Recipe recognition method, recipe recognition system, recipe recognition device, recipe recognition program, and recording medium on which the program is stored.

Abstract

This invention provides a recipe recognition method that can recognize recipes that can achieve sufficient taste and other qualities even when using substitute ingredients. [Solution] The recipe recognition system S performs a process comprising: a basic recipe recognition unit 10 recognizing a basic recipe; an excluded ingredient recognition unit 12 recognizing excluded ingredients which are ingredients to be excluded from the basic recipe; an alternative ingredient recognition unit 13 recognizing alternative ingredients which are ingredients that can be used in place of the excluded ingredients in the basic recipe and consist of a combination of multiple ingredients; an alternative recipe recognition unit 14 recognizing an alternative recipe using the alternative ingredients; and a recipe presentation unit 17 presenting the alternative recipe.

Description

Technical Field

[0001] The present invention relates to a recipe recognition method for recognizing a recipe using other ingredients instead of specific ingredients, a recipe recognition system, a recipe recognition device, a recipe recognition program, and a recording medium recording the program.

Background Art

[0002] When cooking a dish, there may be cases where specific ingredients cannot be used due to reasons such as the constitution or beliefs of the person who eats the dish or the person who cooks the dish. On the other hand, when using alternative ingredients instead of the specific ingredients, depending on the type of alternative ingredients, the dish using the alternative ingredients may not be sufficient in terms of taste, aroma, texture, appearance, nutrition, etc. (hereinafter sometimes simply referred to as "taste, etc.") compared to the dish without using the alternative ingredients.

[0003] Therefore, when using alternative ingredients instead of specific ingredients, a system has been proposed that presents a cooking method according to the alternative ingredients in order to make the taste, etc. of the dish sufficient (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, even for ingredients of the same type, the characteristics of the ingredients are various. Therefore, when replacing a specific ingredient with an alternative ingredient, there is a possibility that simply changing the cooking method may not make the taste, etc. sufficient compared to the dish using the specific ingredient.

[0006] The present invention has been made in view of the above points, and aims to provide a recipe recognition method, a recipe recognition system, a recipe recognition device, a recipe recognition program, and a recording medium on which the program is stored, which can recognize recipes for dishes that can achieve sufficient taste and other qualities even when using substitute ingredients. [Means for solving the problem]

[0007] The recipe recognition method of the present invention is A recipe recognition method that recognizes recipes that use other ingredients instead of specific ingredients, The basic recipe recognition unit recognizes a basic recipe, which is the recipe for the desired dish, based on the user's input. The excluded ingredient recognition unit recognizes excluded ingredients, which are ingredients to be excluded from the basic recipe, based on the user's input. The alternative ingredient recognition unit recognizes, based on the basic recipe and the excluded ingredients, an alternative ingredient that can be used in place of the excluded ingredients in the basic recipe, The alternative recipe recognition unit includes the step of recognizing an alternative recipe, which is a recipe for making the desired dish using the alternative ingredients instead of the excluded ingredients, based on the desired dish, excluded ingredients, and alternative ingredients. The aforementioned substitute ingredients are characterized by consisting of a combination of multiple ingredients.

[0008] Thus, the recipe recognition method of the present invention employs a combination of multiple ingredients as substitute ingredients. This allows for more flexible reproduction of the taste, aroma, texture, appearance, nutrition, etc. (hereinafter sometimes simply referred to as "taste, etc.") of a specific ingredient than when a single ingredient is used as a substitute ingredient.

[0009] Therefore, the recipe recognition method of the present invention can flexibly reproduce the taste and other characteristics of specific ingredients, making it possible to recognize dishes that can achieve sufficient taste and other characteristics even when using substitute ingredients. Consequently, users can become familiar with recipes for dishes that can achieve sufficient taste and other characteristics even when using substitute ingredients.

[0010] Furthermore, in the recipe recognition method of the present invention, Preferably, the substitute ingredient recognition unit recognizes the substitute ingredient using a first generation model that, upon input of second data including the desired dish and the excluded ingredients, outputs third data representing the substitute ingredient, after referring to a plurality of first data representing the characteristics of the ingredients.

[0011] Generally, cooking is a complex process involving numerous elements, using multiple ingredients and going through many steps to complete the dish. Therefore, finding substitute ingredients that can bring the taste of the finished dish close to that of the original dish, or even improve upon it, is difficult. Furthermore, the search becomes even more challenging when the substitute ingredient is a combination of multiple ingredients.

