Method and system for recommending personalized nutritional products

Abstract

The present invention relates to a method and a system for recommending personalized nutritional products and, more specifically, a method and a system for recommending personalized nutritional products, the method comprising the steps of: receiving a selection of at least one first nutritional product; receiving information about the selected first nutritional product; searching for a combination of nutritional products on the basis of optimization criteria for the received first nutritional product; and providing the combination of nutritional products to a user terminal, wherein the searching for the combination of nutritional products involves grouping, in a database storing information about a plurality of nutritional products, one or more second nutritional products on the basis of the optimization criteria, thereby enhancing convenience of product recommendation on the basis of a user's consumption tendencies, and optimizing and providing recommendations.

Description

Personalized nutritional product recommendation method and system

[0001] The present invention relates to a personalized nutritional product recommendation method and system thereof, and more particularly, to a personalized nutritional product recommendation method and system thereof, comprising: a step of providing a selection for at least one first nutritional product; a step of receiving information about the selected first nutritional product; a step of searching for a nutritional product combination according to an optimization criterion for the received first nutritional product; and a step of providing the nutritional product combination to a user terminal, wherein the searching for the nutritional product combination groups at least one second nutritional product according to the optimization criterion in a database in which information about a plurality of nutritional products is stored, thereby increasing the convenience of product recommendation and providing it in an optimized manner according to the user's consumption tendency.

[0002] Health functional foods are foods manufactured using raw materials or ingredients that have beneficial functions for the human body. Nutritional supplements are products formulated with various nutrients, made into tablets or beverages for easy consumption and absorption. With growing interest in health, interest in nutritional products like nutritional supplements and health functional foods is also on the rise. However, online services providing nutritional product information are underdeveloped, making it difficult for consumers to accurately understand and compare nutritional product information.

[0003] There are attempts to provide consumers with information about nutritional products online. However, these services often require users to answer a series of questions to determine health-related information, such as medical conditions or age. This often prevents users from purchasing nutritional products, and often results in recommendations that don't meet their criteria. In other words, conventional nutritional product information services require users to complete questionnaires or collect data, such as medical or prescription records, based on which nutritional product recommendations are made, leading to an excessive amount of input.

[0004] In addition, when a user is taking a nutritional product or wants to purchase a nutritional product, it is impossible to recommend a more effective nutritional product, and there is the inconvenience of having to compare several nutritional products and select a nutritional product that suits the user's taste.

[0005] The technology underlying the present invention is disclosed in Korean Patent Publication No. 10-2022-0063877.

[0006] The present invention is intended to solve the problems of the above-mentioned prior art, and can be performed by an electronic device including at least one processor, and includes the steps of: providing a selection for at least one first nutritional product; receiving information on the selected first nutritional product; searching for a combination of nutritional products according to an optimization criterion for the received first nutritional product; and providing the combination of nutritional products to a user terminal, wherein the searching for the combination of nutritional products is performed by grouping at least one second nutritional product according to the optimization criterion in a database in which information on a plurality of nutritional products is stored, thereby increasing the convenience of product recommendation according to the user's consumption tendency and providing a nutritional product recommendation method and system capable of optimizing and recommending nutritional products.

[0007] In addition, the information on the first nutritional product of the present invention includes at least the nutritional components of the first nutritional product, and the search for the combination of nutritional products is performed by grouping at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided by comparing the nutritional component content of the first nutritional product, thereby providing a nutritional product recommendation method and system thereof that enable the user to consume nutritional components corresponding to the selected nutritional product when optimizing the combination of nutritional products.

[0008] In addition, the present invention further includes, prior to the searching step, a step of analyzing an optimization criterion for the nutritional product based on the usage history information of the user terminal, wherein the optimization criterion includes at least one filter criterion for filtering the nutritional product on the database, and in the searching step, the second nutritional product is grouped to correspond to the at least one analyzed filter criterion, thereby providing a personalized nutritional product recommendation method and system thereof.

[0009] In addition, the analysis of the usage history information of the user terminal of the present invention is performed based on the interaction data for the nutritional product and the response of the user terminal to the content provided to the user terminal, and the filter criteria are determined from a pool of criteria related to medical branch processing and preferences for nutritional product intake methods, thereby providing a nutritional product recommendation method and system capable of filtering and combining nutritional products differently according to the user's usage history.

[0010] In addition, the present invention provides a nutritional product recommendation method and system capable of determining whether an exemplary nutritional product satisfies the provision criteria, including a step of recommending the grouped exemplary nutritional product through a prediction model learned to group at least one nutritional product according to the optimization criteria on the database, and a step of approving the exemplary nutritional product based on preset approval criteria including medical branch processing.

[0011] In addition, the present invention includes a step of providing a plurality of filter criteria to the user terminal when the optimization criteria include a plurality of filter criteria, and the step of providing the filter criteria provides the plurality of filter criteria to the user terminal in a sorted order based on analysis of the usage history information of the user terminal, thereby providing the user with various choices.

[0012] In addition, the present invention further includes, prior to the searching step, a step of extracting at least one keyword from an input provided from a user terminal, and, if the keyword is determined to be a keyword related to a nutritional ingredient, grouping the second nutritional product so as to satisfy the content of the nutritional ingredient related to the keyword, and, if the keyword is determined to be a keyword related to an optimization criterion, grouping the second nutritional product according to the optimization criterion related to the keyword, thereby providing an optimized combination even with a simple search together with the usage history information of the user terminal.

[0013] However, the technical problems to be solved by the embodiments of the present invention are not limited to the technical problems described above, and other technical problems may exist.

[0014] As a technical means for achieving the above-described technical task, a personalized nutritional product recommendation method according to one embodiment of the present invention can be performed by an electronic device including at least one processor, and includes a step of providing a selection for at least one first nutritional product, a step of receiving information about the selected first nutritional product, a step of searching for a nutritional product combination according to an optimization criterion for the received first nutritional product, and a step of providing the nutritional product combination to a user terminal, wherein the searching for the nutritional product combination can group at least one or more second nutritional products according to the optimization criterion on a database in which information about a plurality of nutritional products is stored.

[0015] Additionally, according to one embodiment of the present invention, the information about the first nutritional product includes at least the nutritional components of the first nutritional product, and the search for the nutritional product combination may be to group at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided, compared to the nutritional component content of the first nutritional product.

[0016] In addition, according to one embodiment of the present invention, prior to the searching step, the method further includes a step of analyzing an optimization criterion for the nutritional product based on the usage history information of the user terminal, wherein the optimization criterion includes at least one filter criterion for filtering nutritional products on the database, and in the searching step, the second nutritional product can be grouped to correspond to the at least one analyzed filter criterion.

[0017] In addition, according to one embodiment of the present invention, the analysis of the usage history information of the user terminal is performed based on interaction data for the nutritional product and the user terminal's response to content provided to the user terminal, and the filter criteria may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods.

[0018] In addition, according to one embodiment of the present invention, the searching step may include a step of recommending the grouped example nutritional products through a predictive model learned to group at least one nutritional product according to the optimization criteria on the database and provide the example nutritional products, and a step of approving the example nutritional products based on preset approval criteria including medical branch processing.

[0019] In addition, according to one embodiment of the present invention, when the optimization criteria include a plurality of filter criteria, the step of providing a plurality of the filter criteria to the user terminal may include providing the plurality of the filter criteria to the user terminal in a sorted order based on analysis of the usage history information of the user terminal.

