. 5 is a drawing exemplarily showing a decision-making process performed through a purchaser terminal and a recipient terminal of the present invention shown in FIG. 1.

Hereinafter, preferred embodiments of a gift recommendation system selected by a recipient according to the present invention will be described in detail with reference to the attached drawings.

FIG. 1 is a diagram conceptually illustrating a gift recommendation system selected by a recipient according to the present invention, FIG. 2 is a flowchart sequentially illustrating a gift recommendation method selected by a recipient according to the present invention, FIG. 3 is a diagram conceptually illustrating an algorithm used in the first and second gift extraction steps of the present invention illustrated in FIG. 2, FIG. 4 is a diagram conceptually illustrating an algorithm used in the third gift extraction step of the present invention illustrated in FIG. 2, and FIG. 5 is a diagram exemplarily illustrating a decision-making process performed through a purchaser terminal and a recipient terminal of the present invention illustrated in FIG. 1.

The present invention relates to a gift recommendation system selected by a recipient, which generates a first recommendation gift list selected through information on a recipient, i.e., a gift recipient input by a purchaser, and a second recommendation gift list comprehensively considering open data on preferred products by gender, age, and occupation, and transmits the generated second recommendation gift list to the recipient so that the recipient can directly select a desired gift, thereby maximizing the satisfaction of both the purchaser and the recipient. First, a gift recommendation system (10) selected by a recipient according to the present invention (hereinafter, referred to as 'gift recommendation system (10)') may largely include a purchaser terminal (100), a recipient terminal (200), a gift recommendation server (300), and a gift recommendation platform (400), as illustrated in FIG. 1.

The above-mentioned buyer terminal (100) and recipient terminal (200) are each carried by a person who wishes to order a gift, i.e., a buyer, and a person who will receive a gift ordered by the buyer, i.e., a recipient, and are terminals capable of wireless communication on which a gift recommendation platform (400) provided by a gift recommendation server (300) described below is executed, and service information provided from the gift recommendation server (300) can be received through the gift recommendation platform (400).

In addition, the purchaser terminal (100) and recipient terminal (200) are devices for executing the gift recommendation platform (400) described below, and may include a laptop computer, a smart phone, a tablet, a wearable computer, etc., and may mean any terminal device capable of accessing a web/mobile site or installing and executing a service-only application.

In addition, the purchaser terminal (100) and recipient terminal (200) can perform overall service operations such as service screen configuration, data input, data transmission/reception, and data storage under the control of a web/mobile site or dedicated application.

Next, the gift recommendation server (300) acts as a typical server that creates and manages a gift recommendation platform (400) that the buyer and recipient can access through the buyer terminal (100) and the recipient terminal (200), respectively, and may include a product index DB (310), a product DB (320), an AI learning unit (330), and a platform providing unit (340).

To explain in more detail, the product indicator DB (310) serves to store open data indicators for products that can be collected from the open market (20). The open data indicators stored in the product indicator DB (310) may include product reviews, number of purchases, evaluation scores, evaluation rankings, age-based preferences, gender-based preferences, brands, model names, etc.

Next, the product DB (320) serves to store a list of products that buyers can purchase. Preferred products can be stored by classifying them by gender, age, and occupation, and products classified by the above criteria can be stored by classifying them again by type.

Next, the AI learning unit (330) plays a role in learning AI algorithms to be used for gift extraction in the gift recommendation platform (400) described later, and methods such as machine learning can be used.

The above AI learning unit (330) may include a first learning module and a second learning module. The first learning module and the second learning module are configured to learn the Collaborative Filtering algorithm (hereinafter referred to as the 'CF algorithm') and the Latent Factorization algorithm (hereinafter referred to as the 'LF algorithm'), respectively, which are used in the AI recommendation process for generating a gift recommendation list. A more detailed description thereof will be provided later.

Next, the platform providing unit (340) plays a role in creating and managing the gift recommendation platform (400) described later, and can be configured to provide data stored in the product index DB (310) and product DB (320) to the gift recommendation platform (400), or to perform updates to additionally apply content learned by the AI learning unit (330) to the gift recommendation platform (400).

In addition, although not included in the city, the gift recommendation server (300) may include a configuration for storing and managing member information of buyers, recipients, etc., a configuration for storing and managing evaluations such as reviews of recommended gift lists or gift usage reviews written by buyers and recipients, and a configuration for processing gift cost payments by buyers.

