KR20220012158A - Method and apparatus for providing personalized cooking service - Google Patents

Abstract

A server for generating data for cooking of food and a generating method thereof are disclosed. A server for generating data for cooking of food includes: a communication unit; a storage unit storing one or more instructions; and a processor executing the one or more instructions stored by the storage unit. The processor is configured to: obtain text representing a recipe of food; identify a plurality of semantic elements included in the text based on at least one semantic element applied to a plurality of natural language understanding (NLU) models for analysis of the text and output from each of the plurality of NLU models; generate structured first machine readable recipe (MRR) data including a plurality of recipe parameters by using the identified semantic elements; obtain user log data related to operations which a user has performed for a cooking device; and generate second MRR data personalized to the user by updating the plurality of cooking parameters based on the obtained log data. The present invention can provide more easy and intuitive recipes.

Description

METHOD AND APPARATUS FOR PROVIDING PERSONALIZED COOKING SERVICE

Various embodiments relate to a method of generating data for cooking food.

An artificial intelligence (AI) system is a computer system that implements human-level intelligence, and unlike the existing rule-based smart system, the machine learns, judges, and becomes smarter by itself. The more the AI system is used, the higher the recognition rate and the more accurate understanding of user preferences, and the existing rule-based smart systems are gradually being replaced by deep learning-based AI systems.

Artificial intelligence technology consists of machine learning (deep learning) and element technologies using machine learning. Machine learning is an algorithm technology that classifies/learns the characteristics of input data by itself, and element technology is a machine learning algorithm such as deep learning. It is a technology that simulates the functions of cognition and judgment of the human brain using

The fields where artificial intelligence technology is applied include natural language processing that recognizes and applies/processes human language/text, linguistic understanding including machine translation, dialogue system, question and answer, voice recognition/synthesis, etc. There are technical fields such as inference/prediction that logically reason and predict, knowledge/probability-based reasoning, optimization prediction, preference-based planning, recommendation, and the like.

Meanwhile, in the case of a cooking apparatus (oven, etc.) for cooking food, a user directly reads a recipe, and directly inputs a cooking type, a cooking method, and setting information for cooking according to the recipe. However, it is complicated to set the cooking apparatus according to various recipes, and the user does not know the characteristics of the cooking apparatus well, so that the cooking may not be prepared according to the recipe. Accordingly, it is possible to create a recipe in a standardized data form so that it can be read by a cooking device, process the created standardized data form recipe into personalized data to the user, and provide the recipe in an intuitive form to the user. skills are required

Various embodiments, in a server including a plurality of NLU models, identify a plurality of semantic elements in a recipe in text form, and change into structured data that can be read by various cooking devices based on the identified semantic elements Accordingly, it is possible to provide a server for generating data for cooking food, which can automatically set a cooking apparatus using structured data, and provide cooking information to a user in an intuitive form, and an operating method thereof.

As a technical means for achieving the above-described technical problem, a method for a server to generate data for cooking food according to an embodiment includes: acquiring text indicating a recipe for food; identifying a plurality of semantic elements included in the text by using a plurality of natural language understanding (NLU) models for analyzing the text; generating first Machine Readable Recipe (MRR) data in a structured form including a plurality of cooking parameters by using the identified semantic elements; acquiring log data of a user related to an operation in which the user operated the cooking apparatus; and generating, based on the obtained log data, second MRR data personalized to the user by updating the plurality of cooking parameters, wherein the identifying of the plurality of semantic elements comprises: It may be a method of applying each of a plurality of NLU models, and identifying the plurality of semantic elements based on at least one semantic element output from each of the plurality of NLU models.

In addition, a server for generating data for cooking food according to an embodiment includes: a communication unit; a storage unit for storing one or more instructions; and a processor executing the one or more instructions stored in the storage unit, wherein the processor obtains a text representing a recipe of food, and provides each of a plurality of NLU (Natural Language Understanding) models for analyzing the text. apply, identify a plurality of semantic elements included in the text based on at least one semantic element output from each of the plurality of NLU models, and include a plurality of cooking parameters by using the identified semantic elements Generates first MRR (Machine Readable Recipe) data in a structured form, obtains log data of a user related to an operation in which the user manipulated the cooking apparatus, and based on the obtained log data, the plurality of The server may be, generating the second MRR data personalized to the user by updating the cooking parameters.

A server that provides cooking information according to an embodiment identifies a plurality of semantic elements using a plurality of NLU (Natural Language Understanding) models in text representing a recipe, and generates Machine Readable Recipe (MRR) data based on this. And, by providing a recipe card generated using the MRR data to the user, it is possible to provide an easier and more intuitive recipe.

In addition, when the user cooks according to a recipe, the user can use the cooking apparatus possessed by the user as an optimized setting value by controlling the cooking apparatus using MRR data instead of individually setting the cooking apparatus.

1 is a diagram referenced to explain a method of a server providing food cooking information and controlling cooking apparatuses according to an exemplary embodiment.
2 is a block diagram illustrating a detailed configuration of a server according to an embodiment.
3 is a flowchart illustrating a method in which a server operates to generate first MRR data and second MRR data according to an embodiment.
4 is a diagram for explaining operations of a recipe syntax analysis module included in a server according to an embodiment.
5 is a diagram referenced to explain a method in which a server identifies semantic elements using a plurality of NLU models using a recipe syntax analysis module according to an embodiment.
6 is a diagram for describing a detailed configuration of an additional information generating module that generates additional information based on an identified semantic element included in a server according to an exemplary embodiment.
7 is a diagram referenced to explain a method for a server to generate first MRR data using semantic elements identified from a plurality of NLU models according to an embodiment.
8 is a server according to an embodiment, by using the first MRR data generation module to identify semantic elements from text representing a recipe, and generate a cooking parameter to generate first MRR data that can read the recipe in the cooking device It is a drawing referenced to explain the method.
9 is a diagram for describing a method in which a server acquires user's log data and generates second MRR data based on the acquired log data, according to an embodiment.
10 is a diagram for explaining an embodiment of performing reinforcement learning based on user log data when the server generates second MRR data according to an embodiment.
11 is a diagram referenced to explain a method for a server to generate a recipe card according to an embodiment.
12 is a diagram referenced to explain a method for a server to generate a recipe card for each cooking step according to an embodiment.
13 is a diagram referenced to explain a method for optimizing by changing MRR data generated by a server according to a user's cooking apparatus according to an embodiment;
14 is a diagram referenced to explain a method for a server to change and optimize MRR data generated by a server based on various user information according to an embodiment.
15 is a diagram for explaining the configuration of a cooking information providing system according to an exemplary embodiment.
FIG. 16 is a diagram for explaining an example in which information for cooking food is provided to a user through a client device after MRR data and recipe cards are generated in a server according to an embodiment.
17 is a diagram illustrating a GUI for generating MRR data for a new recipe by using a plurality of pre-generated MRR data in the cooking information providing system according to an embodiment.
18 is a view for explaining an example in which recipe cards for cooking a plurality of foods are provided to a user through a client device in a cooking information providing system according to an embodiment.
19 is a view for explaining an example in which a first recipe card and a personalized second recipe card are provided to a user through a client device in the cooking information providing system according to an embodiment.

Terms used in this specification will be briefly described, and the present invention will be described in detail.

The terms used in the present invention have been selected as currently widely used general terms as possible while considering the functions in the present invention, but these may vary depending on the intention or precedent of a person skilled in the art, the emergence of new technology, and the like. In addition, in a specific case, there is a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the corresponding invention. Therefore, the term used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, rather than the name of a simple term.

The singular expression may include the plural expression unless the context clearly dictates otherwise. Terms used herein, including technical or scientific terms, may have the same meanings as commonly understood by one of ordinary skill in the art described herein. Also, terms including an ordinal number such as 'first' or 'second' used in this specification may be used to describe various components, but the components should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

In the entire specification, when a part "includes" a certain element, this means that other elements may be further included, rather than excluding other elements, unless otherwise stated. In addition, terms such as “unit” and “module” described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software, or a combination of hardware and software.

Hereinafter, with reference to the accompanying drawings, the embodiments of the present invention will be described in detail so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

In the disclosed embodiment, the Natural Language Understanding (NLU) may be an artificial intelligence model trained to interpret text to obtain semantic elements in the text. In this case, the artificial intelligence model may be composed of a plurality of neural network layers. Each of the plurality of neural network layers may include a plurality of weights, and a neural network operation may be performed through an operation between an operation result of a previous layer and the plurality of weights.

In the disclosed embodiment, MRR (Machine Readable Recipe) data is standardized data generated from text representing a recipe, which is data in an unstructured form, and may be data in a structured form converted to be read by a cooking device. The server according to the disclosed embodiment may control various types of different cooking devices using the standardized MRR data. MRR data according to the disclosed embodiment may consist of cooking parameters generated from the identified semantic elements, and the cooking parameter may be a parameter related to a control operation of the cooking apparatus. The server may use the cooking parameters in the MRR data to control various types of different cooking devices so that food is cooked in the cooking device. In addition, the MRR data may include first MRR data and second MRR data.

In the disclosed embodiment, the first MRR data may be generated from text representing a recipe using semantic elements identified using a plurality of NLU models, and various cooking devices read the MRR data in a structured text form, and perform a cooking step. Cooking operation can be performed by applying to

In the disclosed embodiment, the second MRR data refers to MRR data generated based on the first MRR data and user log data. The second MRR data may be data in which cooking parameters in the first MRR data are updated so that food can be cooked by reflecting the user's personal cooking tendency.

In the disclosed embodiment, the recipe card may be data in the form of a card that intuitively provides necessary information for each cooking step to a user who wants to cook food. The recipe card may correspond to each cooking step constituting a recipe, and may provide a user with various functions related to cooking and information related to cooking.

In the disclosed embodiment, the user's log data may be data related to an operation in which the user manipulated the cooking apparatus. The user's log data may include the temperature, cooking time, and number of times of cooking of the cooking device in which the user manipulated the cooking device, and may include cooking evaluation data input by the user for the finished dish.

