Foreign user-customized brand matching method

Abstract

The present invention may provide information, which may be accepted by a foreign user, in a customized manner, in consideration of the position of the foreign user by calculating and storing an SD value in advance by applying a Rasch model on the basis of brand data that is received as an input and foreign user data, and then showing a brand satisfying (1 - SDb value) ≥ SDu value to the foreign user when requested. Also, the feeling from the position of the foreign user may be reflected by reflecting, in the SDb value of the corresponding brand, data regarding an activity carried out by the foreign user with respect to the brand, and accordingly, information may be provided in a customized manner. In addition, when reflecting in the SD value, by applying a KNN model, K nearest neighbors may be found and predicted as new data so as to match between the foreign user and the brand in a more customized manner.

Description

A customized brand matching method for foreign users

[0001] The present invention relates to a brand matching method tailored to foreign users, and more specifically, to an SD obtained by applying a Rasch model to user data directly stored by foreign users. u Based on the value, SD of several brands pre-stored in the relay server in advance b Among the values, (1- SD b value)≥ SD u By configuring the recommendation system to recommend brands that satisfy the value, we can recommend a range of brands that are acceptable to foreign users, enabling us to provide customized brand recommendations to foreign users, unlike brands that are indiscriminately exposed. In addition, as user data changes, such as changes in foreign users' travel schedules or companions, we can reflect this in the SD. u' By recommending brands through value, we can match you with the optimal customized brand. And the brand's SD b When calculating the value, it is possible to provide customized recommendations for brands seen from the perspective of foreign users by reflecting the values ​​applied to the Rasch Model, such as activity data performed by foreign users through the application, such as frequency of service access, number of posts viewed within the service, number of brand searches within the service, number of brand views within the service, number of brand saves within the service, number of brand voucher saves within the service, and number of brand voucher uses within the service.

[0002] Recently, with the advancement of media like the Internet, travelers abroad often utilize apps to obtain information about the country they wish to visit. Patent Documents 1 and 3 below describe technologies that provide diverse tourism information to foreigners.

[0003]

[0004] (Patent Document 1) Korean Patent No. 10-1928076

[0005] The present invention relates to a method for providing foreign tourist location information content using big data, comprising a step of generating multilingual tourism statistical location content, wherein the step of generating multilingual tourism statistical location content includes a step of searching for foreign statistical information matching multilingual content input from an administrator terminal, and generating statistical meta information if there is matching foreign statistical information, a step of checking whether there is surrounding location information matching the input multilingual content, and a step of generating surrounding location information if there is location information, and a step of linking and processing the statistical meta information and surrounding location information.

[0006]

[0007] (Patent Document 2) Korean Patent Publication No. 10-2021-0130298

[0008] A system for providing shopping promotion information for foreign tourists is characterized by including: a promotion server that receives promotion information including period information, item information, and price information from multiple shopping malls and stores the received information in a promotion database; an information input module that receives search information including travel schedule information from a user; an information comparison module that filters the promotion information by comparing the search information with the promotion information to generate personalized shopping information; and a display module that displays the personalized shopping information to the user.

[0009]

[0010] (Patent Document 3) Korean Patent No. 10-2225704

[0011] A tourism service provision system based on linking domestic residents and Korean nationals is provided, and includes a tourism service provision server including at least one foreign terminal for uploading domestic tourism product data in order for domestic residents to sell Korean tourism products to foreign customers of their own country, at least one professional terminal for uploading professional data including information on regions, foreign languages ​​spoken, types of qualifications held, concepts of tourism products, and payment information, a user terminal for selecting and making payment for at least one tourism product uploaded by the at least one foreign terminal and a professional terminal based on information uploaded by the at least one professional terminal, a product management unit for uploading and storing domestic tourism product data from the at least one foreign terminal, a human resources management unit for receiving professional data from the at least one professional terminal and managing a human resources pool, and an allocation unit for transmitting and allocating tourism product and user information to the professional terminal of the selected professional terminal when payment is made for the tourism product and professional personnel selected by the user terminal.

[0012] [Prior Art Literature]

[0013] [Patent Document]

[0014] (Patent Document 1) Korean Patent No. 10-1928076 (Registration Date: December 5, 2018)

[0015] (Patent Document 2) Korean Patent Publication No. 10-2021-0130298 (Published: November 1, 2021)

[0016] (Patent Document 3) Korean Patent No. 10-2225704 (Registration Date: March 3, 2021)

[0017] However, the services provided to these existing foreign users have the following problems.