Next, the gift recommendation platform (400) is configured to be created and managed by the gift recommendation server (300) so that the buyer and recipient can access it through the buyer terminal (100) and the recipient terminal (200), respectively, and may include applications (apps), programs, web pages, etc. that are installed and run on the buyer terminal (100) and the recipient terminal (200) through a network.

To explain in more detail, the gift recommendation platform (400) may include a first data collection unit (410), a second data collection unit (420), and a gift recommendation module (430). First, the first data collection unit (410) serves to receive buyer information and recipient information input through a buyer terminal (100). As shown in FIG. 5, the buyer information may include name, contact information, etc., and the recipient information may include gender, occupation, age, etc., and the amount of the product to be gifted may be included in the buyer information or the recipient information.

Next, the second data collection unit (420) collects open data indicators for products sold from the open market (20). As described above, the open data indicators may include product reviews, number of purchases, evaluation scores, evaluation rankings, etc., and the open data indicators newly collected through the second data collection unit (420) may be additionally stored in the product indicator DB (310) of the gift recommendation server (300).

Next, the gift recommendation module (430) plays a role in generating a list of recommended gifts to be provided to the recipient through AI recommendations using recipient information collected through the first data collection unit (410) and open data indicators collected through the second data collection unit (420), and may include a first extraction unit (431) and a second extraction unit (432).

To explain in more detail, the first extraction unit (431) primarily extracts a list of gifts to be recommended from the product DB (320) of the gift recommendation server (300) using the recipient information collected through the first data collection unit (410). As shown in FIG. 3, a list of recommended gifts suitable for the recipient can be extracted from the list of products stored in the product DB (320) by considering the gender, age, occupation, etc. of the recipient collected through the first data collection unit (410) (hereinafter referred to as the 'first recommended gift list').

Next, the second extraction unit (432) may match the first recommendation gift list extracted by the first extraction unit (431) with the open data indicators collected by the second data collection unit (420) using the CF algorithm to extract the recommendation gift list (hereinafter referred to as the 'second recommendation gift list') to be initially transmitted to the recipient.

The above CF algorithm is a recommendation algorithm based on big data that collects the preferences of various users and is used to predict the user's areas of interest. Since it is an algorithm that is already being used in various fields, a detailed description thereof will be omitted.

In the present invention, a CF algorithm is pre-learned to extract a second recommendation gift list suitable for a recipient by matching a first recommendation gift list extracted by considering gender, age, and occupation, etc. by a first learning module equipped in an AI learning unit (330) of a gift recommendation server (300) with open data indices collected by a second data collection unit (420) or stored in a product index DB (310), thereby extracting a second recommendation gift list suitable for a recipient by AI recommendation in the second extraction unit (432).

In addition, the secondary recommended gift list extracted by the secondary extraction unit (432) can be transmitted to the purchaser terminal (100) and the recipient terminal (200) by the gift recommendation platform (400).

At this time, the secondary recommended gift list can be transmitted to the purchaser terminal (100) in the form of a simple list, and the secondary recommended gift list can be transmitted to the recipient terminal (200) in the form of a page link, so that when the recipient opens the page link, the secondary recommended gift list can be configured to open through the recipient terminal (200), as shown in FIG. 5.

In addition, the gift purchase price can be transmitted together with the second recommended gift list to the purchaser terminal (100), and the purchaser can check the recommended gift list and proceed with payment of the purchase price through the purchaser terminal (100).

At this time, payment of the above purchase amount may be made as a prepayment together with the determination of the gift amount when entering the recipient information using the purchaser's terminal (100), and since various payment methods currently in use may be used, a more detailed description thereof will be omitted.

And, when the recipient selects a gift he or she wishes to receive from the list of secondary recommended gifts transmitted to the recipient terminal (200), the selected product can be sent to the recipient.

Meanwhile, the gift recommendation platform may further include a third data collection unit (440) that collects a behavioral pattern of a recipient in selecting a gift from a secondary recommended gift list, i.e., a usage pattern of a recipient's terminal (200). The usage pattern of the recipient's terminal (200) collected by the third data collection unit (440) may be utilized for learning for more accurate gift recommendations, i.e., learning in an AI learning unit (330) of a gift recommendation server (300), or may be used to generate an additional recommended gift list (hereinafter, referred to as a 'tertiary recommended gift list') when the recipient is unable to select a gift he or she wishes to receive from the secondary recommended gift list.