In the disclosed embodiment, the additional information may be additional information generated to guide the cooking process based on semantic elements identified using a plurality of NLU models in the text representing the input recipe. For example, the additional information may include taste information that is changed when the cooking time is changed in a specific cooking step based on a semantic element related to the cooking time identified in the text representing the recipe. The generated additional information may be included in the recipe card and provided to the user when the recipe card is created, or may be inserted into the MRR data and stored in the MRR data.

1 is a diagram referenced to explain a method of a server providing food cooking information and controlling cooking apparatuses according to an exemplary embodiment.

The server 2000 according to an embodiment acquires the text 101 representing the recipe of food, uses a plurality of natural language understanding (NLU) models to identify semantic elements in the text, and various cooking devices ( 4000) may generate first MRR (Machine Readable Recipe) data 102 and may generate additional information for a cooking guide. In this case, the text 101 indicating the recipe of the food may be received from the client device 3000, but is not limited thereto, and the text 101 indicating the recipe of the food may be transmitted to another external device (eg, a third party). server) or may be pre-stored in a storage unit within the server.

In addition, the server 2000 according to an embodiment obtains log data of the user for the cooking apparatuses 4000 and updates the first MRR data 102 based on the obtained log data, thereby providing personalized to the user. The second MRR data 103 including the cooking method may be generated.

The server 2000 according to an embodiment generates a plurality of first recipe cards 104 for cooking food using the first MRR data 102 and additional information, and the generated plurality of first recipe cards ( 104 ) to the client device 3000 .

In addition, the server 2000 according to an embodiment generates a plurality of second recipe cards 105 for the user's personalized food cooking by using the second MRR data 103 and additional information, and the generated A plurality of second recipe cards 105 may be provided to the client device 3000 . In addition, when the cooking devices 4000 can display the first recipe card 104 or the second recipe card 105 including a display, the server 2000 generates the first recipe cards 104 . Alternatively, the second recipe cards 105 may be provided to the cooking devices 4000 .

The client device 3000 may receive a control command of the cooking apparatus 4000 from the user and transmit a control request of the cooking apparatus 4000 to the server 2000 . In response to the received control request, the server 2000 changes the first MRR data or the second MRR data to a command executable in each cooking device 4000 to match the interface of each cooking device 4000, A command for controlling each cooking device 4000 may be transmitted to each cooking device 4000 .

The cooking apparatus 4000 may perform a cooking operation based on control information for controlling the cooking apparatus received from the server.

2 is a block diagram illustrating a detailed configuration of a server according to an embodiment.

The server 2000 according to an embodiment may be configured to include at least a communication unit 2100 , a processor 2200 , and a storage unit 2300 .

The communication unit 2100 may transmit/receive data or signals to and from an external device (eg, the client device 3000 , the model server 3000 , the processing server 4000 , and the cooking device) under the control of the processor 2200 . .

The communication unit 2100 may transmit/receive data or signals to and from an external device (eg, the client device 3000 , a third-party server (not shown), or a cooking device (not shown)) under the control of the processor 2200 . .

The communication unit 2100 according to an embodiment includes a local area network (LAN), a wide area network (WAN), a value added network (VAN), a mobile radio communication network, and a satellite. It may include one or more components that allow communication through a communication network and a combination thereof.

The communication unit 2100 according to an embodiment may receive a text indicating a recipe from the client device 3000 . Also, in order to control the cooking apparatus, the communication unit 2100 may receive a cooking apparatus control command from the client device 3000 and transmit the cooking apparatus control commands generated based on the MRR data to the cooking apparatuses.

The processor 2200 may control the server 2000 as a whole. The processor 2200 may execute one or more instructions of a program stored in a memory. The processor 2200 may include hardware components that perform arithmetic, logic, input/output operations and signal processing.

The processor 2200 according to an embodiment includes, for example, a central processing unit (Central Processing Unit), a microprocessor, a graphic processing unit (Graphic Processing Unit), Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs). ), DSPDs (Digital Signal Processing Devices), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), APs (Application Processors), Neural Processing Units, or It may consist of at least one of the AI-only processors, but is not limited thereto.

The storage unit 2300 may store instructions, data structures, and program codes readable by the processor 2200 . In the disclosed embodiments, operations performed by the processor 2200 may be implemented by executing instructions or codes of a program stored in the storage 2300 .

The storage unit 2300 is, for example, a flash memory type, a hard disk type, a multimedia card micro type, or a card type memory (eg, SD or XD). Non-volatile memory including at least one of memory, etc.), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic disk, and optical disk and volatile memory such as random access memory (RAM) or static random access memory (SRAM).

The storage unit 2300 may store instructions, data structures, and program codes readable by the processor 2200 . In the disclosed embodiments, operations performed by the processor 2200 may be implemented by executing instructions or codes of a program stored in the storage 2300 . In an embodiment, the storage unit 2300 includes a recipe syntax analysis module 2310 , an additional information generation module 2320 , a first MRR data generation module 2330 , a second MRR data generation module 2340 , and a card recipe generation Data and program command codes corresponding to the module 2350 , the cooking device control module 2360 , and the database 2370 may be stored.

In one embodiment, the database 2370 may include text recipes, NLU models, user log data, card recipes, and MRR data.

The processor 2200 according to an embodiment receives the text representing the recipe, loads and executes a plurality of NLU models stored in the storage 2300 , and performs syntax analysis of the text representing the recipe, thereby first MRR data can create In addition, the processor 2200 obtains log data of a user related to an operation in which the user manipulated the cooking device, and generates second MRR data personalized to the user based on the user's log data and the first MRR data. can

The processor 2200 according to an embodiment executes a plurality of NLU models using the recipe syntax analysis module 2310 stored in the storage 2300, and a recognition target (eg, a cooking tool for each NLU model) , material) can identify semantic elements. For example, the processor 2200 may obtain a text indicating a recipe, apply the text to each of the plurality of NLU models, and identify at least one semantic element output from each of the plurality of NLU models.

The processor 2200 according to an embodiment may generate additional information for the cooking guide by using the additional information generating module 2320 stored in the storage 2300 . For example, the processor 2200 converts a time-related expression among the identified semantic elements into a numeric form based on the semantic element identified by the recipe parsing module, and when the cooking time applied for each cooking step increases or decreases Additional information capable of guiding changing taste information and the like may be generated. In another example, it is possible to identify an expression related to an amount among the identified semantic elements, and generate additional information that can guide change in taste information when the amount of cooking ingredients applied for each cooking step increases or decreases. have. In another example, it is possible to identify a temperature-related expression among the identified semantic elements, and generate additional information capable of guiding taste information that is changed when the cooking temperature is increased or decreased in each cooking step. In addition, the processor 2200 uses the additional information generation module 2320 to identify items such as cooking materials, cooking tools, and cooking devices from the semantic elements identified by the recipe syntax analysis module 2310 , and identify the identified items. You can search for advertisement information related to In addition, the processor 2200 uses the additional information generation module 2320 to identify expressions related to cooking techniques, plating methods, cooking tools, cooking materials, etc. among the semantic elements identified from the recipe syntax analysis module 2310, and , it is possible to search for media content such as video, image, and audio related to the identified expression.

The processor 2200 according to an embodiment generates first MRR (Machine Readable Recipe) data in a structured form including a plurality of cooking parameters by using the first MRR generating module 2330 stored in the storage 2300 . can create For example, the processor 2200 uses the first MRR generation module 2330 to generate entity data corresponding to the semantic elements identified by the recipe syntax analysis module 2310, and to The first MRR data may be generated by inserting the generated cooking parameter into the form of the first MRR data.

The processor 2200 according to an embodiment may generate second MRR data personalized to the user by changing the first MRR data using the second MRR generation module 2340 stored in the storage unit. The second MRR data generation module 2340 may include a log data acquisition module, a first MRR data extraction module, a reinforcement learning module, and an MRR data transformation module.

The processor 2200 acquires log data related to operations in which the user manipulated the cooking apparatus using the log data acquisition module, and cooks corresponding to the log data in the first MRR data using the first MRR data extraction module. parameters can be extracted. The processor 2200 may generate the second MRR data by updating the cooking parameter in the first MRR data based on the obtained log data using the MRR data transformation module.

In addition, the processor 2200 may perform reinforcement learning based on feedbacks on the operation of the user using the cooking apparatuses by using the reinforcement learning module in the second MRR generation module 2340 . In this case, log data of the user may be used when reinforcement learning is performed. The processor 2200 may generate the second MRR data personalized to the user by changing the first MRR data based on the result output from the reinforcement learning module by using the MRR data transformation module.

The processor 2200 according to an embodiment provides information for cooking food based on the generated first MRR data and the generated additional information using the card recipe generation module 2350 stored in the storage 2300 . A plurality of first recipe cards can be created. In addition, the processor 2200 may generate a plurality of second recipe cards for cooking personalized food to the user based on the generated second MRR data and the generated additional information.

3 is a flowchart illustrating a method in which a server operates to generate first MRR data and second MRR data according to an embodiment.

In step S310, the server 2000 may obtain a text indicating a recipe for food. The 'text representing a recipe' is data in the form of text representing a recipe for cooking food, and may be obtained in various ways.

For example, text representing a recipe may be obtained in the form of a recipe made of text in a document format usable by a computer, a recipe page on the Web, text extracted from an image of a recipe including text, etc. The present invention is not limited thereto. In addition, the text indicating the recipe is not necessarily obtained only from text, and may be obtained from other types of data such as images and moving pictures. For example, when a recipe including an image is received, the server 2000 may identify only a text portion in the recipe and use it to generate data for cooking.

In operation S320 , the server 2000 may identify a plurality of semantic elements included in the text by using a plurality of NLU models for analyzing the text. In this case, the server 2000 may determine a plurality of NLU models for parsing text. In an embodiment, the server 2000 may determine a plurality of NLU models to be used for syntax analysis based on the type of food indicated by the obtained text, and identify semantic elements using the determined NLU models. Each of the NLU models may be a model trained to recognize and extract semantic elements of a specific category.

However, the present invention is not limited thereto, and the server 2000 may identify semantic elements using a plurality of predetermined NLU models without determining a plurality of NLU models to be used for syntax analysis with respect to the obtained text.

