KR20220034401A - Blending tea recommendation system - Google Patents

Abstract

An embodiment of the present invention provides an AI concierge service that recommends blending tea suitable for a consumer by using artificial intelligence (AI). In addition, provided is an interface that allows a consumer to directly customize blending tea. In addition, one embodiment stores information of blending tea purchased or customized by the consumer together with identification information of a consumer in a unique pattern and provides the same to the consumer so as to allow the consumer to easily make an order by using a pattern when the consumer makes the order at an offline store.

Description

Blending tea recommendation system {BLENDING TEA RECOMMENDATION SYSTEM}

The present invention provides a way for consumers to more easily access blending tea.

Tea is a favorite food with a long history as a beverage made by brewing tea leaves in hot water. Tea has various tastes, flavors, and effects depending on the type of tea leaves. In particular, Blending Tea, which is made by mixing different types of tea leaves, can produce various tastes and aromas depending on the type of tea leaves to be combined and the ratio of the combination. And the types of blending tea that can be manufactured according to the combination are very vast. In addition, various additional ingredients or substances such as herbs and fruits may be combined in the blending tea.

On the other hand, in the existing tea market, each tea leaf is sold individually or the ingredients for several types of blending tea preferred by the public are sold together. Alternatively, some types of preferred blending teas are made and sold at offline stores. As such, although the theoretical number of types of blending tea is very diverse, the types of blending tea actually consumed are limited. This is because the number of different types of tea leaves and the number of cases that occur according to their blending can be a great barrier to entry for consumers who are new to tea. Nevertheless, many consumers search for many recipes to find blending tea that suits their taste and find their own tea while tasting tea through actual blending, and services that try to reflect these consumers' needs are is appearing in

A technology that can recommend a suitable blending tea for each consumer is required. To this end, there is a need for a system that can analyze consumer tastes based on individual consumer information and recommend a more suitable blending tea by reflecting the consumer's purchase history and feedback.

There is a need for a service that allows consumers to repeatedly purchase customized blending tea while maintaining the same identity.

One embodiment provides a method for recommending a beverage suitable for a user, the method comprising: collecting user information and product information in a database; extracting metadata by receiving information from the database; transmitting the extracted metadata to a control unit; selecting a beverage suitable for the user through deep learning using the metadata as an input value; recommending the selected beverage to a user; generating a pattern including information on the beverage selected by the user and information on the user; And it provides a beverage recommendation method comprising the step of providing the generated pattern to the user.

In addition, after the step of recommending the selected beverage to the user, the method may further include providing an interface for the user to customize the recommended beverage.

In addition, the pattern may be a barcode or a QR code.

In addition, the user information may include one or more of user information, user preference information, concierge history information, user order history information, and teasity information.

In addition, the step of providing the generated pattern to the user may include providing graphic information to the user's user terminal of the user, or providing the pattern in a physically printed form.

In addition, the step of selecting a beverage suitable for the user may include one or more of a user's preference item, a user's non-preference item, and a user's SNS data analysis as the input value.

In addition, in one embodiment, the deep learning includes an input layer including the input value, a hidden layer that readjusts weights through comparison with existing metadata through machine learning, and an output layer from which the selected beverage is derived as a result. may include

In addition, the collecting of the user information and product information in the database may provide an interface through which one or more of a user's preference item and a user's non-preference item can be input.

Also, the beverage may be blending tea in which a plurality of tea leaves are combined.

In addition, the pattern may include information on types of tea leaves constituting the blending tea.

According to an embodiment of the present invention, customers can come in contact with their own new blending tea made based on their own flavor that was difficult to find or implement on their own. Therefore, compared to the case where an individual makes the blending tea by themselves, both time and money can be saved and the desired tea can be accessed.

By using the AI deep learning system, it is possible to utilize individual customer data and data on the tea market, and based on this, blending tea that fits the trend can be provided to the market.

1 is a schematic diagram of an AI concierge/tsnality reading system (recommended order system).
2 is a diagram schematically illustrating an AI concierge system.
3 is a diagram of a database 200 .
4 is a diagram schematically showing a concierge system through deep learning.
5 is a flowchart for using a concierge according to an embodiment.