[0012] Therefore, by using a generative model to search for alternative ingredients in this way, the search or generation is performed by referring to a vast number of samples, making it easy to identify a large number of candidate alternative ingredients. Consequently, it becomes easier to recognize dishes that can achieve sufficient taste and other qualities even when using alternative ingredients. As a result, users can more easily identify recipes that can achieve sufficient taste and other qualities for their own taste from among recipes using alternative ingredients.

[0013] Furthermore, in the recipe recognition method of the present invention, The recipe presentation unit provides the alternative recipe, The cooking parameter recognition unit recognizes cooking parameters, which are parameters relating to at least one of the following aspects of a dish: taste, aroma, texture, appearance, and nutrition. The cooking parameter comparison unit includes the step of comparing the cooking parameters of the desired dish made by the alternative recipe with the cooking parameters of the desired dish made by the basic recipe, The alternative recipe recognition unit recognizes a plurality of alternative recipes, Preferably, the recipe presentation unit presents, along with each of the multiple alternative recipes, the results of comparing the cooking parameters of the desired dish made by the alternative recipe with the cooking parameters of the desired dish made by the basic recipe.

[0014] When using substitute ingredients, differences inevitably arise between dishes made with those substitutes and those made with the original ingredients. These differences can manifest in various ways, including taste, aroma, texture, appearance, and nutritional value, and the importance placed on each difference varies from user to user.

[0015] Therefore, when configuring the system to recognize multiple alternative recipes and allow the user to select the recipe they actually want to use, it is advisable to present a comparison of the desired dish made by each alternative recipe with the desired dish made by the basic recipe, along with each of the alternative recipes, for cooking parameters that relate to at least one of the following: taste, aroma, texture, appearance, and nutrition.

[0016] With this configuration, users can easily understand the cooking parameters of the dishes produced by each alternative recipe. This allows users to easily identify recipes using alternative ingredients that will produce a taste and other qualities that meet their needs.

[0017] Furthermore, in the recipe recognition method of the present invention, in the case of a configuration that compares cooking parameters, The specified parameter recognition unit includes the step of recognizing a specified parameter, which is a cooking parameter specified by the user, from among the cooking parameters, based on the user's input. The recipe presentation unit preferably presents, with greater emphasis than other alternative recipes, alternative recipes in which the difference between the specified parameters in the desired dish made by the alternative recipe and the specified parameters in the desired dish made by the basic recipe is small.

[0018] Thus, when configured to emphasize and display alternative recipes with smaller differences in specified parameters (i.e., cooking parameters that the user values), the user can more easily select a recipe to adopt from among a plurality of alternative recipes based on the cooking parameters that the user values. As a result, the user can more easily grasp a recipe that can achieve a sufficient taste or the like for the user from recipes using alternative ingredients.

[0019] Also, in the recipe recognition method of the present invention, in the case of a configuration for comparing cooking parameters, the cooking point recognition unit includes a step of recognizing a cooking point, which is a cooking point for reducing the difference, for a cooking parameter recognized as being different as a result of the comparison. The recipe presentation unit preferably presents the cooking point together with the alternative recipe.

[0020] Some users may want to use a specific alternative ingredient. On the other hand, when using that specific alternative ingredient, the difference from a dish that does not use the alternative ingredient may become larger than that of a dish using other alternative ingredients in some cases.

[0021] Therefore, when configured to present to the user a cooking point for suppressing the difference caused by using the alternative ingredient in this way, the user can grasp a recipe that uses a specific alternative ingredient that the user wants to use and can achieve a sufficient taste or the like for the user from recipes using the alternative ingredient.

[0022] Also, in the recipe recognition method of the present invention, in the case of a configuration for comparing cooking parameters, The cooking parameter recognition unit and the cooking parameter comparison unit, upon inputting a plurality of fourth data which are text representing the cooking parameters for a dish made by a predetermined recipe, and a plurality of fifth data representing a plurality of recipes, use a second generation model that outputs the cooking parameters for each of the plurality of recipes and a sixth data representing the result of the comparison of the cooking parameters to recognize the cooking parameters in the desired dish made by the alternative recipe, the cooking parameters in the desired dish made by the basic recipe, and the result of the comparison. Preferably, the sixth data includes text representing the results of the comparison.

[0023] Generally, cooking parameters are expressed in a variety of ways, and the effects they produce can differ from person to person, even with the same numerical value. Therefore, even if cooking parameters are presented using a uniform numerical representation, it may not be possible to adequately convey their meaning to the user.

[0024] Therefore, by using a generative model to recognize text representing the content of cooking parameters, it becomes possible to present cooking parameters in a way that is easy for users to understand. This allows users to intuitively understand the differences between alternative recipes and the basic recipe.