When the server 2000 according to an embodiment determines a plurality of NLU models to perform syntax analysis of the obtained text, if the number of semantic element categories of a specific category to be identified is greater than or equal to a preset number, preference is given to the plurality of NLU models. It is possible to assign a rank, determine a plurality of NLU models based on the assigned priority, and identify semantic elements in the obtained text.

The server 2000 according to an embodiment applies the obtained text to a plurality of NLU models sequentially or in parallel, respectively, and performs text recognition for each NLU model, so that each NLU model is a semantic element to be identified. can be obtained.

For example, 'first semantic element' that is at least one semantic element output by applying the obtained text to the first NLU model, and 'second semantic element' that is at least one or more semantic elements output by applying the text to the second NLU model semantic element' can be obtained. The server 2000 may acquire the 'Nth semantic element' that is at least one output semantic element by applying the text to the Nth NLU model in the same manner.

In operation S330, the server 2000 according to an embodiment may generate first MRR (Machine Readable Recipe) data that is structured data based on the identified semantic elements. For example, the first MRR data may be text data in a structured form written in a JSON (JavaScript Object Notation) form, but is not limited thereto, and may be text data in a structured form implemented in another programming language or the like.

The first MRR data according to an embodiment may include a cooking parameter. The cooking parameter may be a parameter indicating a control operation of the cooking apparatus. The cooking parameter may include an operation mode, an operation temperature, an operation time of the cooking apparatus, and the like. For example, in the case of a recipe for oven cooking, the cooking parameter in the first MRR data may be a convection mode, 200° C., and 20 minutes.

In addition, the first MRR data may include basic recipe information and cooking step information. The basic recipe information may include information such as a dish name, a dish type, a recipe creation date, a recipe version, a language, a cooking time, a recipe author, and cooking ingredients. The cooking step information includes the original sentence of the text indicating the recipe, the user's cooking behavior required for each step, the user's action related to the cooking device, media content corresponding to the recipe, cooking ingredients required for each cooking step, cooking tools required for each cooking step, etc. may include

In operation S340 , the server 2000 according to an exemplary embodiment may acquire log data of the user related to operations in which the user manipulated the cooking apparatus. For example, the user's log data may be log data related to operations in which the user manipulated the cooking apparatus for a predetermined period.

In an embodiment, the user's log data may include parameters related to the control operation of the cooking apparatus, such as the operation temperature, operation time, and number of operations of the cooking apparatus, and may include evaluation data of the user on cooking. , but is not limited thereto.

According to an embodiment, the user's log data is received from the cooking apparatus to the server 2000, or the user's log data is received from the cooking apparatus to another external device (eg, a third-party server), and the server 2000 is It may be received from another external device.

In an embodiment, the user's log data may be the user's log data in which data related to operations in which the user manipulated the cooking apparatus are recorded.

In another embodiment, the log data of the user may be log data of the user related to the operation of the user to additionally manipulate the cooking apparatus differently from the cooking parameters included in the first MRR data. For example, when the cooking parameters included in the first MRR data are the convection mode of the oven, 200° C., and 20 minutes, and the user additionally manipulates the oven to cook for 5 minutes at the convection mode, 200° C., the convection mode, 200° C. , 5 minutes may be the user's log data.

In step S350 , the server 2000 according to an embodiment may generate the second MRR data personalized to the user by updating the plurality of cooking parameters included in the first MRR based on the obtained log data of the user. .

The server 2000 according to an exemplary embodiment determines, based on the user's log data, a plurality of cooking parameters included in the first MRR data when the number of times the user operates the cooking device in a specific mode of the cooking device is greater than or equal to a predetermined number of times. Values corresponding to the user's log data can be updated.

For example, when the user cooks pork, the cooking apparatus may complete the pork dish by performing cooking operations based on the first MRR data. In this case, the user of the cooking device may further cook the pork dish by further manipulating the cooking device (for example, heating the cooking device at 200° C. for 5 minutes more) in order to cook according to the user's preference, and the user's User log data according to the operation of the cooking device may be generated.

The server 2000 according to an embodiment changes the cooking parameters included in the first MRR data based on the user's log data (eg, 'convection mode, 200°C, 20 minutes', convection mode, 200° C., change to 25 minutes) to generate second MRR data representing a personalized recipe to the user.

Also, when generating the second MRR data, the server 2000 according to an embodiment may generate the second MRR data by using reinforcement learning.

The server 2000 according to an embodiment may generate the second MRR data by performing reinforcement learning using the reinforcement learning module. In this case, the server 2000 may allocate values included in the user's log data as compensation values. The server 2000 acquires compensation values corresponding to the operations of the cooking device in each cooking stage while the cooking device is operating, and performs the first operation such that the accumulated sum of the compensation values obtained until the cooking completion stage is maximized. By updating the cooking parameters in the MRR data, it is possible to generate second MRR data representing a personalized recipe to the user.

4 is a diagram for explaining operations of a recipe syntax analysis module included in a server according to an embodiment.

In the disclosed embodiment, the semantic element refers to data related to cooking of food recognized through the NLU model in the text representing the recipe. Semantic elements identified using the NLU model may be classified according to a category preset for cooking of food. For example, semantic elements may be classified according to categories such as a cooking device, a cooking technique, a cooking material, a cooking tool, a cooking time, and a cooking amount. When the semantic element according to the disclosed embodiment is changed, the cooking result of the food cooked in the cooking apparatus may be changed.

Also, the server 2000 may identify intent information determined to indicate the intention of the text from the semantic elements. For example, if the input text is 'Make a sauce by mixing 3 spoons of minced garlic, 2 spoons of soy sauce, 2 spoons of oyster sauce, 3 spoons of oligosaccharide, and 1 spoon of melted butter.' can When generating additional information to guide a cooking process based on the identified semantic element, the server 2000 may generate additional information using the identified intent information.

Referring to FIG. 4 , the server 2000 may obtain text indicating a recipe. The text representing the recipe may include various information related to a cooking method, such as a cooking technique, a cooking material, a cooking tool, a cooking device, and a cooking time. The server 2000 may determine a plurality of NLU models for syntax analysis in order to extract semantic elements from the obtained text. In this case, each NLU model may be an NLU model trained to identify a semantic element corresponding to a specific category, and the determined NLU models perform text recognition independently on the obtained text, and each NLU model independently performs text recognition. It is possible to identify a plurality of semantic elements output from the NLU model of .

However, the present invention is not limited thereto, and the server 2000 may identify a semantic element using a plurality of predetermined NLU models without determining a plurality of NLU models to be used for syntax analysis.

In an embodiment, the server 2000 identifies a cooking technique semantic element from the obtained text using a cooking technique NLU model, identifies a cooking ingredient semantic element using a cooking material NLU model, and uses a cooking tool NLU model to identify the cooking tool semantic element, the cooking device semantic element may be identified using the cooking device NLU model, and the cooking time semantic element may be identified using the cooking time NLU model.

A specific embodiment in which the server 2000 identifies semantic elements for each category in the obtained text will be further described with reference to FIG. 5 .

5 is a diagram referenced to explain a method in which a server identifies semantic elements using a plurality of NLU models using a recipe syntax analysis module according to an embodiment.

In an embodiment, the server 2000 may obtain text indicating recipes for various foods. Hereinafter, for convenience of explanation, a case in which the text representing the recipe obtained from the server 2000 is the text representing the 'chicken dish recipe' will be described as an example.

The server 2000 selects a plurality of NLU models to be used to analyze the text from among a plurality of NLU models stored in the storage with respect to the text 510 indicating the acquired 'chicken dish recipe', a cooking technique NLU model, and a cooking material. It may be determined by the NLU model, the cooking tool NLU model, the cooking apparatus NLU model, and the cooking time NLU model. The server 2000 may identify semantic elements for each category corresponding to each NLU model by using the determined NLU models.

For example, the server 2000 may identify at least one semantic element related to the cooking technique by applying the text 510 representing the 'chicken cooking recipe' to the cooking technique NLU model. Specifically, the server 2000, from the text 510 indicating the chicken recipe, 'to remove the smell', 'soak for about 30 minutes and then rinse with running water to remove the water', 'mix well to make the sauce', ' Meaning factors of cooking technique that can change the cooking result according to the user's cooking technique during the cooking process, such as 'sliced chicken', 'add 2/3 of the seasoning and toss', 'marinate for 30 minutes', etc. (520) can be identified.

In another example, the server 2000 may identify at least one semantic element related to the cooking material by applying the text 510 representing the 'chicken dish recipe' to the cooking material NLU model. Specifically, the server 2000, from the text 510 indicating the chicken recipe, 'chicken', 'milk or rice water', '3 spoons minced garlic', '2 spoons soy sauce', '2 spoons oyster sauce', During the cooking process, such as '3 spoons of oligosaccharide', '1 spoon of melted butter', and 'parsley powder', it is possible to identify the elements 530 meaning cooking ingredients that can cause a change in the cooking result according to the change of the cooking material. .

In another example, the server 2000 may identify at least one semantic element related to the cooking tool by applying the text 510 representing the 'chicken cooking recipe' to the cooking tool NLU model. Specifically, the server 2000 identifies, from the text 510 representing the chicken cooking recipe, the cooking tool semantic elements 540 that can cause a change in the cooking result according to the change of the cooking tool, such as 'oven pan'. can

In the same way, the server 2000 applies the text 510 indicating 'chicken cooking recipe' to each of the cooking apparatus NLU model and the cooking time NLU model, and the cooking apparatus semantic element 650, which is output for each model, and cooking A temporal semantic element 660 may be identified.

The server 2000 according to an embodiment may generate the first MRR data and additional information using semantic elements obtained for each category by applying each NLU model to the text representing the recipe.

6 is a diagram for describing a detailed configuration of an additional information generating module that generates additional information based on an identified semantic element included in a server according to an exemplary embodiment.

The server 2000 according to an embodiment may generate additional information for guiding food cooking by using the additional information generating module 2320 . In this case, the additional information generating module 2320 may include a cooking guide information generating module 610 , an advertisement information obtaining module 620 , a content information obtaining module 630 , and an alternative recipe information obtaining module 640 . .

In an embodiment, the server 2000 may generate cooking guide information based on the semantic element identified in the text representing the recipe.