The present invention will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings herein. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in a variety of different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the technical field to which the present invention belongs It is provided to fully inform the possessor of the scope of the invention, and the present invention is only defined by the scope of the claims. On the other hand, the terms used 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. As used herein, "comprises" or "comprising" does not exclude the presence or addition of one or more other elements, steps, in addition to the recited elements, steps. Terms such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

An embodiment of the present invention provides an AI (Artificial Intelligence) concierge service, which is a service combining an AI recommendation system and a customer custom service. Consumers who are new to tea often do not know which tea to drink. In other words, it is to provide a service to first recommend a suitable tea to a customer who is interested in tea, and then customize the flavor of the recommended tea according to the customer's request. The blending tea made by reaching the custom stage is combined with customer information and given a unique identification number to create a unique personality pattern for each customer, and the pattern can be checked through the customer's mobile device or website. Here, teasnality is a compound word of Tea and Personality, and refers to the unique personality and characteristics of blending tea. In addition, the pattern may be attached or engraved on a container such as a tumbler. Various types of patterns such as barcodes and QR codes can be applied to the pattern, and by recognizing the pattern, you can check the recipe for blending tea, for example, the type of tea leaves, the mixing ratio, etc. In other words, since the consumer has information about his/her own blending tea as a pattern and can order using it, there is no need to list the blending tea combinations every time. The service method using the pattern will be described later.

The AI recommendation system may be a system primarily targeted at consumers who are new to cars or are not yet familiar with cars. When a customer registers as a member, when the SNS account is linked, the customer's gender, age group, and keywords of interest are extracted. If it is not linked with the SNS account, the gender, age group, and keywords of interest are extracted when registering as a member on the website. The types of teas to be recommended to customers are primarily selected by comparing and analyzing customer data extracted at the time of membership registration with existing data accumulated in a search engine, for example, keywords of interest by age group, popular cafe-related categories, etc. Thereafter, one type of blending tea most suitable for the customer may be recommended from among predetermined blending tea candidates through a more sophisticated recommendation questionnaire. In addition, customer feedback can be converted into data and reflected in the deep learning system to provide more accurate recommendations. All the data in this process is accumulated in the AI recommendation system and can be used for new menu development, customer custom service, and subscription service that provides various information about tea to customers.

Customer custom service is a service that allows customers to customize products recommended through the AI recommendation system. The candidates for the blending tea provided by the AI recommendation system are blending teas made from recipes that have already been developed by experimentally combining various ingredients, but if the customer wants to customize it more, it can be changed according to the customer's taste. For example, if the customer wants to make the grapefruit flavor in the finally recommended blending tea stronger, it is a system that enables the addition of strength and ingredients. The customer custom service is an additional option that the customer can select, and if the customer is satisfied with the blending tea provided through the AI recommendation system, the customer custom service may be omitted.

Through the AI concierge service, which combines such an AI recommendation service and customer custom service, customers can choose a blending tea that suits their taste, and based on the finally selected tea information and customer information, a customer’s own teasity pattern is created. It can be created, issued and given to a customer. At this time, the given pattern may be confirmed through the application of the smartphone, or may be provided by being engraved on a tumbler. When ordering tea from an offline store, customers can easily select and order the desired tea combination through the teasity pattern without a separate long order. For example, when a teasity pattern is scanned in an offline store, the types of blending tea selected by the customer may be checked together with the customer's information. A plurality of blending teas may be included in one pattern, and a customer may select and order a blending tea from among them. The pattern can be provided in a variety of ways. As described above, it may be provided to the user terminal through the AI concierge service, or may be provided attached or engraved on a related product such as a tumbler. Alternatively, a pattern may be provided to the user in the form of a sticker so that the user can attach it to a tumbler or the like that he owns.

Alternatively, when a customer orders a custom car at an offline store, a pattern may be printed on the receipt so that the customer can check and store the pattern information for the custom car ordered.

1 is a diagram showing a schematic appearance of an AI concierge/tsnality reading system (recommended order system).

The network 110 refers to a communication network as a network environment. The communication network includes a mobile communication network, a wired/wireless Internet, and the like, through which the user terminal 120 and the AI concierge/tsnality reading system 130 can be connected. The user terminal 120 is a communication-capable terminal possessed by consumers, customers, and the like, and may include, for example, a mobile phone, a smart phone, a tablet, a computer, a notebook computer, and the like. The AI concierge/tissue reading system (hereinafter, concierge) 130 is a system for providing an AI concierge service to consumers, and may be, for example, a server providing an AI concierge service. The module included in the concierge 130 is not limited to a physical configuration, and may represent functional elements or processing steps performed through the control unit. The concierge 130 will be described in detail below.