[0025] Furthermore, the recipe recognition system of the present invention is A recipe recognition system that recognizes recipes that use other ingredients instead of specific ingredients, It is characterized by being configured to perform one of the above recipe recognition methods.

[0026] Furthermore, the recipe recognition device of the present invention A recipe recognition device that recognizes recipes using other ingredients instead of specific ingredients, It is characterized by being configured to perform one of the above recipe recognition methods.

[0027] Furthermore, the recipe recognition program of the present invention, A recipe recognition program that causes a computer to execute a recipe recognition method that recognizes recipes using other ingredients instead of specific ingredients, The computer is characterized by being made to execute one of the above-described recipe recognition methods.

[0028] Furthermore, the recording medium of the present invention is The above-mentioned recipe recognition program is recorded, and the recipe recognition program is readable by the computer. [Brief explanation of the drawing]

[0029] [Figure 1] An explanatory diagram showing the schematic configuration of the recipe recognition system according to the embodiment. [Figure 2] A block diagram showing the configuration of the processing unit of the recipe recognition system in Figure 1. [Figure 3] Figure 1 is an illustrative image showing an example of the screen displayed on the user's terminal when the recipe recognition system instructs the system to present a recipe. [Figure 4] This image shows an example of a screen displayed on the user's terminal when multiple candidate recipes are presented by the recipe recognition system shown in Figure 1. [Figure 5] Figure 1 is an illustrative image showing an example of the screen displayed on the user's terminal when recipe details are presented by the recipe recognition system. [Figure 6] Figure 1 is a flowchart showing the processes that the recipe recognition system executes when presenting a recipe. [Modes for carrying out the invention]

[0030] The following describes a recipe recognition system S according to an embodiment, and a recipe recognition method performed using the same, with reference to the drawings.

[0031] The recipe recognition system S described below is a computer system for recognizing recipes that use other ingredients instead of specific ingredients. Specifically, the recipe recognition system S recognizes substitute ingredients that can be used in the dish desired by the user, in place of excluded ingredients that the user has entered, and is used in a service that presents recipes using those substitute ingredients and related information.

[0032] [System Outline] First, the general configuration of the recipe recognition system S will be described with reference to Figures 1 and 2.

[0033] As shown in Figure 1, the recipe recognition system S is comprised of a server 1 owned by the provider of the service provided by the recipe recognition system S.

[0034] Server 1 is configured to communicate with user terminals 2, such as smartphones and tablets, via the internet, public network, etc.

[0035] Furthermore, the recipe recognition system of the present invention is not limited to being composed of a single server; it is sufficient if any of the terminals constituting the recipe recognition system are configured to have the processing unit described later.

[0036] Therefore, for example, the entire recipe recognition system may be configured using multiple servers. Alternatively, the user terminal may implement at least one processing unit or at least a part of the functions of that processing unit, and the system may be configured either through collaboration between the user terminal and the server or solely by the user terminal.

[0037] Furthermore, in the recipe recognition system S of this embodiment, the functions provided by the user terminal 2 may be provided on the terminals that make up the server 1, and configured as an independent recipe recognition device.

[0038] As shown in Figure 2, the user terminal 2 includes, as a function (processing unit) realized by at least one of the implemented hardware configuration and program, an input unit 20 for causing the recipe recognition system S to recognize the conditions of the user's desired recipe, and an output unit 21 for presenting the recipe recognized by the recipe recognition system S to the user.

[0039] In this embodiment, the user terminal is a notebook computer, and its input unit 20 is a keyboard, and its output unit 21 is a monitor.

[0040] However, the user terminal is not limited to such a configuration; it can be any device capable of receiving information input from the user and outputting information to the user. For example, the user terminal may be configured to enable input and output using other devices such as a microphone, camera, and speakers, in addition to a keyboard and monitor.

[0041] Furthermore, in this embodiment, the server 1 and user terminal 2 are described as being composed of real-world objects. However, the devices that input or output information to or from the recipe recognition system of the present invention are not limited to real-world objects, but may be virtual objects, or they may be composed of a combination of real-world and virtual objects.

[0042] Therefore, for example, at least one of the input and output sections of the user terminal 2 in this embodiment may be made virtual, and the user terminal 2 may be configured to allow input and output of information in a virtual space, or in a space that combines the virtual space and the real space.

[0043] [Configuration of each processing unit] Next, the processing units that constitute the recipe recognition system S will be described with reference to Figures 2 to 5.