[0020] In addition, according to one embodiment of the present invention, before the searching step, the method further includes a step of extracting at least one keyword from an input provided from a user terminal, and when the keyword is determined to be a keyword related to a nutritional ingredient, the second nutritional product may be grouped so as to satisfy the content of the nutritional ingredient related to the keyword, and when the keyword is determined to be a keyword related to an optimization criterion, the second nutritional product may be grouped according to an optimization criterion related to the keyword.

[0021] In addition, a nutritional product recommendation device according to one embodiment of the present invention includes a communication unit that provides a second nutritional product grouped from a selection of at least one first nutritional product to a user terminal, and a search unit that searches for a combination of nutritional products according to an optimization criterion for the first nutritional product, wherein the search for the combination of nutritional products may be a process of grouping at least one second nutritional product according to the optimization criterion in a database in which information on a plurality of nutritional products is stored.

[0022] In addition, according to one embodiment of the present invention, the method further includes a criterion analysis unit that derives an optimization criterion for the nutritional product based on the usage history information of the user terminal, wherein the optimization criterion includes at least one filter criterion for filtering nutritional products on the database, and the search unit groups the second nutritional product to correspond to the analyzed at least one filter criterion, and the analysis of the usage history information of the user terminal may be performed based on interaction data for the nutritional product and a response of the user terminal to content provided to the user terminal.

[0023] In addition, a system for providing a personalized nutritional product recommendation service through a service platform according to one embodiment of the present invention includes a user terminal and a nutritional product recommendation device that communicates with the user terminal, wherein the nutritional product recommendation device includes a communication unit that provides the user terminal with a second nutritional product grouped from a selection of at least one first nutritional product, and a search unit that searches for a combination of nutritional products according to an optimization criterion for the first nutritional product, and the search for the combination of nutritional products may be a process of grouping at least one or more second nutritional products according to the optimization criterion on a database in which information on a plurality of nutritional products is stored.

[0024] The above-described problem-solving methods are merely exemplary and should not be construed as limiting the present invention. In addition to the exemplary embodiments described above, additional embodiments may be included in the drawings and detailed description of the invention.

[0025] According to the problem solving means of the present invention described above, the present invention can be performed by an electronic device including at least one processor, and includes a step of providing a selection for at least one first nutritional product, a step of receiving information about the selected first nutritional product, a step of searching for a combination of nutritional products according to an optimization criterion for the received first nutritional product, and a step of providing the combination of nutritional products to a user terminal, wherein the searching for the combination of nutritional products can increase the convenience of product recommendation and optimize and recommend nutritional products according to the user's consumption tendency by grouping at least one or more second nutritional products according to the optimization criterion on a database in which information about a plurality of nutritional products is stored.

[0026] In addition, the information on the first nutritional product of the present invention includes at least the nutritional components of the first nutritional product, and the search for the combination of nutritional products groups at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided by comparing the nutritional component content of the first nutritional product, thereby enabling the consumption of nutritional components corresponding to the selected nutritional product when optimizing the combination of nutritional products.

[0027] In addition, the present invention further includes, prior to the searching step, a step of analyzing an optimization criterion for the nutritional product based on the usage history information of the user terminal, wherein the optimization criterion includes at least one filter criterion for filtering the nutritional product on the database, and in the searching step, the recommendation of the nutritional product can be personalized by grouping the second nutritional product to correspond to the analyzed at least one filter criterion.

[0028] In addition, the analysis of the usage history information of the user terminal of the present invention is performed based on the interaction data for the nutritional product and the user terminal's response to the content provided to the user terminal, and the filter criteria are determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods, thereby allowing nutritional products to be filtered and combined differently according to the user's usage history.

[0029] In addition, the present invention comprises a step of recommending the grouped example nutritional products through a prediction model learned to group at least one nutritional product according to the optimization criteria on the database and provide the example nutritional products, and a step of approving the example nutritional products based on preset approval criteria including medical branch processing, so as to determine whether the example nutritional products meet the provision criteria.

[0030] In addition, the present invention includes a step of providing a plurality of filter criteria to the user terminal when the optimization criteria include a plurality of filter criteria, and the step of providing the filter criteria provides the plurality of filter criteria to the user terminal in a sorted order based on analysis of the usage history information of the user terminal, thereby providing the user with various choices.

[0031] In addition, the present invention further includes, prior to the searching step, a step of extracting at least one keyword from an input provided from a user terminal, and when the keyword is determined to be a keyword related to a nutritional ingredient, the second nutritional product is grouped so as to satisfy the content of the nutritional ingredient related to the keyword, and when the keyword is determined to be a keyword related to an optimization criterion, the second nutritional product is grouped according to the optimization criterion related to the keyword, thereby providing an optimized combination even with a simple search together with the usage history information of the user terminal.

[0032] However, the effects that can be obtained from the present invention are not limited to the effects described above, and other effects may exist.

[0033] Figure 1 is a schematic diagram illustrating a system (1) that provides a personalized nutritional product recommendation service.

[0034] Figure 2 is a schematic block diagram of a nutritional product recommendation device (10) according to one embodiment of the present invention.

[0035] FIG. 3 is a diagram schematically illustrating the interaction between a communication unit (11), an information collection unit (12), and a reference analysis unit (13) according to one embodiment of the present invention.

[0036] Figure 4 is a schematic diagram of information on nutritional products stored in a database.

[0037] FIG. 5 is a diagram illustrating sorting nutritional products on a database by determining recommendation criteria including at least one filter criterion from a reference pool.

[0038] FIG. 6 is a drawing illustrating nutritional products being sorted and provided to a user terminal (30) by a search unit (15) according to one embodiment of the present invention.

[0039] Figure 7 is a drawing illustrating an embodiment in which the recommendation criteria and product sorting change according to input from a user terminal (30).

[0040] FIG. 8 is a diagram illustrating grouping nutritional products on a database by determining optimization criteria including at least one filter criterion from a reference pool.

[0041] FIG. 9 is a schematic diagram illustrating a nutritional product being grouped and provided to a user terminal (30) by a search unit (15) according to one embodiment of the present invention.

[0042] FIG. 10 is a schematic diagram illustrating a nutritional product combination and preset provided together according to an optimization criterion according to one embodiment of the present invention.

[0043] Figure 11 is a schematic flowchart of a step (S1) of providing a nutritional product according to a search of a user terminal according to one embodiment of the present invention.

[0044] Figure 12 is a schematic flowchart of step S14.

[0045] Figure 13 is a schematic flowchart of a step (S2) of optimizing a nutritional product according to one embodiment of the present invention.

[0046] Figure 14 is a schematic flowchart of step S25.

[0047] Below, with reference to the attached drawings, embodiments of the present invention are described in detail so that those skilled in the art can easily implement them. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, irrelevant parts have been omitted for clarity of description, and similar reference numerals have been used throughout the specification to indicate similar elements.

[0048] Throughout the present specification, when a part is said to be "connected" to another part, this includes not only cases where it is "directly connected" but also cases where it is "electrically connected" or "indirectly connected" with another element in between.

[0049] Throughout the present specification, when it is said that a member is located “on”, “above”, “upper”, “lower”, “lower” or “lower” another member, this includes not only cases where a member is in contact with another member, but also cases where another member exists between the two members.

[0050] Throughout this specification, when a part is said to "include" a component, this means that it may include other components, rather than excluding other components, unless specifically stated otherwise.

[0051] Referring to FIG. 1, a system (1) providing a personalized nutritional product recommendation service according to one embodiment of the present invention can recommend nutritional products suitable for the user by communicating with a user terminal (30) via a network. The system (1) may include a nutritional product recommendation device (10) and a user terminal (30). However, the configuration of the system (1) providing a personalized nutritional product recommendation service is not limited thereto.