[0018] (1) From the perspective of Koreans, content, restaurants, interests and activities that foreigners may find interesting or beneficial are provided, but from the perspective of foreigners, it is not easy to stimulate interest because the content does not match the tastes, consumption patterns and preferences of foreigners.

[0019] (2) In addition, most of the content provided in this way is simply selected to target what is frequently visited by foreigners, so there are limitations in recommending content.

[0020] (3) In particular, the content selected by foreign users should vary depending on the time of travel, purpose, and travel companions, but existing services only recommend pre-determined content, which limits the content that foreign users can select.

[0021] (4) On the other hand, information obtained through SNS, etc. has the advantage of being easily obtainable, but it may be provided as biased information as it does not reflect the tastes, consumption patterns, and preferences of foreign users.

[0022] (5) Since information about items, food, and animals that are restricted depending on the country and religion is provided without filtering, there is a risk that inaccurate information will be provided to foreign users, and this also acts as a factor that greatly reduces interest in foreign users.

[0023]

[0024] The present invention has taken these points into consideration, and applies the Rasch Model based on the input brand data to pre-determine SD. b The values ​​are calculated and saved, and the data directly entered by foreign users is applied to the Rasch Model to create SD. u After calculating and saving the value, when there is a request from a foreign user (1- SD b value)≥ SD uThe purpose is to provide a customized brand matching method for foreign users that can provide customized brand data that is acceptable to foreign users by considering the position of foreign users by configuring it to show brand data that satisfies the value to foreign users.

[0025] In addition, the present invention detects activity data such as frequency of service access by foreign users, number of posts viewed within the service, number of brand searches within the service, number of brand views within the service, number of brand saves within the service, number of brand voucher saves within the service, and number of brand voucher uses within the service in numerical form, and compares the numerical value (frequency) with the SD of the corresponding brand. b Another purpose is to provide a customized brand matching method for foreign users that allows them to reflect in their brand selection what they are interested in, have experienced, or have visited, thereby providing more customized data to foreign users.

[0026] In addition, the present invention provides a method for matching foreign users with customized brands, which enables more customized matching between foreign users and brands by applying a KNN (K Nearest Neighbor) model in which data is classified through a definition of surrounding neighbors (k) when reflecting on such SD values, thereby finding K nearest neighbors around a given data point when classifying data or solving a regression problem, and predicting new data through majority vote or average. Another purpose of the present invention is to provide a method for matching foreign users with customized brands, which enables more customized matching between foreign users and brands.

[0027] In order to achieve this purpose, the foreign user customized brand matching method according to the present invention is a matching method in which an application stored in a relay server is downloaded to the user's mobile terminal and used; brand data is input for a brand that is to open an online store, and the Rasch Model is applied to each input brand data to create an SD b Step 1: Calculate the value and store it in the relay server; When there is a foreign user who wants to sign up, enter the user data of the foreign user, and apply the Rasch Model to each entered user data to obtain the SD u The second step is to calculate the value and store it in the relay server; and when a foreign user requests a recommendation or runs the application, it checks whether the foreign user is registered. If not, the second step is performed first, and if the foreign user is registered, (1-SD) b value)≥ SD u It is characterized by including a third step of showing a brand that satisfies the value as an application;

[0028] In particular, the brand data is characterized by including industry, region (accessibility), items handled (menu), price range, and voucher.

[0029] In addition, the user data is characterized by including nationality, gender, age, language, interests, domestic visit schedule, and visiting companion.

[0030] Meanwhile, the present invention applies the Rasch Model to user data to re-enter SD when a registered foreign user changes or adds new personal data input items of his / her choice. u It is characterized by including a fourth step of calculating the value and storing it in a relay server.

[0031] And the present invention, when the activity data including the frequency of service access, the number of posts viewed within the service, the number of brand searches within the service, the number of brand views within the service, the number of brand saves within the service, the number of brand voucher saves within the service, and the number of brand voucher uses within the service are changed, the activity data is added to the brand data of the brand and then the Rasch Model is applied to re-SD b It is characterized by including a fifth step of calculating and applying the value.

[0032] Lastly, the present invention provides the SD of the foreign user u Value and SD of the above brands b When determining the value, it is characterized by being classified through the K Nearest Neighbor (KNN) model, in which data is classified through the definition of the surrounding neighbors (k).