For example, the server 2000 uses the cooking guide information generating module 610 to convert a time-related expression among the identified semantic elements into a numeric form, and changes when the cooking time applied for each cooking step increases or decreases. It is possible to create additional information that can guide the taste information and the like. In another example, it is possible to identify an expression related to an amount among the identified semantic elements, and generate additional information that can guide change in taste information when the amount of cooking ingredients applied for each cooking step increases or decreases. have. In another example, it is possible to identify a temperature-related expression among the identified semantic elements, and generate additional information capable of guiding taste information that is changed when the cooking temperature is increased or decreased in each cooking step.

In an embodiment, the server 2000 may acquire advertisement information based on the semantic element identified in the text representing the recipe.

For example, the server 2000 uses the advertisement information obtaining module 620 to identify items such as cooking materials, cooking tools, and cooking devices from the identified semantic elements, and retrieve advertisement information related to the identified items. can do. In addition, the server 2000 may use the advertisement information acquisition module 620 to identify items such as cooking materials, cooking tools, and cooking devices from the MRR data, and search advertisement information related to the identified items. The advertisement information obtaining module 620 may generate additional information capable of matching the searched advertisement information with a corresponding semantic element and providing advertisement information for each cooking step.

In an embodiment, the server 2000 may acquire content information based on a semantic element identified in the text representing the recipe.

For example, the server 2000 may identify an expression related to a cooking technique, a plating method, a cooking tool, a cooking material, etc. from among the identified semantic elements by using the content information acquisition module 630 . The server 2000 may use the content information obtaining module 630 to search for media content such as video, image, and audio related to the above-described identified expressions. The server 2000 may generate additional information capable of matching the searched content with a corresponding semantic element and providing media content for each cooking stage.

In an embodiment, the server 2000 may obtain the replacement recipe information based on the semantic element identified in the text representing the recipe and the user profile information obtained from the user.

For example, the server 2000 may use the alternative recipe information obtaining module 640 based on the identified recipe information and user profile information (eg, a cooking device possessed by the user, a taste preference of the user, etc.) Accordingly, it is possible to search for a method capable of cooking in a method other than the cooking method written in the currently identified recipe. The server 2000 may generate, as additional information, a recipe in which at least a portion of the current recipe is modified or an alternative recipe of the current recipe, based on the found cooking method.

Although an embodiment of generating additional information for a cooking guide has been described in the description of FIG. 6 , it should be understood that the description is exemplarily described for convenience of explanation, and in addition to the above-described additional information, a cooking guide is used. Various additional information modules may be included for

When generating a recipe card using the card recipe generating module, the server 2000 according to an embodiment may generate additional information corresponding to each cooking step by being included in the recipe card. The server 2000 may provide a recipe card including additional information to the user. Also, the server 2000 may insert additional information into the MRR data when generating the MRR data and store it in the MRR data.

7 is a diagram referenced to explain a method for a server to generate first MRR data using semantic elements identified from a plurality of NLU models according to an embodiment.

Referring to FIG. 7 , the server 2000 may identify a plurality of semantic elements 710 from text representing a recipe by using a plurality of NLU models. Since the method for the server 2000 to identify the plurality of semantic elements 710 has been described above with reference to FIG. 5 , a description thereof will be omitted here for the sake of simplicity.

In an embodiment, the first MRR data 730 is data generated based on the identified semantic elements, and may be data in a structured form so that the cooking apparatus can read it.

The server 2000 according to an embodiment may generate a cooking parameter 720 corresponding to the identified semantic elements 710 by using the first MRR data generation module. In this case, the cooking parameter 720 is each entity included in the first MRR data, and may represent elements related to cooking in a recipe.

In an embodiment, the semantic element may be identified in the form of a sentence. In this case, the server 2000 may extract text that becomes a core keyword from the semantic element by using the first MRR data generation module, and may generate entity data to be included in the first MRR data 730 . For example, the server 2000 extracts 'Stir' as a keyword from 'mix ingredients in a bowl' of the cooking technique semantic element, and generates the extracted 'Stir' as a cooking parameter corresponding to the cooking technique category 721 . can do. In addition, the server 2000 extracts 'Wait' as a keyword from 'Wait for 5 minutes when finished' of the cooking technique semantic element, and generates the extracted 'Wait' as a cooking parameter corresponding to the cooking technique category 721. have.

In an embodiment, the semantic element may be identified in the form of a word. In this case, the server 2000 may generate a cooking parameter to be included in the first MRR data by extracting the semantic element identified in the form of a word as a keyword using the first MRR data generating module. For example, the server 2000 extracts 'Banana 3-4 ea' as a keyword from '3 to 4 bananas' of the cooking ingredient semantic element, and sets the extracted 'Banana 3-4 ea' to the cooking ingredient category 722 ) can be created as a cooking parameter corresponding to In addition, the server 2000 extracts 'Egg 2ea' as a keyword from '2 eggs' of the cooking ingredient semantic element, and generates the extracted 'Egg 2ea' as a cooking parameter corresponding to the cooking ingredient category 722. have.

In an embodiment, the identified semantic element may be a semantic element identified as a common pronoun (eg, 'bowl') due to the nature of the text recipe. In this case, the server 2000 may generate a cooking parameter to be included in the first MRR data by changing the corresponding semantic element to a suitable keyword using the first MRR data generating module. For example, the server 2000 uses the first MRR data generation module to bake 'bowl', which is a semantic element of the identified cooking tool category 723, and 'bowl,' which is another identified semantic element, in an oven at 200 degrees for 10 minutes. ', an 'oven pan', which is a vessel suitable for an oven, may be created as a cooking parameter corresponding to the cooking tool category 723 .

In an embodiment, when the extracted semantic element includes a plurality of keywords necessary for cooking, the server 2000 may generate a cooking parameter in a form in which the plurality of keywords are combined.

For example, the server 2000 extracts two keywords 'preheat the oven to 200 degrees Celsius' from the semantic element of the cooking device category 724, '200ºC and preheat the oven', to the first MRR data 730. You can create cooking parameters to be included. In another example, the server 2000 may be included in the first MRR data 730 by extracting two keywords '200ºC and heating' from 'bake an oven at 200 degrees for 10 minutes', which is a semantic element of the cooking device category 724 . You can create cooking parameters.

In addition, in the case of a semantic element of the cooking time category 725 , the server 2000 may generate a cooking parameter by extracting a keyword indicating which cooking operation is included. For example, the server 2000 may extract two keywords '10min, Heat' from 'bake for 10 minutes', which is a semantic element of the cooking time category 725 , and generate a cooking parameter. In another example, the server 2000 generates cooking parameters to be included in the first MRR data 730 by extracting two keywords '5min, Wait' from 'wait for 5 minutes', which is a semantic element of the cooking time category 725 . can do.

In an embodiment, among the cooking parameters included in the first MRR data 730 , the cooking parameters of the cooking device category 724 and the cooking time category 725 may be related to a control operation of the cooking device.

The server 2000 generates a cooking parameter 720 to be included in the first MRR data 730 from the identified semantic element 710 by using the first MRR data generation module, and generates the first MRR data in a structured form. The first MRR data 730 can be generated by inserting the generated cooking parameters 720 into the form. 7 illustrates the first MRR data written in JSON form for convenience of explanation, but is not limited thereto, and the first MRR data 730 may be implemented in various programming languages and generated in a structured form.

The first MRR data 730 according to an embodiment is name, cuisine, dataPublished, version, language, cookTime, image, author which is basic recipe information for management of recipe information (eg, storing and searching by category). , ingredient information, etc. Basic recipe information may be extracted from text indicating a recipe obtained by the server 2000, and if there is no corresponding information, it is left blank and later updated through information received from a user, information input from an external device, etc. have.

8 is a server according to an embodiment, by using the first MRR data generation module, identifying semantic elements from text representing a recipe, and generating a cooking parameter to generate first MRR data that can read the recipe in the cooking device It is a drawing referenced to explain the method.

The server 2000 may store cooking parameters generated from the semantic elements by substituting them into the structured area using the first MRR generation module. In FIG. 8 , first MRR data 810 corresponding to a preheating step among a plurality of cooking steps included in a recipe is illustrated for convenience of explanation.

Referring to FIG. 8 , the first MRR data may include cooking step information for each cooking step. For example, in the case of the first MRR data 810 corresponding to the oven preheating step, cooking parameters generated from semantic elements may be structured and configured into a plurality of data areas.

For example, the plurality of data areas include an instruction area 811 representing the original sentence of a recipe, an actions area 812 representing data determined as a cooking operation, and data determined as an expression related to a cooking device. It may be a data area divided into a device area 813 indicating a device area 813 and a media area 814 indicating content (eg, video, image, etc.) that is additional information that can be matched with a cooking step.

In this case, among the data areas included in the first MRR data, the actions 812 area includes an ingredients area 815 related to cooking materials and a tools area 816 related to cooking tools. may be included, but is not limited thereto.

In one embodiment, among the data areas included in the first MRR data 810 corresponding to the oven preheating step, the actions area 812 is a sub-structure, 'cooking operation, start, end, material, tool It may include cooking parameters indicating ' and the like. In addition, the device area 813 form is a sub-structure, and may include cooking parameters indicating 'cooking device type, cooking temperature', and the like. Also, the media area 814 is a sub-structure and may include additional information data such as 'content type, content url'. In this case, the additional information data may be generated by the additional information generating module.

The server 2000 according to an embodiment may generate the first MRR data by inserting a cooking parameter into each data area using the first MRR data generating module.

9 is a diagram for describing a method in which a server acquires user's log data and generates second MRR data based on the acquired log data, according to an embodiment.

Referring to FIG. 9 , the second MRR data generation module 2340 according to an embodiment includes at least a log data acquisition module 910 , a first MRR data extraction module 920 , a reinforcement learning module 930 , and MRR data transformation module 940 .

The server 2000 according to an embodiment may acquire the user's log data by using the log data acquisition module 910 . The user's log data may include parameters related to the control operation of the cooking apparatus, such as the operation temperature, operation time, and number of operations of the cooking apparatus, and may include the user's evaluation of the finished dish.