[0052] The nutritional product recommendation device (10) and the user terminal (30) can communicate with each other through a network. The network refers to a connection structure that enables information exchange between each node, such as terminals and servers. Examples of such networks include, but are not limited to, a 3GPP (3rd Generation Partnership Project) network, an LTE (Long Term Evolution) network, a 5G network, a WIMAX (World Interoperability for Microwave Access) network, the Internet, a LAN (Local Area Network), a Wireless LAN (Wireless Local Area Network), a WAN (Wide Area Network), a PAN (Personal Area Network), a Wi-Fi network, a Bluetooth network, a satellite broadcasting network, an analog broadcasting network, and a DMB (Digital Multimedia Broadcasting) network.

[0053] According to one embodiment of the present invention, a nutritional product recommendation device (10) analyzes input transmitted from a user terminal (30) and, based on the input, increases the convenience of product recommendations tailored to the user's consumption tendencies and provides an optimized combination of nutritional products. In other words, the nutritional product recommendation device (10) according to the present invention can minimize the questionnaire provided to the user for nutritional product recommendations while increasing the convenience of product recommendations tailored to the user's tendencies, and can provide one or more nutritional products selected or provided in an optimized manner tailored to the user's tendencies. Through this, the user can receive recommendations for suitable nutritional products or receive a group of nutritional products combined according to optimized criteria within a minimal amount of time without having to answer numerous questions to provide information such as the user's preferences, health status, and types of nutritional products of interest.

[0054] The nutritional product recommendation device (10) may include any type of server, terminal, or device capable of transmitting and receiving data, content, and various communication signals to and from a user terminal (30) via a network, and having data storage and processing capabilities. As described below, the nutritional product recommendation device (10) may provide an API that authorizes access to an internally stored database and data transmission and reception, and may enable the user terminal (30) to execute appropriate commands.

[0055] In the present specification, the term "nutritional product" can be understood to encompass both nutritional supplements and health functional foods. While nutritional supplements are classified as general pharmaceuticals, health functional foods are not. Health functional foods are manufactured and processed using raw materials or ingredients with beneficial functional properties for the human body in the form of tablets, capsules, powders, granules, liquids, pills, etc., and can be understood as foods that display the functional properties of their main ingredients.

[0056] A combination of nutritional products can be understood as a grouping of at least one nutritional product. This grouping may be performed to form a single combination consisting solely of health functional foods, conversely, a single combination consisting solely of nutritional supplements, or a single combination consisting of health functional foods and nutritional supplements. As described below, one or more nutritional products may be grouped according to optimization criteria and provided to the user terminal (30) as a combination of nutritional products.

[0057] Nutritional ingredients can be understood as the content of nutrients such as vitamins, minerals, and essential fatty acids. Nutritional products may contain one or more nutrients and may have main and secondary ingredients. The main ingredient represents the claimed functionality, while the remaining ingredients, excluding the main ingredient and additives, can be understood as the remaining ingredients.

[0058] The optimization criteria may be criteria for grouping or selecting nutritional products within the database to provide a combination of nutritional products to the user terminal. Meanwhile, the recommendation criteria may be criteria for sorting nutritional products within the database to recommend nutritional products to the user terminal. The optimization criteria and recommendation criteria may include at least one filter criterion for filtering nutritional products in the database. Furthermore, the filter criterion may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods, as described below.

[0059] The usage record information can be understood as a record generated between the nutritional product recommendation device (10) and the user terminal (30), and may include activity details (logs) of the user terminal (30) within the environment provided by the nutritional product recommendation device (10). As described below, optimization criteria for nutritional products can be analyzed based on the usage record information of the user terminal.

[0060] A nutritional product recommendation device (10) can communicate with a user terminal (30) to provide a recommendation of a nutritional product or a combination of nutritional products to the user terminal (30). Specifically, the nutritional product recommendation device (10) can sort nutritional products according to recommendation criteria in response to an input from the user terminal (30) and provide the sorted nutritional products to the user terminal (30), and can also provide a related preset to the user terminal. In addition, the nutritional product combination can be searched for and provided to the user terminal (30) according to an optimization criterion based on information on at least one first nutritional product. That is, the nutritional product recommendation device (10) can sort nutritional products according to recommendation criteria and provide the sorted nutritional products to the user terminal (30), thereby allowing the user to check and compare information on nutritional products according to the input or analyzed recommendation criteria, and can suggest modification of the nutritional product combination or provide nutritional products that meet the user's preference by providing a nutritional product combination to the user terminal (30) according to the optimization criterion.

[0061] The nutritional product recommendation device (10) may include any type of computer system or computer device, such as a microprocessor, a mainframe computer, a digital processor, a portable device, or a device controller. The processor is a controller that performs the role of a CPU and controls input / output and internal operations related to each component of a device connected to the processor. The processor may be composed of one or more cores and may include any type of processor for data processing, data analysis, and / or deep learning, such as a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), or a tensor processing unit (TPU) of a computing device.

[0062] In addition, the nutritional product recommendation device (10) can analyze input from a user terminal (30), establish optimization criteria or recommendation criteria based on the input, and sort or group nutritional products accordingly. Conversely, even without input from the user terminal (30), the nutritional product recommendation device can sort and / or group nutritional products by establishing optimization criteria or recommendation criteria based on analysis of the user terminal's usage history information. In other words, the user can optionally provide input for establishing recommendation criteria or optimization criteria through the user terminal (30).

[0063] Referring to FIG. 2, the nutritional product recommendation device (10) may include a communication unit (11), an information collection unit (12), a reference analysis unit (13), a learning unit (14), a search unit (15), and a database.

[0064] The communication unit (11) may be equipped to communicate with a user terminal (30) via a network, and may transmit commands or requests to the user terminal (30) and receive responses. In one embodiment, the communication unit (11) may authorize access of the user terminal (30) based on an API (Application Programming Interface).

[0065] Referring to FIG. 3, the communication unit (11) can obtain usage history information of a user terminal (30) through communication with the user terminal and transmit it to the reference analysis unit (13). The usage history information may include activity logs of the user terminal, and the usage history information can be analyzed to filter out the user's consumption tendencies or desired nutritional products.

[0066] In one embodiment, the communication unit (11) may be configured to provide a second nutritional product grouped from a selection of at least one first nutritional product to the user terminal. Here, the first nutritional product may be at least one nutritional product input from the user terminal (30), and may also be at least one nutritional product pre-stored in the database corresponding to member information. For example, the user may select nutritional products currently being taken on the UI provided through the user terminal (30), and the nutritional products selected on the user terminal (30) may be provided as the first nutritional products. After grouping the second nutritional products according to an optimization criterion in the search unit (15) described below, the grouped second nutritional products may be provided to the user terminal (30). As another example, if a user is identified through logging in or the like, nutritional products taken or selected by the user stored in the database are provided as first nutritional products, and after the search unit (15) groups second nutritional products according to optimization criteria, the grouped second nutritional products can be provided to the user terminal (30).

[0067] Additionally, the communication unit (11) can recommend or optimize nutritional products based on health-related information stored in a server or database when a user logs in via a user terminal (30), by transmitting this information. That is, when an action that can identify a member, such as logging in, is performed on the user terminal (30), nutritional products can be recommended based on the health-related information of the identified member.

[0068] Meanwhile, in cases where the user's health-related information stored in the database cannot be found, such as when the user does not log in on the user terminal (30), or even if it is possible to find it within the database, nutritional products can be recommended by analyzing the usage history information of the user terminal (30) to request minimal information from the user, or keywords can be extracted from input through the UI and analyzed to recommend nutritional products. This will be described later.