To explain in more detail, the gift recommendation module (430) may further include a third extraction unit (434), and the third extraction unit (434) matches new products from the product DB (320) through analysis of the usage pattern of the recipient terminal (200) collected by the third data collection unit (440) to extract a third recommendation gift list, and the aforementioned LF algorithm may be utilized.

The above LF algorithm is a method of extracting a list of gifts to be recommended to the recipient by using latent factors of the recipient. The latent factors may include usage patterns of the recipient's terminal (200) collected through the third data collection unit (440), i.e., products clicked by the recipient using the recipient's terminal (200), time spent on product images, preference clicks (like, dislike, etc.), time spent on detailed information images, number of product thumbnail clicks, time and number of times of revisiting the list, etc., which will be described in more detail later.

In the present invention, the LF algorithm is pre-learned by matching the usage pattern of the recipient terminal (200) collected by the third data collection unit (440) with the second recommendation gift list initially transmitted to the recipient terminal (200) by the second learning module equipped in the AI learning unit (330) of the gift recommendation server (300) with new products stored in the product DB (320) to extract a new recommendation gift list, i.e., a third recommendation gift list, so that the third extraction unit (434) can newly extract a third recommendation gift list to be newly transmitted to the recipient by AI recommendation from the product DB (320).

In addition, the third extraction unit (434) may play a role in transmitting the third recommendation gift list extracted based on the AI recommendation result by the LF algorithm to the purchaser terminal (100) and the recipient terminal (200), respectively.

The recipient can open the 3rd recommended gift list in the form of a newly sent page link and select the gift they wish to receive. Once the gift selection is confirmed, the selected product can be sent to the recipient.

At this time, the usage pattern of the recipient's terminal (200) in the process of selecting a gift to receive from the third recommended gift list can also be collected by the third data collection unit (440), and if the recipient is unable to select a gift to receive from the third recommended gift list, the process of extracting a new recommended gift list from the product DB (320) by AI recommendation using the usage pattern of the recipient's terminal (200) newly collected through the third data collection unit (440) can be repeated.

Meanwhile, the gift recommendation system (10) according to the present invention may further include a fourth data collection unit (450) equipped in the gift recommendation platform (400). The fourth data collection unit (450) serves to collect evaluations such as service usage reviews and gift usage reviews from the purchaser terminal (100) and the recipient terminal (200). The evaluations collected through the fourth data collection unit (450) may be additionally stored in the product index DB (310) of the gift recommendation server (300) or utilized for learning the CF algorithm and LF algorithm through the AI learning unit (330).

In addition, the gift recommendation server (300) may separately have and manage an evaluation DB (not shown) for storing evaluations written by buyers and recipients.

Meanwhile, a gift recommendation method selected by a recipient according to the present invention (hereinafter referred to as a "gift recommendation method") relates to a service provision method that enables a recipient to directly select a gift he or she wishes to receive by sending a list of recommended gifts to the recipient using the aforementioned gift recommendation system (10), and its configuration, as shown in FIG. 2, can largely include a data storage and AI learning step (S10), a recipient information input step (S20), a first gift extraction step (S30), a second gift extraction step (S40), a recommended gift list sending step (S50), a gift selection step (S60), and a gift sending step (S70).

First, the data storage and AI learning step (S10) is a preprocessing, i.e., preparation step for providing a gift recommendation service using a gift recommendation platform (400). It is a process of storing data in the product index DB (310) and gift DB constituting the gift provision server (300), and learning algorithms to be used in the gift list recommendation process through the AI learning unit (330).

That is, in the above product index DB (310), open data indices for various products that can be collected from various open markets (20) can be stored and converted into big data, and in the product DB (320), a list of various products that buyers can purchase can be stored, categorized by gender, age, occupation, preference, or type.

Next, learning through the AI learning unit (330) may include a first learning step (S12) and a second learning step (S14) for AI recommendation to be used in the gift extraction process described later. First, in the first learning step (S12), CF algorithm learning may be performed using the first learning module.

That is, in the first learning step (S12), learning can be performed on a CF algorithm that extracts a second recommended gift list by matching the first recommended gift list extracted first using the recipient information collected by the first data collection unit (410) with the open data index of the open market (20) collected by the second data collection unit (420), thereby enabling AI recommendation in the second gift extraction step (S40) described later.

