KR102617327B1 - recommending system for nutritional supplement and nutraceuticals - Google Patents

Abstract

The present invention relates to a health functional food and nutritional supplement recommendation system, and more specifically, to a health functional food and nutritional supplement recommendation system that can increase user satisfaction by recommending the optimal product considering the user's current health status as well as the user's tendencies and preferences. This is about a nutritional supplement recommendation system.
The present invention for achieving the above object consists of a server, a user terminal connected to the server through communication, and a recommendation module provided in the server to recommend nutritional supplements and health functional foods according to the user's request, and the server It is characterized by including the recommendation module, a customer DB in which user information is stored, a product DB in which information on nutritional supplements and health functional foods is stored, and a survey module that conducts a survey through the user terminal.

Description

{recommending system for nutritional supplement and nutraceuticals}

The present invention relates to a health functional food and nutritional supplement recommendation system, and more specifically, to a health functional food and nutritional supplement recommendation system that can increase user satisfaction by recommending the optimal product considering the user's current health status as well as the user's tendencies and preferences. This is about a nutritional supplement recommendation system.

As electronic commerce becomes widely used due to the development of information and communication technology (ICT), the number of cases of purchasing health functional foods through online markets is increasing significantly, moving away from purchasing health functional foods through offline channels such as pharmacies or hospitals. .

Accordingly, a variety of online markets handling health functional foods are emerging, and users can search for and purchase the health functional foods they want using the desired online market among various online markets. For example, users select health functional foods according to the functions they require, such as fatigue recovery, liver health, lung health, etc., or select health functional foods containing vitamins, magnesium, etc., depending on the ingredients the health functional food contains. You can purchase it.

In the case of health functional foods, there is a high tendency for experts to be indiscriminately exposed to various information that amounts to false and exaggerated advertising, and general users have lower medical knowledge than experts such as doctors and pharmacists, so the ingredients and functions of health functional foods are high. It is often difficult to carefully consider the health functional foods that are suitable for oneself.

In the case of a service providing health functional foods through a conventional online market, categories are separated according to function and ingredient or health functional foods are sorted in order of highest rating so that users can more easily select the health functional foods they want. to provide.

However, since the above-mentioned method collects and manages data on ratings and reviews for each company individually, it may not be easy to collect data for some of the companies due to a lack of users, and the evaluation scale is limited. there is.

In the case of another health functional food service, an expert diagnoses each user's symptoms or health condition online, and according to the diagnosis results, each category of health functional food is provided so that individual users can adjust according to their symptoms or health condition. We provide assistance in selecting health functional foods.

However, even in the case of the above-mentioned method, the user must be clearly aware of his or her symptoms or health condition in order to accurately select the health functional food that suits him or her, so there is a high possibility of purchasing a health functional food that does not suit the user.

So, in order to solve this problem, the technology described in Korean Patent No. 10-2308894, as shown in Figures 1 and 2, was proposed, the technical characteristics of which are based on AI (Artificial Intelligence) and provide health functions necessary for users. An operation server 100 that provides food, storing nutritional information, user's medical information, user's location information, user's identification information including the user's face image and voice information, and product information of the health functional food. Big data management department (101); a condition determination unit 103 that determines the user's condition based on at least one of the face image and the voice information; A location determination unit 102 that obtains the user's location information from the user terminal and determines the user's current location; an artificial neural network-based average face determination unit 104 that generates an average face image of the user using a pre-supervised average face image guessing model; and a recommended health functional food determination unit 105 that determines a recommended health functional food to be provided to the user based on at least one of the location information, the medical information, and the condition.

However, the technology described in Korean Patent No. 10-2308894 has the advantage of recommending optimal health functional foods considering the user's health status and current location, but when recommending health functional foods, it is simply There is a problem that there are many cases where the recommended product is not helpful to the user because the recommendation is only made considering the condition, but the user's preference is not considered at all.

Korean Patent No. 10-2308894 (registered on September 28, 2021)

The present invention was created to solve the above-described problems, and the purpose of the present invention is to recommend products through information on a customer group that shows similar tendencies to the user when recommending nutritional supplements and health functional foods at the user's request. By using a combination of a collaborative filtering recommendation method and a content-based recommendation method that recommends products with similar characteristics based on the user's past preferences, the optimal product is recommended by considering the user's current health status as well as the user's inclinations and preferences. The goal is to provide a health functional food and nutritional supplement recommendation system that can increase user satisfaction.