[0069] In addition, when logging in from a user terminal (30), the usage history information of the user terminal (30) can be stored in a database, and the optimization criteria or recommendation criteria analyzed through the usage history information can be stored in the corresponding member information.

[0070] Continuing with reference to FIGS. 2 and 3, the information collection unit (12) extracts and transmits keywords for searching for nutritional products from a request transmitted or input from a user terminal (30), and generates and returns the operating result of the nutritional product recommendation device (10). In one embodiment, the information collection unit (12) may be provided in the form of a chatbot, but it is not excluded that it may be provided in other ways. As illustrated in FIG. 3, the information collection unit (12) may include a UI module (111), a chatbot module (112), a first NLU module (123), and a second NLU module (124).

[0071] The above UI module (121) can receive a request from a user terminal (30) and return the operation result of the nutritional product recommendation device (10) to the user terminal (30).

[0072] In one embodiment, the information collection unit (12) can extract at least one keyword from the conversational input provided from the user terminal (30), and generate a conversational answer from the search result of the search unit (15) based on the optimization criteria or recommendation criteria analyzed by the criterion analysis unit (13) and transmit the generated conversational answer to the user terminal. To this end, the chatbot module (122) can chat with the user by searching and extracting text information from a database (DB) and outputting the search result. That is, the chatbot module (122) can provide the nutritional product search result transmitted to the user terminal (30) in a conversational format based on the input content through the UI module (121), and can provide voice information or text information by preset rules or artificial intelligence (AI).

[0073] The first NLU module (113) can extract keywords by analyzing input including text or voice transmitted from a user terminal. In addition, the second NLU module (114) can generate a conversational output from the nutritional product search results returned from the search unit. The first NLU module (123) and the second NLU module (124) can analyze the meaning of text data. The first NLU module (123) can generate keywords for reference analysis as a result of analyzing the text data, and the second NLU module (124) can generate a conversational output as a result of analyzing the text data.

[0074] For example, input from a user terminal (30) may be a recommendation for a nutritional product with a specific purpose. As described later in FIG. 7, the first input (S12) from the user terminal (30) may be "recommendation for male eye health nutritional supplements." Such input may be analyzed by the first NLU module (123), and one or more keywords, such as "male," "eyes," and "nutritional supplements," may be extracted. In other words, multiple keywords may be extracted for a single input.

[0075] Meanwhile, the second NLU module (124) can generate an interactive output based on the search results from the search unit (15). For the first input, the interactive output can be generated by sorting nutritional products according to the work performed in the search unit (15). For example, the response message generated and transmitted by the second NLU module (114) can be an interactive message including 'lutein' and 'eye fatigue improvement', and can provide information on the searched nutritional products together with the interactive output (see S14 of FIG. 7).

[0076] Continuing with reference to FIGS. 2 and 3, the criteria analysis unit (13) can analyze recommendation criteria and / or optimization criteria for nutritional products. The criteria analysis unit (13) can analyze recommendation criteria and / or optimization criteria from keywords extracted and transmitted by the information collection unit (12) and usage history information of the user terminal (30) transmitted from the communication unit (11). Nutritional products can be sorted and provided to the user terminal according to the analyzed recommendation criteria, and nutritional products can be grouped and provided to the user terminal according to the optimization criteria. In one embodiment, the nutritional product recommendation device (10) can obtain indirect personal data about the user based on usage history information without providing a questionnaire to the user terminal (30) and provide nutritional product recommendations that match the obtained data. In one embodiment, the criteria analysis unit (13) can derive optimization criteria and / or recommendation criteria for nutritional products based on the usage history information of the user terminal. Additionally, the criteria analysis unit (13) may derive optimization criteria and / or recommendation criteria for nutritional products based on keywords extracted from input from a user terminal. The criteria analysis unit (13) may include a usage history information analysis module (131) and a keyword analysis module (132).

[0077] The usage record information analysis module (131) can analyze the usage record information of the user terminal (30) to derive recommendation criteria and / or optimization criteria for nutritional products. As described above, the usage record information can be understood as a record generated between the nutritional product recommendation device (10) and the user terminal (30), and can include activity details (logs) of the user terminal (30) within the environment provided by the nutritional product recommendation device (10).

[0078] Usage history information may include interaction data regarding nutritional products and the user's response to content provided to the user's terminal. Interaction data may include searching for nutritional products or nutritional ingredients, viewing nutritional ingredients or detailed information or reviews, writing reviews for nutritional products, purchasing nutritional products, liking nutritional products, and saving nutritional products to storage.

[0079] Meanwhile, the content provided to the user terminal may include videos, images, text, columns, papers, etc. The user terminal (30) can react to the content provided to the user terminal.

[0080] The response of a user terminal can be understood as information accessed to content by input from the user terminal (30). If the content provided to the user terminal is of the type such as a video that changes over time within a predetermined area of ​​the page, the response may include information accessed to the video, such as the content's playback time, the number of playbacks, and the number of times the video has been watched to the end (content completion rate). On the other hand, if the content provided to the user terminal is provided on a page, such as an image or text, the response may include information accessed within the page, such as the time spent within the page where the content is displayed, and the number of times the content displayed within the page has been scrolled to the end (content completion rate).

[0081] The usage record information analysis module (131) can analyze usage record information, and the usage record information analysis can be performed based on at least one of interaction data regarding nutritional products and the user terminal's response to content provided to the user terminal. In a preferred embodiment, the usage record information analysis module (131) can analyze usage record information based on both interaction data regarding nutritional products and the user terminal's response to content, and can adopt filter criteria highly related to the user's interests.

[0082] The adoption of highly relevant filter criteria can be performed by a predictive model trained to predict filter criteria that are judged to have a relevance exceeding a threshold based on interaction data about nutritional products and the user terminal's responses to content provided to the user terminal. In this case, interaction data about nutritional products that are directly linked to preferences for nutritional products can be given greater weight than the user terminal's responses.

[0083] Analysis of usage history information can be performed based on the user terminal (30). Usage history can be tracked for each user terminal (30) accessing the environment provided by the nutritional product recommendation device (10), and usage information can be analyzed based on the corresponding user terminal (30). In addition, higher weight can be placed on usage history information recorded more recently in time series on the same user terminal (30).

[0084] Meanwhile, when the member information of a user terminal (30) is recognized through a method such as login, the recommendation criteria and / or optimization criteria derived from the analysis of usage information for the user terminal (30) may be stored in a database, at least temporarily. The stored recommendation criteria and / or optimization criteria may be applied to search for recommendations and / or optimization of nutritional products when a user logs in from another user terminal (30).

[0085] The keyword analysis module (132) can derive recommendation criteria and / or optimization criteria for nutritional products based on keywords transmitted from the information collection unit (12). The keywords transmitted from the information collection unit (12) are keywords for searching nutritional products, and may be keywords related to setting criteria for recommendation or optimization of nutritional products. Unlike the recommendation criteria or optimization criteria, they may also be keywords for nutritional ingredients. The keyword analysis module (132) can adopt filter criteria based on keywords when it determines that the keywords are related to recommendation criteria and / or optimization criteria.

[0086] For example, if the input from the user terminal is "recommendation of nutritional supplements for male eye health," the keywords may be male, eye, and nutritional supplement. Here, "eye" is determined to be a keyword for nutritional ingredients, and thus, the search unit (15) described below may recommend nutritional ingredients related to eye health, such as nutritional ingredients containing lutein. On the other hand, the keyword "male" is determined to be a keyword related to recommendation criteria or optimization criteria, and thus, a filter criterion may be adopted to sort nutritional products that are popular with male users among a pool of criteria related to intake preferences.