2 is a diagram schematically illustrating an AI concierge system.

The database 200 is a basic database for executing the concierge, user information DB 210, product information DB 220, user preference information DB 230, concierge history DB 240, user order history DB 250, It may include a teasality information DB 260 . More details of the database 200 will be described in detail with reference to FIG. 3 .

The metadata extraction module 300 refers to a series of processes of collecting, quantifying and refining data required for deep learning and API (Application Programming Interface) interworking from the database 200, or a functional module for performing the same. Specifically, the metadata extraction module 300 includes a user information metadata extraction module 310 that digitizes the user information DB, a product information metadata extraction module 320 that digitizes the product information DB, and a user preference information DB. User preference information metadata extraction module 330 that quantifies, concierge history metadata extraction module 340 that quantifies concierge history DB, user order history metadata extraction module 350 that quantifies user order history DB, It may include a faciality information metadata extraction module 360 that digitizes the faciality information DB.

The central control system 400 as a central control unit stores and purifies the metadata extracted from each database to enable deep learning and model expression, and data processing as needed. As the cycle rotates, it is finally fed back through the product recommendation module, result data processing module, and final result display module. This can be reflected in the update of the corresponding DB according to each data standard.

The product recommendation module 500 stores the refined data processed by the central control system 400 in a temporary memory, and stores the data selected as the user's preference item using the second preference item selection module in temporary data, (510), using the second non-preference item selection module to store the data selected as the user's non-preference item in temporary data (520), and when the user agrees, news feed information such as Instagram, Facebook, etc. The data is processed and accumulated so that it can be used in the module corresponding to the next result data processing module 600 by extracting and crawling (530) using a technology such as api or python selenium. Finally, the more users use the collected data through deep learning programs (tensorflow, pandas, and keras), the more accurate prediction is possible.

The result data processing module 600 integrates 610 data based on the data processed in the product recommendation module 500 and then visualizes the data to facilitate use in the final result display module 700. (620) The result is stored in temporary data. The exact operation mechanism will be described later.

The final result display module 700 may include a graphic and pattern processing module 710 for outputting a pattern and result text suitable for the user's aesthetics based on the data integrated and visualized by the result data processing module 600 . . The data is sent back to the central control system, and changes are reflected in the DB

3 shows a detailed description of the database 200 .

The item corresponding to the user information DB 210 may include basic information such as age and gender of users who have signed up through social subscription (including social media linkage) and site self-registration.

Factors corresponding to the product information DB 220 may include taste, fragrance, efficacy, price, whether a gift set is present, a sense of season, whether ICE is recommended, a mood, a sales ranking, and the like. As the number of service use increases through deep learning, the probability of a concierge can increase.

The user preference information DB 230 crawls, collects, and stores keywords related to launching items such as beverages, health functional foods, and cafes by gender, age, and lifestyle of users who have an existing order history. This is RowData that has not been processed yet, and it is data that needs processing in the future. Various APIs are required for selective extraction.

The concierge history DB 240 may store selection details and SNS feed information of a user who has previously used the concierge service. It can also include an entry to record whether or not it actually led to a purchase.

The user order history DB 250 is a combination of order information items of users with actual purchase histories, signature menus preferred by gender and age group, and concierge. Also, it is possible to store information about the items for which purchases have not been made.

The teasnality information DB 260 may include the title and content (recommended base tea, recommended aroma, recommended fragrance, recommended search, recommended nutrient) of the appearance information to be used in the output part corresponding to the final stage of the cycle. .

4 is a diagram schematically showing a concierge system through deep learning.

Using konlpy, tensorflow, pandas, and Numpy, you can use the deep learning program in the product recommendation module that fits the deep learning schematic of FIG. The input layer contains metadata, SNS data obtained through SNS extraction, and the selection module value selected by the user through the selection module in the form of a data frame (data that is the target of the initial program must be historical data, but high-quality metadata For data extraction, it can be carried out by replacing it with pre-surveyed data.), in the hidden layer, through machine learning, the weight is readjusted through comparison with the existing metadata, so that the result of tsnality is derived from the output layer. do. This will be stored in the DB again and will be used as a data frame for input later. In other words, the more the program is used, the more accurate the concierge becomes.