In an embodiment, the user's log data may be generated by the cooking device and provided to the server 2000 . For example, the user's log data may be generated for all operations in which the user operated the cooking apparatus, provided to the server 2000, and stored in a database of the server. In another example, the user's log data may be stored in the database of the server 2000 after the log data is generated for operations in which the user manipulated the cooking apparatus differently from the cooking parameters included in the first MRR data.

According to the above-described embodiments, the server 2000 may identify semantic elements in the text 950 indicating the recipe and generate the first MRR data.

Also, the server 2000 may control the cooking apparatus based on the first MRR data. In this case, the cooking apparatus may perform a control operation based on the cooking parameter included in the first MRR data.

For example, the server 2000 may provide control information to the cooking apparatus based on the cooking parameters generated in the cooking step 960 of 'bake in an oven at 200 degrees for 10 minutes'. The cooking apparatus may perform a control operation of 'heat to 200ºC, 10 minutes' based on the received control information. In this case, the user may input an operation for additionally controlling the cooking apparatus. Specifically, when the first MRR data is generated based on a microwave oven having an output of 1000 watts (W), and the user's microwave oven has an output of 700 watts (W), the user uses 700 watts for additional cooking (W) It is possible to additionally input the heating operation of the microwave oven. In addition, when the user prefers crispy food that is further heated according to the user's preference, an operation of heating the oven at 200 degrees for 10 minutes after the cooking operation of the oven heated to 200 degrees for 4 minutes may be additionally input.

The server 2000 according to an exemplary embodiment may collect log data of a user related to an operation of controlling the user's cooking apparatus and store the data in a database of the server. Also, the server 2000 may use the log data acquisition module 910 to acquire stored user log data, and generate second MRR data 970 based on the acquired log data.

In an embodiment, the server 2000 may extract a cooking parameter related to a control operation of the cooking apparatus from the first MRR data using the first MRR data extraction module 920 . In this case, the server 2000 may generate the second MRR data 970 by updating the cooking parameter in the first MRR data using the MRR data transformation module 940 .

Specifically, the server 2000 uses the first MRR data extraction module 920 to extract the cooking parameters generated from the cooking step 960 of 'bake in the oven at 200 degrees for 10 minutes' in the first MRR data. can In this case, the extracted cooking parameters may indicate 'heat to 200ºC, 10 minutes'. Also, the user's log data may be 'additional heating to 200ºC, 4 minutes'. The server 2000 uses the MRR data transformation module 940 to update the above-mentioned cooking parameter 'heat to 200ºC, 10 minutes' to 'heat to 200ºC, 14 minutes' in the first MRR data, thereby personalized to the user. It is possible to generate the second MRR data 970 that is the MRR data.

Various methods may be applied to a method in which the server 2000 generates the second MRR data 970 by updating the first MRR data according to an embodiment.

For example, when the number of times the user operates the cooking apparatus in a specific mode of the cooking apparatus is greater than or equal to a predetermined number of times, the server 2000 sets the plurality of cooking parameters included in the first MRR data to the user's Values corresponding to log data can be updated.

In another example, the server 2000 may generate the second MRR data 970 by performing reinforcement learning using the user's log data using the reinforcement learning module. This will be described in detail in the description of FIG. 10 .

10 is a diagram for explaining an embodiment in which reinforcement learning is performed based on user log data when the server 2000 generates second MRR data according to an embodiment.

Referring to FIG. 10 , the server 2000 according to an embodiment may provide control information for controlling the cooking device to the cooking device based on first MRR data. As the server 2000 provides control information to the cooking apparatus, the cooking apparatus may perform control operations.

In an embodiment, the server 2000 may obtain compensation values for control operations of the cooking apparatus corresponding to the control information. In this case, the reward values may be determined based on the user's log data. For example, the compensation values may be determined based on cooking parameters related to control operations of the cooking device in which the user manipulated the cooking device, and specifically, the operation mode, operating temperature, and operating time of the cooking device operated by the user. , the number of operations, etc. may be determined as compensation values. In this case, each of the compensation values may have a positive value or a negative value.

Specifically, in one embodiment, the user's log data may include a cooking device operation history of 'additional heating to 200 ºC, 4 minutes'. In this case, in order to perform reinforcement learning so that the cooking device performs a control operation of 'additional heating to 200ºC, 4 minutes', the reward value may be determined as a positive value. In another embodiment, the log data of the user includes the cooking device operation history of 'additional heating to 200ºC, 4 minutes', but may further include a negative evaluation of the user for a corresponding control operation or finished food. In this case, in order to proceed with reinforcement learning so that the cooking device does not perform the control operation of 'additional heating to 200ºC, 4 minutes', the compensation value may be determined as a negative value.

The server 2000 according to an embodiment may obtain compensation values for control operations of the cooking apparatus corresponding to the control information.

For example, the server 2000 provides the first control information to the cooking apparatus, and as the first control operation 1010 is performed in the cooking apparatus, a first compensation value corresponding to the first control operation 1010 is provided. can be obtained Specifically, when the first control operation 1010 is performed to match the user's log data within a predetermined error range, the server 2000 may acquire the first compensation value.

In this case, the first control operation 1010 includes not only the control operation performed by the cooking apparatus receiving the first control information from the server, but also first input data of the user 1000 operating the cooking apparatus while the cooking apparatus is operating. It may include receiving and additionally operating. Also, the first input data received from the user 1000 while the cooking apparatus is operating may be collected as log data of the user and stored in a database within the server 2000 . The server 2000 may allocate the first input data, which is the stored log data, as a compensation value.

In addition, the server 2000 provides the second control information to the cooking apparatus, and as the second control operation 1020 is performed by the cooking apparatus, a second compensation value corresponding to the second control operation 1020 may be obtained. can Specifically, when the second control operation 1020 in the cooking apparatus is performed to match the user's log data within a predetermined error range, the server 2000 may obtain the second compensation value.

In this case, the second control operation 1020 includes not only the control operation performed by the cooking apparatus receiving the second control information from the server, but also input data of the user in which the user 1000 operates the cooking apparatus while the cooking apparatus is operating. It may include receiving and additionally operating. Also, referring to FIG. 10 , user input data is not received for the second control operation 1020 , and the second compensation value may be obtained as the control operation is performed in the cooking apparatus.

In the same way, the server 2000 may provide the N-th control information to the cooking apparatus, and as the N-th control operation is performed by the cooking apparatus, the server 2000 may obtain an N-th compensation value for the N-th control operation.

The server 2000 according to an embodiment may receive evaluation data of a user on food cooked by an operation of the cooking apparatus. In this case, the server 2000 may receive the user's evaluation data from the user's client device 3000 , or the user 1000 may input the evaluation data into the cooking apparatus and receive the user's evaluation data from the cooking apparatus. .

In addition, evaluation data for cooked food received from the user 1000 may be collected as log data of the user and stored in a database within the server 2000 . The server 2000 may allocate user evaluation data, which is stored log data, as a reward value.

The server 2000 according to an embodiment updates the second MRR data by updating the cooking parameters included in the first MRR data based on the accumulated sum of the compensation values obtained in the control operations corresponding to the cooking steps of the food. can create For example, the server 2000 may update the cooking parameters included in the first MRR data so that the accumulated sum of the obtained compensation values becomes the maximum value, but is not limited thereto.

A method in which the server 2000 generates second MRR data by updating cooking parameters included in the first MRR data according to an exemplary embodiment has been described above in the exemplary embodiment of FIG. 9 , and thus a description thereof will be omitted.

11 is a diagram referenced to explain a method for a server to generate a recipe card according to an embodiment.

The server 2000 according to an embodiment may generate a plurality of first recipe cards by using the generated first MRR data and the generated additional information. The generated first recipe card may be provided to the client device (3000).

In the description of FIG. 11 , the first MRR data 1110 corresponding to the preheating step among the plurality of cooking steps included in the recipe described in FIG. 8 will be described as an example.

Referring to FIG. 11 , the first MRR data 1110 corresponding to the preheating stage may include 'preheat', which is information on a cooking operation, 'oven', '200ºC', etc., which are information about a cooking device. In addition, additional information related to the cooking step (eg, a cooking video provided in the preheating step) may exist.

The server 2000 according to an embodiment may generate the first recipe card 1120 corresponding to the above-described preheating cooking step by using the first MRR data and additional information.

The first recipe card 1120 generated using the first MRR data 1110 includes summary information 1121 representing a representative summary of the current cooking stage, and additional information corresponding to the current cooking stage (eg, an image). , image) may include content information 1122 indicating content related to the current cooking stage, text sentence information 1123 indicating the original recipe indicating the current cooking stage, and information 1124 indicating functions that can be linked in the current cooking stage.

Summary information 1121 of the first recipe card 1120 according to an embodiment, the server 2000 interprets the text using the NLU model, intent information determined to indicate the intention of the text It can be created based on what is obtained. For example, when the text for the cooking step is 'Preheat the oven to 200 degrees', the intent information determined as the intention of the text may be 'Preheat'. Although not shown in the drawing, in another example, the intent information determined as the intention of the corresponding text may be a 'cooking preparation stage'. When the first recipe card 1120 is generated, the server 2000 may determine summary information 1121 to be included in the first recipe card 1120 by using intent information.

Content information 1122 of the first recipe card 1120 according to an embodiment may be obtained by searching for images, videos, etc. related to the current cooking stage, or may be obtained together when obtaining text indicating the recipe. have. For example, when the text for the cooking step is 'Preheat the oven to 200 degrees', the server 2000 may obtain an image for guiding the oven preheating operation based on the semantic element identified from the text. In another example, the server 2000 may acquire an image or an image for guiding matters necessary for cooking preparation based on the 'cooking preparation stage', which is intent information determined as the intention of the text. The server 2000 may determine the content information 1122 to be included in the first recipe card 1120 when the first recipe card is generated.

Recipe sentence information 1123 of the first recipe card 1120 according to an embodiment may be a recipe original sentence based on generating the first recipe card.

In an embodiment, when the server 2000 recognizes the text using the NLU model, and as a result of recognizing the text, the recipe sentence for the corresponding cooking step is long and inappropriate to display on the first recipe card 1120 , the semantic element identified in the text Based on these, the sentences can be divided and displayed.

In another embodiment, when the recipe sentences are short and the meaning indicating the recipe step is not formed in each sentence unit, the server 2000 based on the semantic elements identified in the text, the first recipe card by generating the recipe sentence Recipe sentence information 1123 to be included in 1120 may be determined.