Another purpose of the present invention is to check accurate information on nutritional supplements and health functional foods received through the Ministry of Food and Drug Safety and the public data portal, extract accurate index words for each product, and assign weights according to the frequency of occurrence of each index word. By storing the query index word according to the user's preference in the user information as well, the product can be recommended including the query index word when recommending at the user's request, thereby recommending more accurate products that fit the user's preference. The goal is to provide a recommendation system for health functional foods and nutritional supplements.

The present invention to solve these problems;

It is characterized by consisting of a server, a user terminal connected to the server through communication, and a recommendation module provided in the server to recommend nutritional supplements and health functional foods according to the user's request.

Here, the server includes the recommendation module, a customer DB in which user information is stored, a product DB in which information on nutritional supplements and health functional foods is stored, and a survey module that conducts a survey through the user terminal. It is characterized by

In addition, the recommendation module is a first recommendation module that uses a content-based recommendation method that recommends products with similar characteristics according to the user's past preferences, and a collaboration that recommends products through information on a customer group that shows similar tendencies to the user. It is characterized by consisting of a second recommendation module that uses a filtering method, and a main recommendation module that generates a final recommendation list by assigning weights to products recommended by the first recommendation module and the second recommendation module.

Here, the first recommendation module derives a recommendation list from the products stored in the product DB through the survey information, health checkup information, constitution, health status, interests, health food information of interest, and user search query stored in the customer DB. It is characterized by:

At this time, the server is further equipped with an index word module, and the index word module generates a tendency index word according to the user's tendency through the survey information, health checkup information, constitution, health condition, field of interest, and health food information of interest stored in the customer DB. It is characterized by extracting and storing in the customer DB, and storing in the product DB the product index words that are weighted according to the frequency of occurrence of the index words extracted from the information on nutritional supplements and health functional foods received from the public portal or the Ministry of Food and Drug Safety. Do it as

Meanwhile, the second recommendation module creates a user profile based on gender, age, treatment interest, health food of interest, average purchase amount and number of purchases for products stored in the customer DB, and creates a user profile that standardizes each configuration. It is characterized by storing the user profile in a customer DB, deriving similar customer groups through Pearson, Euclidean, or cosine similarity methods, and recommending nutritional supplements and health functional foods through the information on the derived similar customer groups.

At this time, the weight is applied to the recommended product using Equation 1.

Formula 1: st = a×s1 + (1-a)×s2

(st: final score, s1: score for each recommended product derived through the first recommendation module, s2: score for each recommended product derived through the second recommendation module, a: score derived according to the user's preference)

According to the present invention of the above configuration, when recommending nutritional supplements and health functional foods at the user's request, a collaborative filtering recommendation method that recommends products through information on a customer group showing similar tendencies to the user and the user's past preference type By using a complex recommendation method that recommends products with similar characteristics, the optimal product is recommended considering the user's current health status as well as the user's tendencies and preferences, which has the effect of increasing user satisfaction. there is.

In addition, the present invention verifies accurate information on nutritional supplements and health functional foods received through the Ministry of Food and Drug Safety and the public data portal, extracts accurate index words for each product, assigns weights according to the frequency of occurrence of each index word, and stores the information on the server. By storing the query index according to the user's preference in the user information as well, the product can be recommended including the query index when recommending at the user's request, which has the effect of recommending more accurate products that fit the user's preference. there is.

Figure 1 is a concept of a conventional method of providing customized health functional foods.
Figure 2 is a block diagram of a conventional method of providing customized health functional foods.
Figure 3 is a conceptual diagram of a health functional food and nutritional supplement recommendation system according to the present invention.
Figure 4 is a block diagram of a health functional food and nutritional supplement recommendation system according to the present invention.
Figure 5 is a flow chart of the health functional food and nutritional supplement recommendation system according to the present invention.
Figure 6 is an exemplary diagram of a user profile of the health functional food and nutritional supplement recommendation system according to the present invention.
Figure 7 is a conceptual diagram showing the process of recommending products according to user inquiries in the health functional food and nutritional supplement recommendation system according to the present invention.

Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the attached drawings. The same reference numerals are used for the same components in the drawings, and duplicate descriptions for the same components are omitted. Additionally, it should be understood that the present invention can be implemented in many different forms and is not limited to the described embodiments.

Figure 3 is a conceptual diagram of a health functional food and nutritional supplement recommendation system according to the present invention, Figure 4 is a block diagram of a health functional food and nutritional supplement recommendation system according to the present invention, and Figure 5 is a health functional food and nutritional supplement recommendation system according to the present invention. is a flowchart, Figure 6 is an example of a user profile of the health functional food and nutritional supplement recommendation system according to the present invention, and Figure 7 is a diagram showing a product recommendation system according to the user's inquiry in the health functional food and nutritional supplement recommendation system according to the present invention. This is a conceptual diagram showing the process.

The present invention relates to a health functional food and nutritional supplement recommendation system, and as shown in FIGS. 3 to 7, the composition includes a server 200 provided in a remote location and a user terminal 300 connected to the server 200 through communication. and a recommendation module 230 provided on the server 200 to recommend nutritional supplements and health functional foods according to the user's request.

Therefore, when a user needs a nutritional supplement or health functional food and requests a recommendation through an inquiry through the user terminal 300, the recommendation module 230 takes into account the user's current health status as well as the user's inclination and preference. The optimal product is recommended, and the recommended list is transmitted to the user terminal 300 through a communication network.

In other words, in the past, nutritional supplements or health functional foods were recommended simply considering the user's health condition. However, many nutritional supplements and health functional foods have many products that produce similar effects, and among them, there are products that fit the user's preferences or inclinations, but they are not recommended at all. By not taking this into account, there was a problem that users did not take recommended nutritional supplements or health foods.

However, the present invention not only recommends products with proven effectiveness among products with similar functions and effects, but also recommends the optimal product by considering the user's preferences or inclinations, thereby ensuring that the user can take it consistently, thus protecting the user's health. Not only does it improve performance, but it also increases user satisfaction.

In addition, the server 200 includes the recommendation module 230 that recommends nutritional supplements and health functional foods to the user, a customer DB 210 in which user information is stored, and a product DB in which information on nutritional supplements and health functional foods is stored ( 220) and a survey module 240 that conducts a survey through the user terminal 300.

Here, although not shown in the drawing, the server 200 is connected to the external public portal 410, the Ministry of Food and Drug Safety 420, and the National Health Insurance Service 430 through a communication module provided inside to provide accurate nutritional information. In addition to information on health functional foods, the user's health checkup information is received and stored in the customer DB 210 and product DB 220, so that nutritional supplements or health functional foods suitable for the user's situation can be recommended.

Of course, when accessing the National Health Insurance Corporation (430) and receiving health checkup information for each user, it is natural to obtain consent from each user in advance and receive a certificate before proceeding, and to ensure security accordingly. It is also natural.

Meanwhile, the survey module 240 stores a number of questionnaires written by experts and transmits the questionnaires to the user terminal 300 not only when the user first signs up, but also at set periods and under special circumstances to conduct the survey. do.

Here, when the user signs up, the user's information such as health checkup information, constitution, health status, areas of interest, health food information of interest, etc. is entered, and the user's inclination and preference are confirmed and stored in the customer DB (210). .

In addition, the recommendation module 230 consists of a main recommendation module 232, a first recommendation module 234, and a second recommendation module 236, and the user makes a specific inquiry using the user terminal 300. Upon request, nutritional supplements and health functional foods are recommended.

Here, the first recommendation module 234 uses a content-based recommendation method (VSM, vector space model) to recommend products with similar characteristics according to the user's past preference type, and the second recommendation module 236 A collaborative filtering method (CF model) is used to recommend products through information on customer groups that show similar tendencies to the user.

At this time, the main recommendation module 232 generates a final recommendation list by assigning weight to the products recommended by the first recommendation module 234 and the second recommendation module 236, which will be helpful to the user's health. In addition, it is possible to recommend products that are more suited to the user's inclinations and preferences.