[0033] The method for matching foreign users with customized brands according to the present invention has the following effects.

[0034] (1) Customized travel information tailored to foreign users can be provided by providing customized brand data based on data including schedules and companions provided by foreign users.

[0035] (2) At this time, foreign user data and brand data are each SD based on the Rasch Model. b Values ​​and SD u Calculate the values ​​and these (1- SD b value)≥ SD u By providing brand data that satisfies the value, you can further enhance the customized effect by allowing foreign users to evaluate and measure each brand relatively and then view brands that are deemed to be tolerable to foreign users.

[0036] (3) In addition, by applying activity data based on the frequency of service access, number of post views within the service, number of brand searches within the service, number of brand views within the service, number of brand saves within the service, number of brand voucher saves within the service, and number of brand voucher uses within the service obtained by foreign users through the application according to the present invention to the brand in question and recalculating and applying the SD value, the brand evaluation from the foreigner's perspective can be reflected, thereby further enhancing the customized effect.

[0037] (4) And as foreign users change their data, the SD value is recalculated to reflect the changed data, so that a customized brand can be recommended based on the reflected data.

[0038]

[0039] [Figure 1] is a block diagram for explaining a foreign user customized brand matching method according to the present invention.

[0040] [Figure 2] is a time flow chart for explaining a method of matching a customized brand to a foreign user according to the present invention.

[0041] [Figure 3] shows the SD of foreign users calculated according to the present invention. u Value and Brand SD b Here is an example showing how to apply values.

[0042] [Figure 4] is a time flow diagram for explaining a method of matching a customized brand by reflecting the activity data of a foreign user according to the present invention.

[0043] [Figure 5] is a time flow diagram for explaining a method of matching a customized brand by reflecting changed user data of a foreign user according to the present invention.

[0044] Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the attached drawings. Prior to this, it should be noted that the terms and words used in this specification and claims should not be construed as limited to their conventional or dictionary meanings. Rather, they should be interpreted in a way that is consistent with the technical spirit of the present invention, based on the principle that the inventor can appropriately define the concepts of terms to best explain his or her invention.

[0045] Therefore, the embodiments described in this specification and the configurations illustrated in the drawings are only the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention. Therefore, it should be understood that there may be various equivalents and modifications that can replace them at the time of filing this application.

[0046]

[0047] [Brand Matching Methods Tailored to Foreign Users]

[0048] The foreign user customized brand matching method according to the present invention receives brand data of a brand that is to be recommended to a foreigner in advance as shown in [Fig. 1] to [Fig. 5] and performs SD using the Rasch Model on the input data. b The values ​​are calculated and stored, and the SD is applied through the Rasch Model to the data directly entered by foreign users. u After calculating and storing the value, when a request is made by a foreign user or an application is executed, (1- SD b value)≥ SD u By showing brands that satisfy the value to foreign users, the SD obtained as a relative value among all registered foreign users u Among all the brands subscribed based on value, we recommend brands that are relatively more tolerable to foreign users, allowing us to recommend the optimal customized brand.

[0049] At this time, when activity data that occurs when a foreign user uses the application, such as frequency of service access, number of posts viewed within the service, number of brand searches within the service, number of brand views within the service, number of brands saved within the service, number of brand vouchers saved within the service, and number of brand vouchers used within the service, the frequency of this frequency data is applied to the Rasch Model for the brand in question and then SD is generated again. b' By calculating the value, we can recommend brands in a customized way that reflects the needs of foreign users by reflecting what they have experienced or gone through directly.

[0050] In addition, when a foreign user changes his / her user data, such as a change in schedule or a change in companion, the Rasch Model is applied to the changed data and SD is recalculated. u' By calculating the value, we can directly reflect the SD value for foreign users and brands, enabling more customized brand recommendations.

[0051] Finally, the SD value obtained in this way is used to determine the SD of foreign users. u SD of value and brand b When determining the value, the data is classified through the KNN (K Nearest Neighbor) model, which defines the surrounding neighbors (k), thereby enabling foreign users to receive recommendations for customized brands that are more suitable for them.

[0052]

[0053] This configuration is described in more detail below with reference to the attached drawings. The method for matching foreign users with customized brands according to the present invention comprises three steps, and each step will be described here. Furthermore, the present invention can add two additional steps beyond these steps, and these additional steps will be described below.