Information 1124 indicating an interoperable function of the first recipe card 1120 according to an embodiment may be generated by the server using additional information generated based on semantic elements identified in text.

In an embodiment, the information 1124 indicating an interlockable function may provide advertisements and purchases related to a control interlocking function capable of controlling a cooking device operating in the current cooking stage, cooking tools and cooking materials required for the cooking stage, and the like. Advertisement/purchase link function that can be used, content providing alternative cooking method content that can replace the cooking method of the corresponding stage, taste information that changes when the cooking time applied for each cooking step increases or decreases It may include an interlocking function for providing guide information that can guide the result of a change, and the like.

In addition, the server 2000 according to an embodiment may generate a second recipe card including cooking information personalized to the user by using the second MRR data and additional information. This is the same as the method of generating the first recipe card 1120 described above, and thus a description thereof will be omitted.

The user of the client device is provided with the first recipe card 1120 or the second recipe card according to an embodiment, and is provided with the interaction with the server and the cooking apparatus through the first recipe card 1120 or the second recipe card. can

For example, the user of the client device controls the cooking apparatus or a cooking tool required for a cooking step by clicking an icon indicating information 1124 indicating an interlockable function of the first recipe card 1120 according to an embodiment. It is possible to purchase, etc., receive content related to cooking, or receive information guiding a cooking method.

12 is a diagram referenced to explain a method for a server to generate a recipe card for each cooking step according to an embodiment.

12 , the recipe card may be a first recipe card generated from the first MRR data, and may be a second recipe card including cooking information personalized to the user generated from the second MRR data. For convenience of explanation, hereinafter, both the first recipe card and the second recipe card will be referred to as a 'recipe card'.

Referring to FIG. 12 , the server 2000 may generate a recipe card for each cooking step for a 'banana bread recipe' by using the generated MRR data and the generated additional information. In this case, the cooking step of the 'banana bread recipe' may consist of first to fourth steps.

The server 2000 may determine the type of the cooking step in the recipe by using the identified semantic elements. The server 2000 may determine the type of recipe card corresponding to each cooking stage based on the determined type of cooking stage, MRR data, and additional information. In addition, one recipe card is not necessarily limited to one type, and one recipe card may include a plurality of types.

For example, the determined type of recipe card may be a cooking device control type. In an embodiment, the first cooking step corresponding to the recipe card 1210 representing the first cooking step is 'preheating', and the third cooking step corresponding to the recipe card 1230 representing the third cooking step is 'heating'. ', the first cooking step and the third cooking step may require control of the oven, which is a cooking device. In this case, the recipe card corresponding to each of the first cooking step and the third cooking step may include cooking device control buttons 1215 and 1235 . The server 2000 may receive a cooking apparatus control request according to an input of a user selecting the cooking apparatus control buttons 1215 and 1235 from the client device 3000 . When receiving a cooking device control request, the server 2000 according to an embodiment may provide cooking device control information included in MRR data to the cooking device for controlling the cooking device.

In another example, the determined type of recipe card may be a cooking item purchase type. In an embodiment, the recipe card 1220 indicating the second cooking stage may include a cooking item purchase button 1225 . In this case, the server 2000 may receive a purchase information provision request according to an input of the user selecting the cooking item purchase button 1225 from the client device 3000 . Cooking items that can be purchased from the cooking item purchase button 1225 may include various items related to cooking of a corresponding recipe, such as cooking materials and cooking tools. When generating the recipe card, the server 2000 may include additional information, such as advertisement information and purchase information, in the recipe card in relation to the cooking item included in the cooking step. When the server 2000 receives a request for providing cooking item purchase information, the server 2000 may provide the cooking item purchase information to the client device 3000 .

In another example, the determined type of recipe card may be a cooking information provision type. In an embodiment, the recipe card 1220 indicating the second cooking stage and the recipe card 1240 indicating the fourth cooking stage may include cooking information providing buttons 1227 and 1245, respectively. In this case, the server 2000 may receive a cooking information provision request according to an input of the user selecting the cooking information providing button 1227 from the client device 3000 .

The additional information related to the provision of cooking information may include an image of a cooking technique, photo information, sub-recipe information derived from a corresponding recipe, cooking time information, plating information on a finished dish, and the like. The server 2000 may provide, to the client device 3000 , cooking information, etc. on a cooking step corresponding to the recipe card according to an input of the user selecting the cooking information providing button 1245 from the client device 3000 . .

In the above embodiments, the cooking device control type, the cooking item purchase type, and the cooking information provision type have been described as the types of recipe cards, but these are only examples, and various functions related to cooking and cooking using MRR data and additional information are provided. The type of recipe card can be defined to provide relevant information.

13 is a diagram referenced to explain a method for optimizing by changing the MRR data generated by the server according to the user's cooking apparatus according to an embodiment.

In describing FIG. 13 , the MRR data may be the first MRR data according to the above-described embodiments or the second MRR data personalized to the user. Hereinafter, for convenience of description, both the first MRR data and the second MRR data will be referred to as 'MRR data'.

When the generated MRR data is not suitable for the user's cooking apparatus, the server 2000 may change the MRR data so that food can be cooked by another user's cooking apparatus.

For example, in the cooking stage, MRR data 1310 generated for 'bake in an oven at 200 degrees for 10 minutes' may be generated based on the oven. In this case, the server 2000 may receive identification information for identifying the user's cooking apparatus from the client device 3000 .

When the received user's cooking device identification information is a 2000W microwave oven, the server 2000 changes 'cooking for 10 minutes in an oven at 200 degrees for 10 minutes', which is data representing a cooking operation among MRR data, to 'cooking for 10 minutes in a 2000W microwave oven' Thus, it is possible to generate the MRR data 1320 for a 2000W microwave oven.

In addition, when the received user's cooking device identification information is a 700W microwave oven, the server 2000 converts 'cooking for 10 minutes at 200 degrees in an oven at 200 degrees for 10 minutes', which is data representing the corresponding operation among MRR data, to 'cooking for 15 minutes in a 700W microwave oven'. By changing to , it is possible to generate the MRR data 1330 for a 700W microwave oven.

In another example (not shown), in the MRR data generated for the cooking step 'warm in lukewarm water for 2 days (Sous vide)', the sous vide cooking method is a function available only in a specific cooking device, and the user has This may be an unsupported feature. In this case, the server 2000 may receive identification information for identifying the user's cooking apparatus from the client device 3000 .

Based on the received user's cooking device identification information, the server 2000 may change the MRR data to a function related to the user's cooking device that can replace the Sous vide cooking method.

In another example (not shown), the MRR data generated for ‘mixing ingredients in a bowl’ of the cooking step may be a function provided by a specific cooking device, but may be a function not provided by other cooking devices. In this case, the server 2000 may receive identification information for identifying the user's cooking apparatus from the client device 3000 .

When the function of the cooking device indicated by the MRR data is a function that cannot be performed by the user's cooking device based on the received user's cooking device identification information, the server 2000 prepares a recipe card corresponding to the MRR data by the user directly. It can be provided by changing the content to guide you to mix the ingredients.

The server 2000 may obtain identification information of the other cooking apparatus and change at least a part of the MRR data so that food can be cooked by the other cooking apparatus based on the identification information of the other cooking apparatus.

14 is a diagram referenced to explain a method for a server to change and optimize MRR data generated by a server based on various user information according to an embodiment.

14, the MRR data may be the first MRR data according to the above-described embodiments or the second MRR data personalized to the user. Hereinafter, for convenience of description, both the first MRR data and the second MRR data will be referred to as 'MRR data'.

Referring to FIG. 14 , the server 2000 may obtain a text indicating a recipe for 'whole pork belly pasta' and generate MRR data according to the above-described embodiments.

In an embodiment, the server 2000 may receive the characteristic information about the characteristic of the food input by the user from the client device 3000 . The food characteristic information input by the user may include taste characteristics of food (eg, spiciness, sweetness or saltiness), baking characteristics of food (eg, rare, medium rare), etc. The present invention is not limited thereto and may include various information that can be changed by a user in relation to cooking of food.

The server 2000 may change at least a portion of the MRR data generated from the text indicating the recipe of 'whole pork belly pasta' based on the received characteristic information.

For example, the server 2000 may change the MRR data based on the user's taste preference included in the characteristic information received from the user. When the user prefers spicy food, the server 2000 converts MRR data indicating '3 to 5 pepperoncino, 1 teaspoon of red pepper powder', etc. related to spiciness among the MRR data to '5 to 8 pepperoncino, red pepper powder'. 2 teaspoons' data can be changed.

In another embodiment, the server 2000 may acquire information on ingredients that the user has. The food ingredient information held by the user is a smart home appliance (eg, a refrigerator) capable of receiving ingredient information input by the user from the client device 3000 to the server 2000 or performing data communication with the server 2000 . ) may be received from the server 2000 or from another server (eg, a cloud server) to the server 2000 .

The server 2000 may change at least a portion of the first MRR data generated from the text indicating the recipe of 'whole pork belly pasta' based on the ingredient information held by the user.

For example, the server 2000 does not need to purchase new ingredients, and when it is possible to cook alternatively with ingredients owned by the user, the server 2000 converts MRR data indicating 'peperoncino' among MRR data to pepperoncino. It can be changed to data of 'Cheongyang pepper' that can be replaced.

However, embodiments in which the server 2000 changes the MRR data are not limited thereto, and the server 2000 changes the MRR data using various user information that can be applied to a recipe, thereby changing the MRR data to a user tendency. It can be optimized according to /state/situation, etc.

For example, the server 2000 may change the MRR data based on the user's country/region information. In the country/region where the user is located, when it is difficult to purchase 'peperoncino', a material required in the recipe, the cooking information providing server 2000 collects data related to 'peperoncino' from among the MRR data. It can be changed to data related to 'Cheongyang pepper' that can be replaced.

In another example, the server 2000 may change the MRR data based on the user's health information. When the user is on a diet, the server 2000 may change data related to 'whole pork belly' among the MRR data to data related to 'chicken breast'.