Meanwhile, as shown in FIG. 7, when a user makes a request through a specific inquiry, the first recommendation module 234 provides questionnaire information, health checkup information, constitution, health status, areas of interest, etc. stored in the customer DB 210. A recommended list is derived from the products stored in the product DB 220 through interest health food information and the user's search query.

Here, the server 200 is further equipped with an index module 250. The index module 250 stores survey information, health checkup information, constitution, health status, areas of interest, and health of interest stored in the customer DB 210. A propensity index word according to the user's propensity is extracted through food information, stored in the customer DB 210, and used when searching in the first recommendation module 234. The index word module 250 is a morphological analyzer provided inside. Through (252), prefixes, suffixes, endings, etc. are removed and analyzed based on nouns and adjectives.

At this time, the index word module 250 extracts the index word from the information on nutritional supplements and health functional foods received from the public portal or the Ministry of Food and Drug Safety and enters the product index word weighted according to the frequency of occurrence of the extracted index word in the product DB (220). ) and used when searching in the first recommendation module 234, as described above. The first recommendation module 234 is a VSM that identifies words in the text and expresses the extracted index as a t-dimensional vector. (vector space model) is used.

Therefore, when recommending nutritional supplements and health functional foods, we consider accurate information on nutritional supplements and health functional foods obtained through the public portal (410) and the Ministry of Food and Drug Safety (420), as well as the user's inclinations and preferences to ensure products are suitable for the user. can be recommended.

And, as shown in FIG. 6, the second recommendation module 236 recommends the user through the average purchase amount and number of purchases for gender, age, treatment interest, health food of interest, and product stored in the customer DB 210. A profile is created.

Here, since the units of each component of the user profile are different, they are standardized, and the standardized user profile is stored in the customer DB 210 and used when recommending nutritional supplements and health functional foods.

At this time, if there is a request for recommendation from the user, the second recommendation module 236 derives a similar customer group using the user profile stored in the customer DB 210 using the Pearson, Euclidean, or cosine similarity method, and the derived similar customer group is Through the information, nutritional supplements and health functional foods are recommended.

Meanwhile, as described above, the main recommendation module 232 applies weights to the products recommended by the first recommendation module 234 and the second recommendation module 236, from the order of each product to the set ranking. A final recommendation list of products is created, and weights are applied through [Formula 1].

[Formula 1]

st = a×s1 + (1-a)×s2

(st:final score,

s1: Score for each recommended product derived through the first recommendation module,

s2: Score for each recommended product derived through the second recommendation module,

a:Score derived according to user’s preference)

Here, when the first recommendation module 234 and the second recommendation module 236 recommend each product, there is a score derived for recommendation for each product, and the parameters used in Equation 1 are the scores of these products. will be utilized.

At this time, the products recommended by the first recommendation module 234 and the second recommendation module 236 may be different, but the same product may be recommended, so each product may have both s1 and s2, or only one score. Since there may be both, the recommendation score is higher when both sides are present, so there is a higher chance of being included in the final list.

Meanwhile, the parameter a is a score derived according to the user's personality, and depending on the user's personality, the content-based recommendation method of the first recommendation module 234 may be appropriate, or the collaborative filtering recommendation method of the second recommendation module 236 may be appropriate. The method may be right.

Therefore, the parameter a is derived from the user information stored in the customer DB 210 through Bayesian optimization to reflect the user's tendency as much as possible.

In addition, in the present invention, although not shown in the drawing, information from the Health Insurance Review and Assessment Service can be received with the user's consent, and the user can check the information on the recently prescribed medicine through the information received from the Health Insurance Review and Assessment Service. This information can be stored in the customer DB (210).

Here, when the user requests a recommendation for nutritional supplements or health functional foods, the main recommendation module 232 can check the medicines the user has recently taken through the information stored in the customer DB 210, and add them to the recommendation list. If you have taken a medicine containing ingredients that conflict with existing nutritional supplements or health functional foods, you can check whether you are currently taking it through the user terminal 300.

At this time, if you are not taking the drug in question, the existing recommendation list is transmitted to the user terminal 300, and if you are taking the drug in question, a recommended list of nutritional supplements or health functional foods containing ingredients that conflict with the ingredients of the drug in question are sent to the user terminal 300. to be excluded from.