5 is a flowchart for using a concierge according to an embodiment.

Step C01 is a process of visiting a site, and step C02 represents a page movement of the user. Step C03 proceeds to step C04 if the page moved by the user in step C02 is the concierge page, otherwise it returns to the previous step. Step C04 is a step that asks for the user's consent to link with SNS and provide personal information. If you do not agree to the provision, after notifying that the provision of the teasity information pattern is not possible, the window for entering basic personal information is displayed, and if you agree, it moves to step C05. Step C05 prepares to run a deep learning learning program by analyzing feed information through SNS api linkage and crawling. In step C06, if the account to be linked is a non-public account, a warning window is displayed again through step C04-1 and returns to step C04. In the case of a public account, it moves to step C07, and if the concierge checks three options for question 1 (check preference), it goes to the next step, otherwise it is not moved.

Step C08 also performs the same process as step C07.

In step C09, depending on whether the previously started step C05 is completed, step C09-1 (if not completed) moves to step C10. When step C05 is completed, the last node moves to step C011 (concierge result window and final custom window provision page), and after giving custom permission, moves to step C12 and ends.

Although the above description has been focused on blending tea combining tea, the present invention is not limited thereto, and the technical idea of the present invention may be applied to beverages other than tea. For example, in the case of coffee or other beverages other than tea, when various raw materials can be combined and customized, the AI concierge service provided in the embodiment of the present invention may be applied. That is, the above description is based on tea as a specific example of a beverage that can be customized by combining raw materials, but the scope of the present invention is not limited thereto. may be applied.

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and used by those skilled in the art of computer software. Examples of the computer-readable recording medium include hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform the processing according to the present invention, and may be a set of program instructions and vice versa, and the user application itself for executing them. Specifically, it may be a program itself that can be downloaded through a server or a storage medium and installed on a client computer.

In the above, the present invention has been described with specific matters such as specific components and limited embodiments and drawings, but these are provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. , various modifications and variations can be devised from these descriptions by those of ordinary skill in the art to which the present invention pertains.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and not only the claims described below, but also all modifications equivalently or equivalently to the claims described below belong to the scope of the spirit of the present invention. will do it

In addition, embodiments of the present invention are not mutually exclusive, and the configuration of one embodiment may be applied to other embodiments. Embodiments of the present invention are provided as examples of some of the various forms that can be derived from various combinations of components, and are not limited to specific embodiments of the present invention itself.

110: network
120: user terminal
130: Concierge
200 : database
300: metadata extraction module
400: central control system
500: product recommendation module
600: result data processing module
700: final result display module

Claims (10)

As a method of recommending a beverage suitable for a user,
collecting user information and product information in a database;
extracting metadata by receiving information from the database;
transmitting the extracted metadata to a control unit;
selecting a beverage suitable for the user through deep learning using the metadata as an input value;
recommending the selected beverage to a user;
generating a pattern including information on the beverage selected by the user and information on the user; and
Comprising the step of providing the generated pattern to the user, beverage recommendation method. The method of claim 1,
After the step of recommending the selected beverage to the user, the method further comprising the step of providing an interface for the user to customize the recommended beverage, beverage recommendation method. The method of claim 1,
The pattern is a barcode or QR code, beverage recommendation method. The method of claim 1,
The user information includes at least one of user information, user preference information, concierge history information, user order history information, and teasity information, beverage recommendation method. The method of claim 1,
The step of providing the generated pattern to the user,
Provided as graphic information to the user terminal of the user, or the pattern is provided in a physically printed form, beverage recommendation method. The method of claim 1,
The step of selecting a beverage suitable for the user,
A beverage recommendation method comprising at least one of a user's preference item, a user's non-preference item, and a user's SNS data analysis as the input value. 7. The method of claim 6,
The deep learning includes an input layer including the input value, a hidden layer for re-adjusting weights through comparison with existing metadata through machine learning, and an output layer from which the selected beverage is derived as a result, beverage recommendation method . The method of claim 1,
The step of collecting the user information and product information in the database, the beverage recommendation method for providing an interface for inputting one or more of the user's preference items and the user's non-preference items. 9. The method according to any one of claims 1 to 8,
The beverage is a blending tea (Blending Tea) combining a plurality of tea leaves, beverage recommendation method. 10. The method of claim 9,
The pattern includes information on the type of tea leaves constituting the blending tea, beverage recommendation method.

Images