In another example, the server 2000 may change the MRR data generated by using the cooking device information preferred by the user. When the user's preferred cooking device is an 'air fryer', the server 2000 may change data related to the operation of the 'oven' among the MRR data into data related to the operation of the 'air fryer'.

15 is a diagram for explaining the configuration of a cooking information providing system according to an exemplary embodiment.

Referring to FIG. 15 , a cooking apparatus system according to an embodiment may include a server 2000 , an IOT cloud server 1500 , a client device 3000 , and user's cooking-related apparatuses 4000 .

The server 2000 according to an embodiment may obtain text indicating a recipe from the client device 3000 or an external source, and analyze the text to generate first MRR data and a first recipe card.

The server 2000 may transmit the first recipe card to the client device 3000 to provide cooking information to the user. Also, the server 2000 may control the user's cooking-related devices 4000 using the generated first MRR data.

Also, the server 2000 may obtain log data of the user related to the operation of the user's cooking-related apparatuses 4000 from the user's cooking-related apparatuses 4000 . The server 2000 may generate second MRR data personalized to the user by changing at least a portion of the first MRR data based on the user's log data. Also, the server 2000 may generate a second recipe card capable of providing personalized cooking information to the user based on the second MRR data.

The server 2000 may transmit the second recipe card to the client device 3000 to provide cooking information to the user, and use the second MRR data to control the user's cooking-related apparatuses.

Operations of the server 2000 according to an embodiment may be performed using the IOT cloud server 1500 . For example, the IOT cloud server 1500 may provide a function of registering the user's cooking-related devices 4000 to the user's account and managing the user's cooking-related devices 4000 . The IOT cloud server 1500 may collect log data from the user's cooking-related devices 4000 and transmit the user's account information including the collected log data to the server 2000 .

In addition, the IOT cloud server 1500 receives MRR data (eg, first MRR data or second MRR data, etc.) from the server 2000, and based on the received MRR data, the cooking device registered in the user's account can control them.

FIG. 16 is a diagram for explaining an example in which information for cooking food is provided to a user through a client device after MRR data and recipe cards are generated in a server according to an embodiment.

In describing FIG. 16 , the MRR data may be the first MRR data according to the above-described embodiments or the second MRR data personalized to the user. In addition, the recipe card may be a first recipe card or a second recipe card according to the above-described embodiments. Hereinafter, for convenience of explanation, both the first MRR data and the second MRR data will be referred to as 'MRR data', and both the first recipe card and the second recipe card will be referred to as 'recipe cards'.

A cooking information providing system according to an embodiment may include a server 2000 and a client device 3000 . Also, the cooking information providing system may further include a cloud server to provide cooking in connection with the user's cooking device.

Referring to FIG. 16 , the server 2000 according to an embodiment may receive a text indicating a recipe from the client device 3000 , and may generate MRR data and a recipe card according to the above-described embodiments.

The server 2000 according to an embodiment may provide the generated MRR data and the recipe card to the client device 3000 so that a user who cooks food controls the cooking apparatus and is provided with information about food cooking. . In this case, the client device 3000 may be a mobile device such as a smart phone or a tablet PC that the user moves and can carry. In addition, the client device 3000 may be a desktop PC, a notebook PC, a TV, a refrigerator, etc. which are various devices capable of executing an application, a web, and the like. Also, the client device 3000 may be a cooking apparatus that may receive information about cooking food including a display.

The server 2000 according to an embodiment may receive texts representing a plurality of recipes from the client device 3000 and generate MRR data and recipe cards for each of the texts representing the plurality of recipes. MRR data and recipe cards generated by the server may be received by the client device 3000 .

In an embodiment, the MRR data and recipe cards generated by the server 2000 may be provided to the user in the form of an application in the client device 3000 . In this case, the client device 3000 may display a list 1610 (hereinafter, referred to as a recipe list) of recipes in which MRR data is generated.

When receiving an input for selecting any one recipe from the user of the client device from the recipe list 1610 , the client device 3000 may display recipe cards corresponding to the recipe selected by the user.

For example, when the user of the client device selects 'recipe D 1611' from the recipe list 1610, the client device 3000 may display recipe cards 1620 corresponding to 'recipe D'. have. The client device 3000 uses the recipe cards 1620 to provide a user with various functions related to cooking (eg, control of a cooking device, purchase of cooking items) and information related to cooking (eg, video of material handling). etc. can be provided.

In one embodiment, when the server 2000 provides a function related to cooking to the user using the recipe cards 1620, the server 2000 uses the IOT cloud server 1630 to control the cooking devices owned by the user. It can work with cloud services (eg, SmartThings) provided by . In this case, when the user logs into the IOT cloud service with the user's account, the server 2000 may receive the user's cooking device list and identification information for identifying the user's cooking devices from the IOT cloud server 1630 . In addition, the user's cooking device list and identification information for identifying the user's cooking devices may be transmitted from the IOT cloud server 1630 to the client device 3000 , and then transmitted from the client device 3000 to the server 2000 . have.

The server 2000 may receive the user's cooking device list owned by the user, and may change at least a portion of the MRR data to be data suitable for the cooking device owned by the user. Since this corresponds to the embodiments described above in FIG. 13 , the description will be omitted for simplicity of description.

In addition, the server 2000 receives a cooking device control request according to an input of a user selecting a cooking device control button from the recipe card from the client device 3000 , and includes cooking device control information included in MRR data to control the cooking device. may be transmitted to the cloud server 1630 . In this case, the cloud server 1630 may control the user's cooking devices based on the received cooking device control information. Since this corresponds to the embodiments described above in FIG. 12 , the description will be omitted for simplicity of description.

In an embodiment, the server 2000 may generate MRR data for a new recipe by using the pre-generated MRR data. For example, when the user of the client device selects the new recipe creation button 1612, the client device 3000 displays a user interface 1640 for creating a new recipe, and receives the user's input for editing the recipe. It can be transmitted to the server 2000 . The client device 3000 may display the user interface 1650 for sharing the created new recipe to share the new recipe created by the user.

17 is a diagram illustrating a GUI for generating MRR data for a new recipe by using a plurality of pre-generated MRR data in the cooking information providing system according to an embodiment.

17, the MRR data may be the first MRR data according to the above-described embodiments or the second MRR data personalized to the user. Hereinafter, for convenience of description, both the first MRR data and the second MRR data will be referred to as 'MRR data'.

Referring to FIG. 17 , MRR data and recipe cards generated according to the above-described embodiments may be provided to a user in the form of an application. In this case, the client device 3000 may display a list of recipes in which MRR data is generated.

In an embodiment, the client device 3000 may receive an input for selecting the recipe creation button 1710 from the user of the application. The client device 3000 may display the recipe editing tool 1720 in response to the user's request for creating a recipe. The user may search the MRR data for the food by inputting the name of the dish for which the user wants to create a recipe in the recipe editing tool. In this case, the client device 3000 may search for MRR data in the database within the server 2000 according to the disclosed embodiment. In addition, the client device 3000 may search for MRR data stored in the memory of the client device.

In an embodiment, the client device 3000 may display a list 1730 of a plurality of recipes corresponding to the name of the dish input by the user. For example, when the user of the client device searches for tteokbokki, recipes such as 'Baek Jong-won's tteokbokki recipe', 'Sinchon oil tteokbokki recipe', and 'sausage soy sauce tteokbokki recipe' may be displayed.

In an embodiment, the client device 3000 may configure the components included in the MRR data of the selected recipe based on a user input for selecting one of a plurality of recipes from the list of searched recipes 1730 by the recipe editing tool 1720 ) can be displayed. Also, the client device 3000 may generate MRR data corresponding to a new recipe based on a user input for editing the components 1740 included in the MRR data using the recipe editing tool 1720 .

For example, based on a user input for selecting a 'sausage soy sauce tteokbokki (1732)' recipe from among a plurality of recipes in the list 1730 of the searched recipes, the client device 3000 receives the 'sausage soy sauce tteokbokki' from the server 2000. (1732)' MRR data corresponding to the recipe may be obtained. Alternatively, the client device 3000 may acquire MRR data corresponding to the 'sausage soy sauce tteokbokki (1732)' recipe from a memory in the client device.

The client device 3000 may display, on the screen, cooking ingredients and cooking steps, which are components included in the MRR data corresponding to the acquired 'sausage soy sauce tteokbokki (1732)' recipe. In this case, the item of 'Cooking ingredients' includes cooking ingredients such as 'Tteokguk-tteok, sausage, onion, and green onion', and the sub-item of 'Cooking step' includes 'Sausage… , to the hot fan... , the broth… Cooking steps such as , ' may be included.

In an embodiment, the user of the client device may change the recipe component 'Tteokbokki Tteok' to 'Tteokbokki Tteok' in the recipe editing tool 1720 . When a component included in the MRR data is changed in the recipe editing tool 1720 , the client device may transmit information related to the changed component to the server 2000 .

The server 2000 according to an embodiment may receive the components of the MRR data changed by the user from the client device 3000, and may generate MRR data related to a new recipe personalized to the user. The client device 3000 may display a user interface 1750 for sharing the created user's new recipe. The user may share the new recipe created by the user to other users by using the user interface 1750 for sharing the new recipe.

18 is a diagram illustrating an example in which recipe cards for cooking a plurality of foods are provided to a user through a client device in the cooking information providing system according to an embodiment.

18 , the recipe card may be a first recipe card, and may be a second recipe card including cooking information personalized to the user. For convenience of explanation, hereinafter, both the first recipe card and the second recipe card will be referred to as a 'recipe card'.

Referring to FIG. 18 , a user of a client device according to an embodiment may be a user who wants to cook a plurality of foods in parallel.

In an embodiment, the client device 3000 may display a list 1810 (hereinafter, a recipe list) of recipes in which MRR data is generated. In order to cook the plurality of foods in parallel, the user of the client device may select recipes for the plurality of foods from the recipe list 1810 displayed on the client device 3000 .

In an embodiment, the client device 3000 may receive an input for selecting a plurality of recipes from the recipe list 1810 from the user.

For example, the client device 3000 receives, from the user, a seasoning steak recipe 1812, a kimchi stew recipe 1814, and a spinach pickle recipe 1816 from the recipe list 1810 in order for the user to simultaneously cook a plurality of foods. You can choose.