To explain the overall flow of the health functional food and nutritional supplement recommendation system of the present invention described above, when a user requests a product recommendation while inputting a query into the user terminal 300, the server 200 receives the request and Recommended products are derived through the content-based recommendation method of the first recommendation module 234 and the collaborative filtering recommendation method of the second recommendation module 236.

Here, the main recommendation module 232 creates a final recommendation list by assigning weights to the products derived from the first and second recommendation modules 234 and 236, and the final recommendation list thus generated is sent to the user terminal 300. transmitted to satisfy the user's needs.

Although the preferred embodiments of the present invention have been described above, the scope of the rights of the present invention is not limited thereto, and the scope of the rights of the present invention extends to the scope substantially equivalent to the embodiments of the present invention. Various modifications can be made by those skilled in the art without departing from the above.

The present invention relates to a health functional food and nutritional supplement recommendation system, and more specifically, to a health functional food and nutritional supplement recommendation system that can increase user satisfaction by recommending the optimal product considering the user's current health status as well as the user's tendencies and preferences. This is about a nutritional supplement recommendation system.

200: Server 210: Customer DB
220: Product DB 230: Recommended module
232: Main recommendation module 234: First recommendation module
236: Second recommendation module 240: Survey module
250: Index word module 300: User terminal
410: Public portal 420: Ministry of Food and Drug Safety
430: National Health Insurance Corporation

Claims (7)

It consists of a server, a user terminal connected to the server through communication, and a recommendation module provided in the server to recommend nutritional supplements and health functional foods according to the user's request,
The server conducts a survey through the recommendation module, a customer DB in which user information is stored including information on medications recently prescribed by the user, a product DB in which information on nutritional supplements and health functional foods is stored, and the user terminal. Includes a survey module and an index module,
The recommendation module is a first recommendation module that uses a content-based recommendation method that recommends products with similar characteristics according to the user's past preferences, and a collaborative filtering method that recommends products through information on a customer group that shows similar tendencies to the user. It consists of a second recommendation module that uses and a main recommendation module that generates a final recommendation list by assigning weights to the products recommended by the first recommendation module and the second recommendation module,
The index module extracts a tendency index word according to the user's tendency through survey information, health checkup information, constitution, health status, interest area, and health food information of interest stored in the customer DB and stores it in the customer DB,
Index words extracted from information on nutritional supplements and health functional foods received from the public portal or the Ministry of Food and Drug Safety are weighted according to the frequency of occurrence of the index words, and the product index words are stored in the product DB.
The main recommendation module determines whether the user is taking a drug that contains ingredients that conflict with the nutritional supplements or health functional foods in the final recommendation list based on the information on the drug recently prescribed by the user stored in the customer DB, and recommends the drug. If it is confirmed that the user is not taking the drug, the existing final recommendation list is sent to the user terminal. If it is confirmed that the user is taking the drug, nutritional supplements or health functional foods containing ingredients that conflict with the drug's ingredients are excluded from the final recommendation list. A health functional food and nutritional supplement recommendation system characterized by:
delete delete According to paragraph 1,
The first recommendation module derives a recommendation list from the products stored in the product DB through the questionnaire information, health checkup information, constitution, health status, interests, health food information of interest, and user search query stored in the customer DB. Features a health functional food and nutritional supplement recommendation system.
delete According to paragraph 1,
The second recommendation module creates a user profile through the average purchase amount and number of purchases for the gender, age, treatment interest, health food of interest, and product stored in the customer DB, and creates a user profile that standardizes each configuration in the customer DB. Save it to
Similar customer groups are derived from the user profile using Pearson, Euclidean, or cosine similarity methods,
A health functional food and nutritional supplement recommendation system characterized by recommending nutritional supplements and health functional foods through derived information on similar customer groups.
According to paragraph 1,
A health functional food and nutritional supplement recommendation system characterized by applying the above weights to recommended products through Formula 1.
Formula 1: st = a×s1 + (1-a)×s2
(st: final score, s1: score for each recommended product derived through the first recommendation module, s2: score for each recommended product derived through the second recommendation module, a: score derived according to the user's inclination)

Images