[0087] The determination of keywords can be performed by a learning model trained to derive filter criteria from keywords, and can also be performed according to a preset algorithm to derive nutritional ingredients or filter criteria for specific keywords. In addition, the determination of keywords can adopt multiple filter criteria by referring to the correlation between two or more keywords, and depending on the keywords being analyzed, nutritional products that do not contain nutritional ingredients related to the keywords can be searched for, or analyzed to exclude them from the filter criteria. For example, if the keywords extracted from the input of the user terminal (30) are 'vitamin B2' and 'without', nutritional products that do not contain vitamin B2 as a main ingredient among the nutritional ingredients can be searched for.

[0088] The database may store data including information about nutritional products, a reference pool, member information for user specific purposes, and health-related information of the user corresponding to the member information.

[0089] Referring to Figure 4, an example of information about a nutritional product can be understood. Information about a nutritional product can include product-related information such as dosage instructions, whether it is a nutritional supplement or a health functional food, and main ingredients. Furthermore, information about the nutritional product can include information about its manufacturing process, such as raw material certification tags, country of origin, manufacturer, and manufacturing method. Furthermore, it can include sales-related information, such as price, model, ranking on sales platforms, and purchase data.

[0090] Referring to Figures 5 and 8, the criteria pool can be understood as a set of criteria that allow for filtering, sorting, or grouping nutritional products in the database according to recommendation criteria or optimization criteria. The criteria pool may include criteria related to medical branch processing and criteria related to the user's preference for how to consume nutritional products. Criteria related to medical branch processing may include criteria related to the effects of nutritional components, such as side effects of nutritional components or nutritional products, whether the effects increase or decrease when nutritional components are taken simultaneously, and restrictions on taking nutritional components due to diseases or allergies. Meanwhile, criteria related to consumption preferences may include criteria that may influence nutritional product selection based on preferences and consumption tendencies, such as nutritional product formulation, manufacturing-related certification, the number of nutritional products consumed, and consumption methods.

[0091] Member information for user identification can be understood as information that enables user identification, such as login, and health-related information of the user corresponding to the member information may include the user's gender, age, allergies, health status, nutritional products currently being consumed, etc. The user's health-related information may be obtained in response to a questionnaire provided in advance to the user terminal (30) from the nutritional product recommendation device (10), but may also be obtained through methods other than the questionnaire.

[0092] Referring again to Figure 2, the learning unit (14) can build a neural network and train a prediction model that predicts a target result value based on the provided data.

[0093] A neural network for recognizing training data may include multiple network nodes with weights. The multiple network nodes may include a deep learning model. The neural network may be composed of a set of interconnected computational units, referred to as nodes (neurons), interconnected by one or more links, and one or more nodes connected through the links may form a relationship between input nodes and output nodes. Some of these may form a layer based on distances from the initial input node, and the neural network may form a deep neural network (DNN) through multiple hidden layers in addition to an input layer and an output layer. The neural network can optimize and set the weight values ​​within the neural network by training the training data and identify the desired output data.

[0094] The input layer of a neural network may receive information extracted from acquired or preprocessed data. The input layer may include a normalization layer for arranging and / or normalizing the data input to input nodes. The hidden layer of the neural network may perform operations in at least one layer, and may include a convolution layer and a dropout after the convolution layer. As data passes through the hidden layer, data representing the features of the input data may be generated. The feature extraction region within the hidden layer may include multiple repetitions of the convolution layer and the pooling layer, or the convolution layer may be repeated multiple times. Thereafter, the desired output data may be derived from the output layer.

[0095] In one embodiment, the predictive model may be trained to provide example nutritional products from at least one keyword and at least one filter criterion determined based on usage history information of the user terminal (30). That is, the predictive model may be trained to provide example nutritional products from a database from analyzed recommendation criteria or optimization criteria. When recommendation criteria including at least one filter criterion are determined, the predictive model may be trained to provide example nutritional products sorted and provided from the database according to the recommendation criteria. Furthermore, when optimization criteria including at least one filter criterion are determined, the predictive model may be trained to provide example nutritional products grouped and provided from the database according to the recommendation criteria. For training the predictive model, nutritional products filtered within the database according to at least one filter criterion selected with priority may be used as labeled learning data.

[0096] In another embodiment, the prediction model may learn to adopt filter criteria within a reference pool from keywords derived from the information collection unit (12), and may also learn to adopt filter criteria within the reference pool from usage history information of the user terminal.

[0097] Furthermore, the predictive model can determine priorities among filter criteria through learning. Nutritional products within the database can be sorted or grouped based on the priorities of the determined filter criteria and recommended to the user terminal (30).

[0098] The above learning performed in the learning unit (14) may be a known artificial intelligence model, such as a DNN, a CNN, a network having a structure including one or more encoders and one or more decoders corresponding thereto, a network configured to perform knowledge distillation learning in which one network transfers knowledge acquired through learning to another network to enable learning, etc.

[0099] Referring again to FIG. 2, the search unit (15) may be configured to search for combinations of nutritional products within the database based on optimization criteria for the first nutritional product. Furthermore, the search unit (15) may also sort the nutritional products within the database based on recommendation criteria. In one embodiment, the search unit (15) may search for nutritional products using at least one keyword and a predictive model trained to provide example nutritional products based on usage history information of the user terminal (30).

[0100] In one embodiment, the search unit (15) may perform branch processing within the filtered nutritional product range based on filter criteria or keywords, and then receive recommendations for exemplary nutritional products based on a prediction model. Afterwards, it may be determined whether the exemplary nutritional products meet the approval criteria and then approved, thereby providing the user terminal (30) with a sorting or grouping of the nutritional products. Here, the approval criteria are for verifying the recommendation of the nutritional products, and may be preset with reference to the Ministry of Food and Drug Safety notice, etc., and may include medical branch processing. Approval criteria for verifying the recommendation may include dosage restrictions, whether simultaneous intake is recommended, and whether nutritional ingredients including main ingredients are satisfactory. In other words, when providing exemplary nutritional products based on an artificial intelligence prediction model, exemplary nutritional products with a reliability level below a threshold are not provided to the user terminal (30).

[0101] The search unit (15) may include a sorting module (not shown) that sorts nutritional products within a database according to recommendation criteria. Sorting nutritional products may involve sorting nutritional products according to recommendation criteria including at least one filter criterion in a database storing information on multiple nutritional products.

[0102] Referring to Figure 5, the recommendation criteria may be determined based on the user terminal's usage history information and may include at least one filter criterion for filtering nutritional products in the database. Furthermore, the filter criterion may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods. Based on at least one filter criterion determined as the recommendation criterion, the sorting module may sort and provide nutritional products in the database.

[0103] FIG. 6 illustrates a process in which nutritional products are sorted and provided to a user terminal (30) by a search unit (15) according to one embodiment of the present invention. Filter criteria for sorting nutritional products may be displayed as interactive output by the second NLU module (124) described above on the user terminal, and nutritional products sorted according to the filter criteria may be provided. For example, if the filter criterion is low price, "same main ingredient but lower price ~" may be provided as interactive output by the second NLU module (124), and nutritional products sorted in descending order of price may be provided. At this time, the sorting and recommendation of nutritional products may be performed within the range of nutritional products that satisfy the content of the main ingredient of the selected nutritional product. That is, nutritional products that satisfy the content of at least one nutritional ingredient that is predetermined or provided may be sorted and provided according to the recommendation criteria by comparing them with the nutritional ingredient content of the first nutritional product selected or provided before sorting.