In addition, as shown in Fig. 3, in the process of matching the first recommended gift list with the open data indicators, weights may be given to specific indicators among the open data indicators, such as the number of purchases and preferences, so that a more sophisticated recommended gift list can be extracted, or a matching range may be expanded by setting a search similar group for age and occupation, so that selection errors may be prevented from occurring when recommending the recommended gift list.

At this time, the search similar group can be set to similar groups in their 30s and 50s if the age is in the 40s, and if the occupation is office workers, the similar group can be set to housewives or students according to gender.

Next, in the second learning step (S14), learning of the LF algorithm using the second learning module can be performed, which is for AI recommendation in the third gift extraction step (S90) described later.

That is, in the second learning step (S14), the LF algorithm can be learned by using the usage pattern of the recipient's terminal (200) for the recommended gift list sent to the recipient, that is, the product clicked by the recipient, the time spent on the product image, the preference click (like, dislike, etc.), the time spent on the detailed information image, the number of product thumbnail clicks, the time and number of times the list is revisited, etc., which are collected by the third data collection unit (440) of the gift recommendation platform (400). As shown in FIG. 4, if the usage pattern of the recipient's terminal (200) is displayed by dividing it by product included in the recommended gift list and then dimensioning it into space to display it, the product preferred by the recipient, that is, the product with a high probability of being selected, can be identified.

At this time, the time spent on a specific product image (stay time) can be counted until another event occurs or until a background process transition event occurs, and for improved accuracy, statistics can classify it as unused time if the stay time exceeds a preset threshold.

In the above, the first and second learning stages (S12, S14) were described as preliminary stages, but it is of course possible to configure the AI recommendation in the gift recommendation platform (400) to be continuously or periodically updated by continuously or periodically performing the first and second learning stages (S12, S14) by utilizing newly added data during the service provision process.

Next, the recipient information input step (S20) is a process in which a buyer who wishes to purchase a gift inputs the information of the recipient who will receive the gift using the buyer terminal (100). The buyer first inputs the buyer information such as name, contact information, and payment method while connected to the gift recommendation platform (400) generated by the gift recommendation server (300) using the buyer terminal (100), and then the recipient information including gender, occupation, age, etc. can be input.

Next, the first gift extraction step (S30) is a process of initially extracting a list of recommended gifts to be provided to the recipient from the product DB (320) of the gift recommendation server (300). The first list of recommended gifts can be extracted by utilizing internal data corresponding to the information of the recipient, such as gender, occupation, and age, entered in the recipient information input step (S20), i.e., preference data according to gender, occupation, age, etc., stored separately in the product DB (320).

Next, the second gift extraction step (S40) is a process of extracting a second recommended gift list to be provided to the recipient by utilizing the first recommended gift list extracted in the first gift extraction step (S30) and open data indicators for products that can be collected from an open market (20), and may include an open data collection step (S42) and a first product indicator matching step (S44).

To explain in more detail, the open data collection step (S42) is a process of collecting open data indicators for products sold from various open markets (20) using the second data collection unit (420) of the gift recommendation platform (400). The open data indicators may include product reviews, number of purchases, evaluation scores, evaluation rankings, age-based preferences, gender-based preferences, brands, model names, etc., and the collected open data indicators may be stored in the product indicator DB (310) of the gift recommendation server (300).

Next, the first product index matching step (S44) is a process of extracting a second recommended gift list to be provided to the recipient by AI recommendation by matching the first recommended gift list extracted in the first gift extraction step (S30) with the open data indexes of the corresponding products collected in the open data collection step (S42), i.e., products included in the first recommended gift list. The first product index matching step (S44) can be performed in the second extraction unit (432) of the gift recommendation platform (400), and the CF algorithm learned in the first learning step (S12) described above can be used for the AI recommendation.

Next, the recommended gift list sending step (S50) is a process of sending the recommended gift list extracted by AI recommendation in the secondary gift extraction step (S40), i.e., the secondary recommended gift list, to the buyer and the recipient, respectively. The secondary extraction unit (432) of the gift recommendation platform (400) sends the secondary recommended gift list extracted by AI recommendation to the buyer terminal (100) and the recipient terminal (200), respectively, thereby informing the buyer of the recommended gift type and purchase price, and notifying the recipient of the fact that a gift has been provided by the buyer, while allowing the recipient to select the gift he or she wishes to receive.

At this time, as described above, the gift purchase amount and payment link can be transmitted together with the second recommended gift list to the purchaser terminal (100), and the purchaser can proceed with payment of the purchase amount through the purchaser terminal (100) while checking the recommended gift list.