The client device 3000 may display recipe cards 1820 corresponding to recipes selected by the user. In this case, the client device 3000 chronologically displays the recipe cards 1820 for the plurality of foods based on the time required for each food to be cooked so that the user can cook the plurality of foods in parallel. can be displayed.

For example, the client device 3000 displays a recipe card 1822 corresponding to the first cooking step of the seasoned steak recipe 1812 and cooks so that the user performs the first cooking step of the seasoned steak recipe 1812 . information can be provided.

Next, the client device 3000 displays a recipe card 1824 corresponding to the first cooking step of the kimchi stew recipe 1814, and after the user performs the first cooking step of the seasoning steak recipe 1812, kimchi Cooking information may be provided to perform the first cooking step of the stew recipe 1814 .

Next, the client device 3000 sequentially displays the recipe card 1826 corresponding to the first cooking step of the spinach-freezing recipe 1816 and the recipe card 1828 corresponding to the second cooking step, respectively, so that the user After the first cooking step of the kimchi stew recipe 1814 is performed, the cooking information may be provided to sequentially perform the first cooking step and the second cooking step of the spinach dressing recipe 1816 .

In the same way, when the user selects a plurality of recipes to cook a plurality of foods, the client device 3000 sequentially displays the recipe cards 1820 for the plurality of foods based on the time required to cook them. By displaying, it is possible to complete cooking of each food at a similar time period.

19 is a view for explaining an example in which a first recipe card and a personalized second recipe card are provided to a user through a client device in the cooking information providing system according to an embodiment.

Referring to FIG. 19 , the client device 3000 provides a list 1910 of recipes in which the first MRR data, the second MRR data, the first recipe card, and the second recipe card are generated according to the above-described embodiments (hereinafter, recipe list) can be displayed.

When receiving an input for selecting any one recipe from the user of the client device from the recipe list 1910 , the client device 3000 may display recipe cards corresponding to the recipe selected by the user.

For example, when the user of the client device selects 'recipe D' from the recipe list 1910, the client device 3000 may display recipe cards 1920 corresponding to 'recipe D'.

In this case, each recipe card (recipe card A, recipe card B, recipe card C, etc.) included in the recipe cards 1920 may be a first recipe card or a second recipe card personalized to a user.

For example, in one embodiment, the recipe card B 1930 displayed on the client device 3000 may be the first recipe card 1932 generated using the first MRR data. In this case, the client device 3000 may receive a user's selection in order to provide the user with a personalized, customized dish. When the client device 3000 receives an input requesting display of the user's second recipe card, the client device 3000 may display the second recipe card 1934 generated using the second MRR data.

The server 2000 receives a cooking device control request according to an input of a user selecting a cooking device control button from the recipe card from the client device 3000 , and uses cooking device control information included in MRR data to control the cooking device. to control the cooking device. In this case, when the user's input is an input to the first recipe card, the first MRR data may be used, and when the user's input is an input to the second recipe card, the second MRR data may be used.

After the user receives the first recipe card or the second recipe card and completes cooking, the client device 3000 may display a feedback card 1940 that allows the user to input evaluation data related to the cooked food. . The user's evaluation data is transmitted from the client device 3000 to the server 2000, and the server 2000 may generate the second MRR data using the user's evaluation data. Since this has been described in the description of FIG. 10 , a description thereof will be omitted herein.

The method for providing information for cooking food according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

In addition, the cooking information providing method according to the disclosed embodiments may be provided by being included in a computer program product. Computer program products may be traded between sellers and buyers as commodities.

The computer program product may include a S/W program and a computer-readable storage medium in which the S/W program is stored. For example, computer program products may include products (eg, downloadable apps) in the form of S/W programs distributed electronically through manufacturers of electronic devices or electronic markets (eg, Google Play Store, App Store). have. For electronic distribution, at least a portion of the S/W program may be stored in a storage medium or may be temporarily generated. In this case, the storage medium may be a server of a manufacturer, a server of an electronic market, or a storage medium of a relay server temporarily storing a SW program.

The computer program product, in a system consisting of a server and a client device, may include a storage medium of the server or a storage medium of the client device. Alternatively, if there is a third device (eg, a smart phone) that is communicatively connected to the server or the client device, the computer program product may include a storage medium of the third device. Alternatively, the computer program product may include the S/W program itself transmitted from the server to the client device or the third device, or transmitted from the third device to the client device.

In this case, one of the server, the client device and the third device may execute the computer program product to perform the method according to the disclosed embodiments. Alternatively, two or more of a server, a client device, and a third device may execute a computer program product to distribute the method according to the disclosed embodiments.

For example, a server (eg, a cloud server or an artificial intelligence server) may execute a computer program product stored in the server to control a client device communicatively connected with the server to perform the method according to the disclosed embodiments.

Although the embodiments have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention as defined in the following claims are also included in the scope of the present invention. belongs to

Claims (20)

A method for a server to generate data for cooking food,
obtaining a text representing a recipe of the food;
identifying a plurality of semantic elements included in the text by using a plurality of natural language understanding (NLU) models for analyzing the text;
generating first Machine Readable Recipe (MRR) data in a structured form including a plurality of cooking parameters by using the identified semantic elements;
acquiring log data of a user related to an operation in which the user manipulated the cooking apparatus; and
generating second MRR data personalized to the user by updating the plurality of cooking parameters based on the obtained log data;
The identifying of the plurality of semantic elements may include applying the text to each of the plurality of NLU models, and identifying the plurality of semantic elements based on at least one semantic element output from each of the plurality of NLU models. How to. According to claim 1,
The generating of the second MRR data includes:
providing control information for controlling the cooking device to the cooking device based on the first MRR data;
obtaining compensation values for control operations of the cooking apparatus corresponding to the control information; and
generating the second MRR data by updating the plurality of cooking parameters based on the accumulated sum of the obtained compensation values;
wherein the reward values are determined based on the obtained log data of the user. 3. The method of claim 2,
The generating of the second MRR data includes:
receiving input data of a user operating the cooking device from the cooking device while the cooking device is operating according to the control information for controlling the cooking device;
acquiring user evaluation data for food cooked by the operation of the cooking device; and
and assigning the received user's input data and user's evaluation data to the reward values. The method of claim 1,
generating a plurality of first recipe cards for cooking the food by using the generated first MRR data; and
and providing the plurality of first recipe cards to a client device. 5. The method of claim 4,
generating a plurality of second recipe cards for cooking of the food personalized to the user by using the generated second MRR data; and
and providing the plurality of second recipe cards to the client device. According to claim 1,
Based on the type of food, the method further comprising the step of determining the plurality of NLU models. 5. The method of claim 4,
Further comprising the step of determining cooking steps for cooking the food by using the identified plurality of semantic elements,
The step of generating the plurality of first recipe cards,
Based on the generated first MRR data and the determined cooking steps, determining the types of the plurality of first recipe cards,
The type of the plurality of first recipe cards includes at least one of a cooking device control type, a cooking item purchase type, and a cooking information provision type. According to claim 1,
acquiring identification information of another cooking device; and
The method further comprising the step of changing at least a portion of the first MRR data and the second MRR data so that the food can be cooked by the other cooking device based on the obtained identification information of the other cooking device; Way. According to claim 1,
receiving, from a client device, characteristic information about a characteristic of the food input by the user;
The method further comprising the step of changing at least a portion of the first MRR data and the second MRR data based on the received characteristic information. According to claim 1,
obtaining information on ingredients owned by the user;
The method further comprising the step of changing at least a portion of the first MRR data and the second MRR data based on the obtained food ingredient information. In the server for generating data for food cooking,
communication department;
a storage unit for storing one or more instructions; and
a processor executing the one or more instructions stored in the storage unit;
including,
The processor is
obtain a text representing the recipe of the food,
A plurality of semantic elements included in the text are applied to a plurality of NLU (Natural Language Understanding) models for analyzing the text, respectively, and based on at least one semantic element output from each of the plurality of NLU models. identify,
using the identified semantic elements to generate first MRR (Machine Readable Recipe) data in a structured form including a plurality of cooking parameters;
Acquire log data of a user related to an operation in which the user manipulated the cooking device,
Based on the obtained log data, the server generates second MRR data personalized to the user by updating the plurality of cooking parameters. 12. The method of claim 11,
The processor is
providing control information for controlling the cooking device to the cooking device based on the first MRR data;
obtaining compensation values for control operations of the cooking apparatus corresponding to the control information;
generating the second MRR data by updating the plurality of cooking parameters based on the accumulated sum of the obtained compensation values;
The server, wherein the reward values are determined based on the obtained log data of the user. 13. The method of claim 12,
The processor is
Receive input data of a user who operates the cooking device from the cooking device while the cooking device is operating according to the control information for controlling the cooking device,
Obtaining user evaluation data for food cooked by the operation of the cooking device,
and assigning the received user's input data and user's evaluation data to the reward values. 12. The method of claim 11,
The processor is
Using the generated first MRR data to generate a plurality of first recipe cards for cooking the food,
A server providing the plurality of first recipe cards to a client device. 15. The method of claim 14,
The processor is
Using the generated second MRR data to generate a plurality of second recipe cards for cooking the food personalized to the user,
The server provides the plurality of second recipe cards to the client device. 12. The method of claim 11,
The processor is
Based on the type of food, the server determines the plurality of NLU models. 15. The method of claim 14,
The processor is
determining cooking steps for cooking the food by using the identified plurality of semantic elements;
Based on the generated first MRR data and the determined cooking steps, determining the types of the plurality of first recipe cards,
The type of the plurality of first recipe cards includes at least one of a cooking device control type, a cooking item purchase type, and a cooking information provision type. 12. The method of claim 11,
The processor is
Acquire identification information of other cooking devices,
Based on the obtained identification information of the other cooking device, the server changes at least a portion of the first MRR data and the second MRR data so that the food can be cooked by the other cooking device. 12. The method of claim 11,
The processor is
Receiving characteristic information about the characteristic of the food input by the user from the client device,
Based on the received characteristic information, the server to change at least a part of the first MRR data and the second MRR data. A computer-readable recording medium in which a program for executing the method of any one of claims 1 to 10 in a computer is recorded.

Images