[0104] Here, satisfaction can be understood as meaning the same or greater nutritional content. However, it can also be understood as satisfying a nutritional content exceeding a set threshold ratio. Furthermore, depending on the embodiment, satisfaction can be understood as not exceeding the set recommended daily intake of a nutritional component. Furthermore, satisfaction can be interpreted as meaning "less" or the like based on keywords extracted from the input of the user terminal (30) or analysis of the user terminal's usage history information. In other words, satisfaction can be interpreted as being suitable for sorting or grouping nutritional products based on the input of the user terminal (30) or usage history information.

[0105] In addition, although FIG. 6 illustrates that nutritional products are sorted and provided according to one filter criterion, the embodiment of the present invention does not exclude that nutritional products are sorted and provided according to two or more filter criteria, and when two or more filter criteria are applied, nutritional products may be sorted and provided according to the priority of the filter criteria.

[0106] Figure 7 illustrates an embodiment in which recommendation criteria and product sorting vary depending on input from a user terminal (30). First, when a first search input is performed from a user terminal, the information collection unit (12) can extract keywords from the input of the user terminal (S12). The criteria analysis unit (13) can analyze recommendation criteria based on keywords and usage history information of the user terminal. The search unit (15) can sort and provide nutritional products according to keywords and recommendation criteria (S14). At this time, the recommendation criteria analysis and nutritional product sorting can be performed by a prediction model learned by the learning unit (14).

[0107] For additional nutritional product searches, a second search may be input from the user terminal. When the second search input is performed, the information collection unit (12) may extract keywords from the second input (S12). When input from the user terminal is performed chronologically to search for nutritional products, nutritional ingredients or filter criteria related to keywords extracted through subsequent inputs may be prioritized for nutritional product searches. As illustrated, when the second input "Find me liquid, not pills" is performed, keywords including pills and liquids may be extracted, and through input analysis, filter criteria may be added to the recommendation criteria to exclude pills and include liquid formulations. These filter criteria may be applied to the sorting of nutritional products using an "and" condition with the filter criteria within the existing recommendation criteria, and filter criteria related to formulations may take precedence over other filter criteria.

[0108] Additionally, a preset may be provided to the user terminal along with the rearranged nutritional products (S15). The preset provided to the user terminal may include filter criteria, information about the nutritional products, and other nutritional ingredients. In a preferred embodiment, the preset provided to the user terminal may be provided based on analysis of the user terminal's usage history information. For example, if the user terminal (30) has a search history for nutritional products for gifting, nutritional products based on the search results are stored in the storage, and nutritional products for which information about the nutritional products is viewed are nutritional products with high sales rankings, a preset titled "Recommendation of Products with High Brand Awareness" may be provided based on this. In this way, the preset provided may vary depending on the user terminal's (30) usage history information.

[0109] In addition, the search unit (15) may include a grouping module (not shown) that groups nutritional products within the database according to optimization criteria. In this case, the search for a combination of nutritional products may be performed by grouping at least one second nutritional product according to optimization criteria in a database in which information on multiple nutritional products is stored, based on information on a first nutritional product. The selection of the first nutritional product may be performed by input from a user terminal (30), such as by specifying or checking at least one nutritional product, or may be performed by retrieving information previously stored in the database.

[0110] Referring to Figure 8, the optimization criteria may be determined based on the user terminal's usage history information and may include at least one filter criterion for filtering nutritional products in the database. Furthermore, the filter criterion may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods. Based on at least one filter criterion determined as the optimization criterion, the grouping module may group and provide nutritional products in the database.

[0111] At this time, the search unit (15) can group at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided by comparing it with the nutritional component content of the first nutritional product.

[0112] FIG. 9 illustrates nutritional products being grouped and provided to a user terminal (30) by a search unit (15) according to one embodiment of the present invention. Optimization criteria for searching for nutritional product combinations may be displayed as interactive output on the user terminal by the second NLU module (124) described above, and combinations of nutritional products grouped according to the optimal criteria may be provided. For example, if the filter criterion is low price, "Find the lowest price product combination" may be provided as output by the second NLU module (124), and second nutritional products grouped to have lower prices than the first nutritional products selected may be provided.

[0113] At this time, the search for a combination of nutritional products performed in the search unit (15) may group at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided, compared to the nutritional component content of the first nutritional product. That is, at least one second nutritional product may be provided so as to have the same or satisfactory nutritional component content as or satisfy the nutritional component content of the grouped nutritional product combination, compared to the nutritional component content of the first nutritional product. Preferably, the second nutritional product may be grouped so as to satisfy the content of the main ingredient of the first nutritional product, compared to the nutritional component content of the first nutritional product.

[0114] Although Figure 9 illustrates grouping nutritional products using the "lowest price" optimization criterion as the filter criterion, optimization criteria that apply other filter criteria may be applied. For example, a filter criterion related to the number of nutritional products may be determined to provide a selection of five or more primary nutritional products and group three or fewer secondary nutritional products. Furthermore, a filter criterion related to the formulation may be determined to group jelly nutritional products of different formulations, providing a selection of refined nutritional supplements. Furthermore, a filter criterion related to over-consumption may be determined to prevent the intake of nutrients exceeding the recommended daily allowance, and a filter criterion considering side effects or interactions may be determined. Furthermore, the optimization criterion may be a combination of two or more of these filter criteria. As described above, the optimization criterion may be derived by analyzing the usage history information of the user terminal.

[0115] Also, referring to FIG. 10, a preset may be provided to a user terminal along with a grouped combination of nutritional products. The preset provided to the user terminal may include filter criteria, information about nutritional products, other nutritional ingredients, etc. In a preferred embodiment, the preset provided to the user terminal may be provided based on analysis of the usage history information of the user terminal. For example, if the user terminal (30) has a search history for nutritional products for gifts, nutritional products based on the search results are stored in the storage, and nutritional products for which information about nutritional products is viewed are nutritional products with high sales rankings, a preset called "for gifts" may be provided based on this. In this way, the preset provided may vary depending on the usage history information of the user terminal (30). Here, the filter criteria provided as a preset may be sorted based on analysis of the usage history information of the user terminal, and may be provided to the user terminal in the sorted ranking. Such provision of presets may also be performed when the optimization criteria include multiple filter criteria.

[0116] The user terminal (30) may be, for example, a smartphone, a smart pad, a tablet PC, a wearable device, and all kinds of wireless communication devices such as a PCS (Personal Communication System), a GSM (Global System for Mobile communication), a PDC (Personal Digital Cellular), a PHS (Personal Handyphone System), a PDA (Personal Digital Assistant), an IMT (International Mobile Telecommunication)-2000, a CDMA (Code Division Multiple Access)-2000, a W-CDMA (W-Code Division Multiple Access), a Wibro (Wireless Broadband Internet) terminal, and a fixed terminal such as a desktop computer or a smart TV. The user terminal (30) is equipped to be able to communicate with the nutritional product recommendation device (10), and may include an input module for input, and an output module for displaying the results transmitted from the nutritional product recommendation device (10).

[0117] Through the above description, the nutritional product recommendation device (10) and the system (1) providing a personalized nutritional product recommendation service according to embodiments of the present invention can be understood. Hereinafter, a personalized nutritional product recommendation method according to one embodiment of the present invention will be described based on the contents described in detail above. The personalized nutritional product recommendation method of the present invention can be performed by the nutritional product recommendation device (10) or the system (1) providing a personalized nutritional product recommendation service through a service platform, and can analyze the input transmitted from the user terminal (30) and, based on the input, increase the convenience of product recommendation according to the user's consumption tendency and provide an optimized combination of nutritional products. The personalized nutritional product recommendation method can include a step (S1) of providing a nutritional product according to a search of the user terminal, and a step (S2) of optimizing the nutritional product.