[0054] A. Step 1

[0055] The first step is to input and store brand data to be provided in a customized manner to foreign users, as shown in [Figure 1] to [Figure 3], and to prepare to provide brand data in a customized manner according to the request of foreign users, which will be described later.

[0056]

[0057] In this first stage, as shown in [Figure 1], for categories expected to be visited by many foreign users, we divide them into each category and have several brands suitable for each category open online, and input brand data for the opened brands and send it to SD. b The values ​​are calculated and stored. Here, the categories are categorized into, for example, restaurants, entertainment, shopping, and accommodations, but are not limited to these.

[0058]

[0059] In addition, the brand data includes data on industry, region (accessibility), items handled (menu), price range, and vouchers, and this brand data is input by the brand side. SD is applied to all brands that are entered. b By calculating and storing it on the relay server, each brand has its own unique SD b Assign a value and SD of foreign users as described later u It will be possible to display brands that meet certain conditions in a customized manner so that foreign users can see them by comparing them to their values.

[0060]

[0061] At this time, the brand data is divided into several groups by input data item and a predetermined value is assigned to each brand in each group, or the values ​​are determined by ordering each item in order of popularity or popularity among foreign users, or an appropriate SD value is determined in advance for each brand through standard distribution by item.

[0062]

[0063] Meanwhile, the above SD b The value is obtained as a numerical value by applying the numerical (frequency) value determined for each brand data to the Rasch model. At this time, the Rasch model is a conventional technique for estimating the subject's ability using only one variable. In the present invention, it refers to the value obtained sequentially for all brands by using the selected value for each brand data as a variable.

[0064]

[0065] SD obtained in this way b The values ​​are calculated for each brand, and the calculated SD b The value is stored on the relay server and the SD of the foreign user described later u It is used to extract appropriately matched brands through the relationship formula.

[0066]

[0067] Step 2

[0068] Step 2 is, as shown in [Figure 1] to [Figure 3], when a foreign user tries to register as a foreign user through an application downloaded to a mobile terminal, the user data required for registration is entered, and the entered user data is stored in an SD card. b This is the step of calculating the value and storing it in the relay server.

[0069]

[0070] At this time, it is desirable to structure the user data entered by the foreign user to include data necessary to distinguish each foreign user and data necessary to match them with brands. Data necessary for distinguishing foreign users, such as contact information, country, and name, sufficient to distinguish one foreign user from another, is sufficient. Data necessary for brand matching, for example, includes nationality, gender, age, language, interests, domestic travel schedule, and visiting companions. This allows for customized brand offers tailored to each travel companion and schedule.

[0071]

[0072] Each of these user data is quantified and the Rasch Model is applied as described above to determine the SD. u The values ​​are calculated and stored on the relay server. At this time, the numerical value of each user data can be divided into several groups in order of the countries that visit Korea the most by nationality, for example, and a set numerical value (frequency) can be assigned. In addition, the visit schedule is assigned a numerical value (frequency) according to the length of the visit, such as whether it is a one-day, two-day, three-four-day, or a week-long schedule. The visiting companion is assigned a numerical value (frequency) so that the appropriate brand can be selected depending on whether the visiting companion is with a friend, parent, or family. Gender and age are used to distinguish appropriate brands that are likely to be frequently visited according to the places or activities that the user wants to see or do, and these values ​​can be used as frequencies as they are.

[0073]

[0074] Step 3

[0075] The third step, as shown in [Figure 1] to [Figure 3], is the step of showing recommended brands to foreign users when they request or run the application. The recommended brands at this time are the SD of the foreign user. u SD of each brand stored on the relay server based on the value b By comparing the values, the brand that satisfies [Mathematical Formula 1] below is shown.

[0076]

[0077]

[0078] At this time, the recommended brands can show all brands that satisfy the above formula, but the number may be too large, so it is desirable to configure it so that foreign users can select the number to show for their convenience, or to limit the number to a fixed number, for example, the top 5 to 10.

[0079]

[0080] That is, as shown in [Figure 3], target A, a foreign user, uses the SD according to the present invention. u Target B, a foreign user with a value of 20, is an SD according to the present invention. u When the value is 60, target A is SD according to the above [Mathematical Formula 1] b It is recommended that the value be less than 80, and target B is SD b You will be recommended brands with a value of 20 or less. In this case, Target A is SD b Because the number of brands with values ​​below 80 can be overwhelming, we've limited the number of brands displayed to 5 to 10. Of course, if foreign users want to see more, we can show more.