In addition, a secondary recommended gift list in the form of a page link may be transmitted to the recipient terminal (200) so that when the recipient opens the page link, the secondary recommended gift list opens and the recipient can check detailed information on products included in the list.

Next, the gift selection step (S60) is a step in which the recipient selects a gift that he or she wishes to receive. The recipient can carefully review the products included in the secondary recommended gift list transmitted to the recipient terminal (200), select the gift he or she wishes to receive, and register the selected gift on the gift recommendation platform (400) using the recipient terminal (200).

Next, the gift sending step (S70) is a process of sending the gift selected in the gift selection step (S60) to the recipient. The address to which the gift is sent uses the recipient address entered by the purchaser in the recipient information input step (S20). If the purchaser does not know the recipient's address, the gift selection step (S60) may be configured so that the recipient directly inputs the address at which he or she wishes to receive the gift using the recipient terminal (200).

Meanwhile, the gift recommendation method according to the present invention may further include a third gift extraction step (S90). The third gift extraction step (S90) is a process for re-extracting a list of recommended gifts to be provided to the recipient when the recipient is unable to select a gift they wish to receive in the gift selection step (S60).

To explain in more detail, the third gift extraction step (S90) may include a recipient pattern collection step (S92) and a product index matching step (S94). First, the recipient pattern collection step (S92) is a process of collecting a behavioral pattern performed by the recipient to select a gift in the gift selection step (S60), i.e., a usage pattern of the recipient terminal (200). The usage pattern of the recipient terminal (200), such as a product clicked by the recipient using the recipient terminal (200), time spent on a product image, preference clicks (like, dislike, etc.), time spent on a detailed information image, number of product thumbnail clicks, time and number of list revisits, etc., may be collected through the third data collection unit (440) of the gift extraction platform.

Next, the product index matching step (S94) is a process of creating a new recommended gift list, i.e., a third recommended gift list, through AI recommendation by matching additional products from the product DB (320) through analysis of the recipient terminal (200) usage pattern collected in the recipient pattern collection step (S92). Hereinafter, the product index matching step (S94) will be referred to as the second product index matching step (S94) to distinguish it from the first product index matching step (S44) of the second gift extraction step (S40) described above.

That is, the second product index matching step (S94) is a process of generating a new third recommendation gift list through AI recommendation using the LF algorithm learned in the second learning step (S14) described above, and the third extraction unit (434) of the gift recommendation platform (400) can match the analysis results of the usage pattern of the recipient terminal (200) collected by the third data collection unit (440) using the learned LF algorithm, i.e., the product clicked by the recipient using the recipient terminal (200), the time spent on the product image, the preference click (like, dislike, etc.), the time spent on the detailed information image, the number of product thumbnail clicks, the time and number of times the list is revisited, etc., with the products stored in the product DB (320) to generate a third recommendation gift list, and the newly generated third recommendation gift list can be transmitted back to the purchaser terminal (100) and the recipient terminal (200) by the third extraction unit (434).

As described above, when the extraction of the third recommended gift list is completed, the recommended gift list sending step (S50) and the gift selection step (S60) described above are repeated, and this repetition of the process can be continued until the final gift selection is made by the recipient.

In addition, even if a gift is selected by the recipient in the gift selection step (S60), the recipient pattern collection step (S92) described above may be performed. The recipient terminal (200) usage patterns collected in the recipient pattern collection step (S92) may be additionally stored in the product index DB (310) of the gift recommendation server (300), or may be utilized for learning the CF algorithm and LF algorithm through the AI learning unit (330).

Meanwhile, the gift recommendation method according to the present invention may further include a review registration step (S80). The review registration step (S80) is a process in which the buyer and the recipient input evaluations, such as usage reviews and gift usage reviews, for the gift recommendation service using the buyer terminal (100) and the recipient terminal (200), respectively. The evaluations input by the buyer terminal (100) and the recipient terminal (200) may be collected by the fourth data collection unit (450) of the gift recommendation platform (400) as described above and stored in the evaluation DB of the gift recommendation server (300), and the evaluation results may be utilized for learning the CF algorithm and the LF algorithm through the AI learning unit (330).