[0118] Referring to FIGS. 11 and 12, a step (S1) of providing a nutritional product according to an embodiment of the present invention may be performed to provide a search for a nutritional product that matches the user's consumption tendencies from an input of a user terminal, and may be understood as a method of providing a nutritional product according to a user search, depending on the embodiment. Step S1 may include a step (S11) of collecting information related to nutritional product recommendation criteria, a step (S12) of extracting keywords from an input of a user terminal, a step (S13) of analyzing keywords and recommendation criteria, a step (S14) of sorting and providing nutritional products according to keywords and recommendation criteria, and a step (S15) of providing a preset to a user terminal.

[0119] Step S11 can be understood as a step for collecting recommendation criteria-related information from the user terminal to derive recommendation criteria. In one embodiment, the communication unit (11) may collect usage history information from the user terminal (30). Furthermore, if the user can be identified through login or other means, information for deriving recommendation criteria stored in the database can be retrieved.

[0120] Step S12 may extract keywords from the user terminal input for search. Step S12 may be performed by the information collection unit (12), and may extract and transmit keywords for nutritional product search from a request transmitted or input from the user terminal (30). For example, the information collection unit (12) may be provided in the form of a chatbot, and may extract at least one keyword from the user terminal input.

[0121] Step S13 may be performed in the criteria analysis unit (13), and keywords and recommendation criteria may be analyzed for nutritional product recommendations. If the keyword is determined to be related to a nutritional ingredient, nutritional products containing the nutritional ingredient related to the keyword as a main ingredient may be searched within the database. Furthermore, if the keyword is determined to be related to a recommendation criterion, recommendation criteria related to the keyword may be derived. As described above, the recommendation criterion may include at least one filter criterion for filtering nutritional products in the database, and the filter criterion may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods.

[0122] Step S14 can be performed in the sorting module of the search unit (15), and can be understood as a process for sorting nutritional products based on keywords and recommendation criteria and providing them to a user terminal. Nutritional products can be sorted based on recommendation criteria including at least one filter criterion in a database storing information on multiple nutritional products.

[0123] In one embodiment, step S14 may be performed using a learned predictive model that sorts nutritional products based on recommendation criteria in a database and provides example nutritional products. In this case, step S14 may include a step of branching within a filtered nutritional product range (S141), a step of recommending example nutritional products (S142), and a step of approving example nutritional products (S143), as illustrated in FIG. 12 .

[0124] Step S141 can be understood as filtering nutritional products based on filter criteria or keywords. Recommendation criteria may vary based on the user terminal's usage history, and keywords may also vary based on the user terminal's input. Therefore, filtered nutritional products may be processed separately and displayed differently for each user terminal based on different usage history information and extracted keywords.

[0125] Step S142 is a step for recommending exemplary nutritional products using a predictive model, and exemplary nutritional products may be recommended from a pre-trained predictive model. As described above, the predictive model may be trained to provide exemplary nutritional products based on at least one keyword and at least one filter criterion determined based on the usage history information of the user terminal (30). In step S142, exemplary nutritional products may be provided in sorted order.

[0126] Step S143 determines whether the example nutritional product meets the provision criteria, allowing the user to approve the example nutritional product or provide another example nutritional product. Based on preset approval criteria, including medical branch processing, if the example nutritional product meets the approval criteria, a list of nutritional products can be provided to the user terminal. If the example nutritional product does not meet the approval criteria, the user can receive another example nutritional product.

[0127] In one embodiment, multiple inputs may be provided for search from a user terminal. For example, as illustrated in FIG. 7, a first search input may be provided from a user terminal, and a first nutritional product recommendation may be made based on keywords and recommendation criteria (S14). If a second search input is provided to perform the search again or to better suit the user's intent, steps S12 to S14 may be re-performed. In this case, when steps S12 to S14 are re-performed, nutritional products may be searched by prioritizing nutritional ingredients or filter criteria related to keywords extracted by subsequent inputs. In other words, nutritional products may be sorted by prioritizing or assigning high weight to nutritional products or filter criteria related to keywords extracted by the second search input.

[0128] Referring back to FIG. 11, step S15 may be performed by the communication unit, and may provide the user terminal with presets, such as filter criteria, information about nutritional products, and other nutritional ingredients. In one embodiment, the presets provided to the user terminal may be provided based on analysis of the user terminal's usage history information. Additionally, multiple presets may be provided to the user terminal in a sorted order based on analysis of the user terminal's usage history information.

[0129] Referring to FIGS. 13 and 14, the step (S2) of optimizing a nutritional product may be performed to provide an optimized combination of nutritional products to suit the user's consumption tendencies from a selection provision for at least one first nutritional product, and may be understood as a personalized nutritional product optimization method depending on the embodiment. The step S2 may include a step of providing a selection for the first nutritional product (S21), a step of receiving information about the first nutritional product (S22), a step of collecting information related to optimization criteria (S23), a step of analyzing optimization criteria for the nutritional product (S24), a step of searching for a combination of nutritional products according to the optimization criteria (S25), and a step of providing the combination of nutritional products to the user terminal (S26).

[0130] Step S21 provides a selection of at least one first nutritional product to be optimized. This selection may be provided through input from a user terminal, or, in some cases, through nutritional products stored in a database. For example, the user may select or input multiple nutritional products currently being consumed on the user terminal (30). Alternatively, in a situation where the user can recognize the nutritional product, such as logging in, the user may select the nutritional product currently being consumed, and the selected nutritional product may be stored in the database and then provided.

[0131] In one embodiment, a user may select a nutritional product provided in step (S1) of providing the aforementioned nutritional product. During this process, multiple nutritional products may be selected by repeating step S1 multiple times. The selected nutritional product may be provided as the first nutritional product for optimization.

[0132] Step S22 can receive information about the selected first nutritional product, including information including nutritional components of the first nutritional product stored in the database.

[0133] Step S23 can be understood as a step of collecting optimization criteria-related information from the user terminal to derive optimization criteria. In one embodiment, the communication unit (11) can collect usage history information from the user terminal (30). Furthermore, if the user can be identified through login or other means, information for deriving optimization criteria stored in the database can be retrieved.

[0134] Step S24 may be performed in the criteria analysis unit (13), which analyzes optimization criteria for exploring nutritional product combinations. Analysis of optimization criteria for nutritional products may be based on user terminal usage history information. Analysis of user terminal usage history information may be performed based on interaction data regarding nutritional products and the user terminal's responses to content provided to the user terminal.

[0135] The optimization criteria derived based on the usage history information of the user terminal may include at least one filter criterion for filtering nutritional products in the database, and the filter criterion may be determined from a pool of criteria related to medical branch treatment and preferences for nutritional product intake methods.

[0136] Step S25 may be performed in the grouping module of the search unit (15), and may search for a combination of nutritional products based on optimization criteria for the received first nutritional product. The search for a combination of nutritional products may be understood as grouping at least one second nutritional product based on the optimization criteria in a database in which information on multiple nutritional products is stored. Since the optimization criteria include one or more filter criteria, step S25 may group the second nutritional products to correspond to at least one of the analyzed filter criteria.

[0137] In step S25, the search for a combination of nutritional products may group at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided, compared to the nutritional component content of the first nutritional product.