[0081]

[0082] Step 4

[0083] Step 4 is, as shown in [Figure 4], when a registered foreign user modifies his / her user data, the modified data is reflected and re-uploaded to SD. u' This is the step that calculates and stores the value.

[0084]

[0085] This is a step to change brand recommendations accordingly as user data changes, such as when a foreign user changes with accompanying people or schedules. To this end, when a foreign user requests a change in user data, the changed user data is input, and the Rasch Model is applied to the foreign user based on the input data, and then SD is re-evaluated. u' Calculate the value. And when there is a brand recommendation through [Example 1] described above, this SD u' As you apply values, you can customize your brand recommendations based on the changed data.

[0086]

[0087] Step 5

[0088] Step 5, as shown in [Figure 5], is a step to enable foreign users to recommend brands that have been evaluated from a foreigner's perspective when making recommendations to other foreign users by reflecting the evaluations made by registered foreign users regarding the recommended brands.

[0089]

[0090] To this end, the activity data of foreign users, i.e. frequency of service access, number of posts viewed within the service, number of brand searches within the service, number of brand views within the service, number of brands saved within the service, number of brand vouchers saved within the service, and number of brand vouchers used within the service, are quantified whenever they change, and the numbers obtained in this way are applied to the Rasch Model described above and re-SD. b'By calculating and storing the values, you can recommend the brands that foreign users search for the most, visit the most, or save and use the most vouchers for to other foreign users.

[0091]

[0092] As described above, the present invention quantifies user data provided by foreign users and brand data provided by a brand to be recommended to foreign users, and applies these values ​​to the Rasch Model to generate SD. b Values ​​and SD u Get the value and based on this value (1- SD b value)≥ SD u By recommending brands that satisfy the value to foreign users, we can provide customized brands that match the foreigner's travel purpose, travel companion, and travel period.

[0093]

[0094] Meanwhile, the present invention provides SD obtained in this manner b Values ​​and SD u When determining the value, by using the KNN (K Nearest Neighbor) model, which classifies data through the definition of surrounding neighbors (k), the data to be evaluated based on the number of neighbors (K) in the given data can be easily classified numerically and used by assigning the most numerous class to the class of the corresponding data point. Therefore, using user data, SD u When determining values ​​or using brand data, SD b It allows for easy and convenient classification and matching when determining values.

Claims

1. In a matching method that allows users to download and use applications stored on a relay server to their mobile devices, Enter brand data for brands that are planning to open an online store, and apply the Rasch Model to each entered brand data to determine SD. b Step 1: Calculating the value and storing it in the relay server; When there is a foreign user who wants to sign up, the user data of the foreign user is entered, and the Rasch Model is applied to each entered user data to determine the SD. u Step 2: Calculating the values ​​and storing them in the relay server; When a foreign user requests a recommendation or runs an application, check whether the foreign user is registered. If not, perform the second step above first. If the foreign user is registered, (1-SD) b value)≥ SD u A method for matching brands tailored to foreign users, characterized by including a third step of showing brands that satisfy the value as an application; 2. In paragraph 1, The above brand data is, A brand matching method tailored to foreign users, featuring industry, region (accessibility), products handled (menu), price range, and vouchers.

3. In paragraph 1, The above user data is, A method for matching brands tailored to foreign users, characterized by including nationality, gender, age, language, interests, domestic visit schedule, and visiting companions.

4. In paragraph 1, Registered foreign users, As you change or add new personal data entries that you have chosen, the user data is re-SD-processed by applying the Rasch Model. uA method for matching foreign users with customized brands, characterized by including a fourth step of calculating a value and storing it in a relay server.

5. In paragraph 1, The registered foreign user, When activity data including service access frequency, number of posts viewed within the service, number of brand searches within the service, number of brand views within the service, number of brand saves within the service, number of brand voucher saves within the service, and number of brand voucher uses within the service change, the activity data is added to the brand data of the brand in question and then the Rasch Model is applied to re-SD. b A method for matching foreign users' customized brands, characterized by including a fifth step of calculating and applying a value.

6. In any one of paragraphs 1 to 5, SD of the above foreign user u Value and SD of the above brands b When determining the value, A foreign user-tailored brand matching method characterized by classification through the KNN (K Nearest Neighbor) model in which data is classified through the definition of surrounding neighbors (k).

Images