Therefore, according to the gift recommendation system according to the present invention as described above, the second recommendation gift list is generated by comprehensively considering the first recommendation gift list selected through the recipient information and open data on preferred products by gender, age, and occupation, and the generated second recommendation gift list is transmitted to the recipient so that the recipient can directly make the final selection of the gift, thereby improving the satisfaction of both the purchaser and the recipient, and in the case where the recipient is unable to select a gift from the recommended gift list, the recommended gift list is additionally generated by considering the behavioral pattern of the recipient using the recipient terminal (200) and provided again to the recipient, thereby enabling more precise recommendations of gifts desired by the recipient, and at the same time, recommending a more accurate gift list through artificial intelligence (AI) learning using various information that can identify the tendencies and preferences of the recipient, etc., has various advantages.

Although the above-described embodiments have described the most preferred examples of the present invention, it is not limited to the above-described embodiments, and it is obvious to those skilled in the art that various modifications are possible within a scope that does not depart from the technical spirit of the present invention.

The present invention relates to a gift recommendation system selected by a recipient, and more specifically, to a gift recommendation system selected by a recipient, which generates a first recommendation gift list selected through information on a recipient, i.e., a gift recipient input by a purchaser, and a second recommendation gift list comprehensively considering open data on preferred products by gender, age, and occupation, and transmits the generated second recommendation gift list to the recipient, thereby allowing the recipient to directly select a desired gift, thereby maximizing the satisfaction of both the purchaser and the recipient.

10: Gift recommendation system 20: Open market
100: Buyer terminal 200: Recipient terminal
300: Gift recommendation server 310: Product index DB
320: Product DB 330: AI Learning Department
340: Platform provider 400: Gift recommendation platform
410: First data collection unit 420: Second data collection unit
430: Gift recommendation module 431: 1st extraction section
432: 2nd extraction section 434: 3rd extraction section
440: 3rd data collection unit 450: 4th data collection unit
S10: Data storage and AI learning stage S12: First learning stage
S14: Second learning stage S20: Recipient information input stage
S30: 1st gift extraction stage S40: 2nd gift extraction stage
S42: Open data collection stage S44: First product indicator matching stage
S50: Recommended gift list sending stage S60: Gift selection stage
S70: Gift delivery stage S80: Review registration stage
S90: 3rd gift extraction stage S92: Recipient pattern collection stage
S94: Second product index matching stage

Claims (7)

A buyer terminal and a recipient terminal capable of wireless communication, each possessed by a buyer who wishes to purchase a gift and a recipient terminal who wishes to receive the gift,
It consists of a product index DB that stores open data indicators for products that can be collected from open markets, a product DB that stores a list of products that buyers can purchase, and a gift recommendation server that can be accessed through buyer terminals and recipient terminals.
The above gift recommendation server is,
A list of recommended gifts to be provided to the recipient is created using the recipient information entered through the above buyer terminal and open data indicators for products that can be collected from the open market, and is transmitted to the buyer terminal and the recipient terminal.
A gift recommendation system selected by a recipient, characterized in that a gift selected by the recipient from the list of recommended gifts transmitted to the recipient terminal can be registered by the recipient to the gift recommendation server.
In paragraph 1,
The above gift recommendation server is,
It includes a first data collection unit that receives buyer information and recipient information input through a buyer terminal, a second data collection unit that collects open data indicators for products from an open market, and a gift recommendation module that generates a recommended gift list using the recipient information collected through the first data collection unit and the open data indicators collected through the second data collection unit.
The above gift recommendation module is,
A first extraction unit that primarily extracts a list of gifts to be recommended from the product DB using the recipient information collected through the first data collection unit;
A gift recommendation system selected by a recipient, characterized in that it includes a secondary extraction unit that secondarily extracts a gift list to be recommended to a recipient by matching the gift list extracted by the first extraction unit with open data indicators collected by the second data collection unit using a CF algorithm.
In the second paragraph,
The above gift recommendation server is,
It further includes a third data collection unit that collects the recipient's terminal usage pattern for selecting a gift to be received from the recommended gift list.
The gift recommendation module is characterized in that it further includes a third extraction unit that extracts a list of gifts to be recommended to the recipient by matching new products from the product DB with the results of the analysis of the recipient's terminal usage patterns collected by the third data collection unit using the LF algorithm. The gift recommendation system selected by the recipient is characterized in that the gift recommendation module further includes a third extraction unit that uses the LF algorithm to extract a list of gifts to be recommended to the recipient.
In the third paragraph,
The above gift recommendation server is,
A gift recommendation system selected by a recipient, characterized in that it further includes an AI learning unit for learning the CF algorithm and LF algorithm to be used in the gift recommendation module.


delete delete delete