[0138] In one embodiment, step S25 may be performed using a learned predictive model that groups nutritional products based on optimization criteria in a database to provide a combination of example nutritional products. In this case, step S25 may include a step of branching within a filtered nutritional product range (S251), a step of recommending example nutritional products (S252), and a step of approving example nutritional products (S253), as illustrated in FIG. 14 .

[0139] Step S251 can be understood as filtering nutritional products based on filter criteria or keywords. Optimization criteria may vary depending on the user terminal's usage history, and keywords may vary depending on the user terminal's input. Therefore, filtered nutritional products may be processed separately and displayed differently for each user terminal based on different usage history information and extracted keywords.

[0140] Step S252 is a step for recommending exemplary nutritional products using a predictive model. Recommendations for exemplary nutritional products may be obtained from a pre-trained predictive model. As described above, the predictive model may be trained to provide exemplary nutritional products based on at least one keyword and at least one filter criterion determined based on the usage history information of the user terminal (30). In step S252, a combination of exemplary nutritional products may be provided.

[0141] Step S253 determines whether the combination of example nutritional products meets the approval criteria. The combination can be approved or another example nutritional product can be provided. Based on preset approval criteria, including medical branch processing, if the example nutritional product satisfies the approval criteria, the combination of nutritional products can be provided to the user terminal. If the approval criteria are not met, the user can be provided with another combination of example nutritional products.

[0142] When the search for a combination of nutritional products is completed, the combination of nutritional products can be provided to the user terminal (30) in step S26.

[0143] Additionally, step S2 may further include a step (S27) of providing filter criteria to the user terminal. If the optimization criteria include multiple filter criteria, the multiple filter criteria may be provided to the user terminal. In this case, step S27 may provide the multiple filter criteria to the user terminal in a sorted order based on analysis of the user terminal's usage history information. As illustrated in FIG. 10, the filter criteria may be provided as a preset along with a combination of nutritional products provided to the user terminal according to the optimization results.

[0144] Additionally, step S2 may further include a step (S28) of extracting keywords from input from a user terminal. Step S28 may be performed before the searching step (S25). Step S28 may extract at least one keyword from input provided from a user terminal.

[0145] Depending on the keyword analysis, the search at step S25 may vary. Specifically, if the keyword is determined to be related to nutritional content, the second nutritional product may be grouped based on the content of the nutritional content associated with the keyword. Alternatively, if the keyword is determined to be related to an optimization criterion, the second nutritional product may be grouped based on the optimization criterion associated with the keyword.

[0146] According to the above description, nutritional products that match the user's consumption tendencies or purchasing intentions may be recommended based on the user's search input, and nutritional products may be provided in an optimized manner to match the user's purchasing intentions based on the user's usage history information from the user's selection of nutritional products.

[0147] In the above description, each step of steps S1 to S2, S1 and / or S2 may be further divided into additional steps or combined into fewer steps, depending on the implementation of the present invention. Furthermore, some steps may be omitted as needed, and the order of the steps may be changed.

[0148] A personalized nutritional product recommendation method according to one embodiment of the present invention may be implemented in the form of program commands that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program commands, data files, data structures, etc., either singly or in combination. The program commands recorded on the medium may be those specifically designed and configured for the present invention or may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. The above hardware devices may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

[0149] Additionally, the aforementioned personalized nutritional product recommendation method can also be implemented in the form of a computer program or application executed by a computer and stored in a recording medium.

[0150] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single entity may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined manner.

[0151] The scope of the present invention is indicated by the claims described below rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention.

Claims

1. A personalized nutritional product recommendation method performed by an electronic device including at least one processor, A step of being offered a choice of at least one first nutritional product; A step of receiving information about the selected first nutritional product; A step of searching for a combination of nutritional products according to the optimization criteria for the first nutritional product received, and A step of providing the above nutritional product combination to a user terminal is included, A personalized nutritional product recommendation method, characterized in that the search for the above nutritional product combination comprises grouping at least one second nutritional product according to the above optimization criteria in a database in which information on multiple nutritional products is stored.

2. In paragraph 1, Information about the first nutritional product includes at least the nutritional components of the first nutritional product, A personalized nutritional product recommendation method, characterized in that the search for the above nutritional product combination comprises grouping at least one second nutritional product so as to satisfy the content of at least one nutritional component that is predetermined or provided, compared to the nutritional component content of the first nutritional product.

3. In paragraph 2, Before the above searching step, a step of analyzing the optimization criteria for the nutritional product based on the usage history information of the user terminal is further included. The above optimization criteria include at least one filter criterion for filtering nutritional products on the database, A personalized nutritional product recommendation method characterized in that, in the above-mentioned searching step, the second nutritional product is grouped to correspond to at least one of the analyzed filter criteria.

4. In paragraph 3, Analysis of the usage history information of the above user terminal, Interaction data for the above nutritional products, and It is performed based on the user terminal's response to the content provided to the user terminal, A personalized nutritional product recommendation method, characterized in that the above filter criteria are determined from a pool of criteria related to preferences for medical branch treatment and nutritional product intake methods.

5. In paragraph 3, The above searching steps are: A step of recommending the grouped example nutritional products through a prediction model learned to group at least one nutritional product according to the optimization criteria on the database and provide the example nutritional products, and A personalized nutritional product recommendation method, characterized in that it comprises a step of approving the example nutritional product based on preset approval criteria including medical branch processing.

6. In paragraph 3, If the above optimization criteria include multiple filter criteria, a step of providing multiple filter criteria to the user terminal is included. The step of providing the above filter criteria is: A personalized nutritional product recommendation method characterized in that a plurality of the above filter criteria are provided to the user terminal in a sorted order based on analysis of the usage history information of the user terminal.

7. In paragraph 2, Before the above searching step, the method further includes a step of extracting at least one keyword from an input provided from a user terminal, If the keyword is determined to be a keyword related to nutritional ingredients, the second nutritional product is grouped to satisfy the content of the nutritional ingredients related to the keyword. A personalized nutritional product recommendation method characterized in that the second nutritional product is grouped according to the optimization criterion related to the keyword when the keyword is determined to be a keyword related to the optimization criterion.

8. As a nutritional product recommendation device, A communication unit that provides a second nutritional product grouped from a selection of at least one first nutritional product provided to a user terminal; Includes an exploration unit that explores a combination of nutritional products based on optimization criteria for the first nutritional product, A nutritional product recommendation device characterized in that the search for the above nutritional product combination groups at least one second nutritional product according to the above optimization criteria in a database in which information on a plurality of nutritional products is stored.

9. In paragraph 8, It further includes a criteria analysis unit that derives an optimization criterion for the nutritional product based on the usage history information of the user terminal. The above optimization criteria include at least one filter criterion for filtering nutritional products on the database, The above search unit groups the second nutritional products to correspond to at least one of the analyzed filter criteria, Analysis of the usage history information of the above user terminal, Interaction data for the above nutritional products, and A personalized nutritional product recommendation device characterized in that it is performed based on the user terminal's response to the content provided to the user terminal.

10. A system that provides personalized nutritional product recommendation services through a service platform. User terminal, and Including a nutritional product recommendation device that performs communication with the above user terminal, The above nutritional product recommendation device is, A communication unit that provides a second nutritional product grouped from a selection of at least one first nutritional product provided to the user terminal; Includes an exploration unit that explores a combination of nutritional products based on optimization criteria for the first nutritional product, A system for providing a personalized nutritional product recommendation service, characterized in that the search for the above nutritional product combination groups at least one second nutritional product according to the above optimization criteria in a database in which information on multiple nutritional products is stored.

Images