[0044] As shown in Figure 2, Server 1 includes, as functions (processing units) realized by at least one of the implemented hardware configuration and program, a basic recipe recognition unit 10, a recipe information storage unit 11, an excluded ingredient recognition unit 12, an alternative ingredient recognition unit 13, an alternative recipe recognition unit 14 (cooking point recognition unit), a specified parameter recognition unit 15, a cooking parameter recognition unit 16 (cooking parameter comparison unit), and a recipe presentation unit 17.

[0045] The basic recipe recognition unit 10 recognizes the desired dish (the dish the user wants) and the basic recipe (the recipe for that desired dish) based on the user's input.

[0046] In the recipe recognition system S, the basic recipe recognition unit 10 first recognizes the dish entered via the input unit 20 of the user terminal 2 as the desired dish. Then, the basic recipe recognition unit 10 searches the recipe information storage unit 11 for the basic recipe corresponding to that desired dish and recognizes it. In this embodiment, pudding is assumed to be the desired dish (see Figure 3).

[0047] Furthermore, the basic recipe recognition unit of the present invention is not limited to this configuration; it is sufficient if it is configured to recognize a basic recipe, which is the recipe for the desired dish, based on user input.

[0048] Therefore, for example, the basic recipe recognition unit may be configured to recognize basic recipes using a database or the like located outside the system. Alternatively, the basic recipe recognition unit may be configured to recognize desired dishes corresponding to basic recipes based on information entered by the user when using other services.

[0049] The recipe information storage unit 11 stores information about recipes for dishes that could become the desired dish, as well as information about the characteristics of the ingredients used in those recipes.

[0050] Here, "information about the recipe" includes, for example, the recipe itself, as well as text data describing the dish made using that recipe. Furthermore, "information about the characteristics of the ingredients used in the recipe" includes, for example, the cooking methods that can be employed, the changes in taste, aroma, texture, appearance, nutrition, etc. (hereinafter sometimes simply referred to as "taste, etc.") due to the cooking methods, and the characteristics when combined with other ingredients.

[0051] The excluded ingredient recognition unit 12 recognizes excluded ingredients, which are ingredients to be excluded from the basic recipe, based on user input. In the recipe recognition system S, the excluded ingredient recognition unit 12 recognizes ingredients entered via the input unit 20 of the user terminal 2 as excluded ingredients. In this embodiment, eggs are assumed to be excluded ingredients (see Figure 3).

[0052] It should be noted that the excluded ingredient recognition unit of the present invention is not limited to this configuration, and may be configured to recognize excluded ingredients based on user input.

[0053] Therefore, for example, the excluded ingredient recognition unit may be configured to recognize excluded ingredients based on information entered by the user when using other services.

[0054] Furthermore, the excluded food recognition unit may recognize the reason why the user wants to exclude certain foods, and based on that reason, it may also be configured to exclude foods other than those specified as excluded foods. Specifically, for example, if the excluded foods are selected to avoid allergies, the excluded food recognition unit may be configured to recognize not only the foods directly designated as excluded foods, but also other foods that are presumed to cause cross-reaction allergies in relation to those designated foods.

[0055] The substitute ingredient recognition unit 13 recognizes substitute ingredients that can be used in place of excluded ingredients in the basic recipe, based on the basic recipe and excluded ingredients. Here, the substitute ingredients may be a combination of multiple ingredients that do not include the excluded ingredients, and may substitute one ingredient or multiple ingredients. In this embodiment, it is assumed that a combination of silken tofu and agar, a combination of yogurt and gelatin, etc., are recognized as substitute ingredients in place of the excluded ingredient, egg (see Figure 4).

[0056] In the recipe recognition system S, the substitute ingredient recognition unit 13 recognizes substitute ingredients by referring to multiple data representing the characteristics of ingredients (first data of the present invention) and, upon input of data including the desired dish and excluded ingredients (second data of the present invention), using a generation model (first generation model of the present invention) that outputs data representing substitute ingredients (third data of the present invention).

[0057] More specifically, the substitute ingredient recognition unit 13 recognizes substitute ingredients by using a model that uses machine learning to correlate ingredients with their roles in dishes using data representing the characteristics of ingredients stored in the recipe information storage unit 11, and then generates and outputs combinations of ingredients that can be substituted for excluded ingredients based on the role of excluded ingredients in the input desired dish. For example, a so-called large-scale language model can be used as such a model.

[0058] Generally, cooking is a complex process involving numerous elements, using multiple ingredients and going through many steps to complete the dish. Therefore, finding substitute ingredients that can bring the taste of the finished dish close to that of the original dish, or even improve upon it, is difficult. Furthermore, the search becomes even more challenging when the substitute ingredient is a combination of multiple ingredients.

[0059] Therefore, the recipe recognition system S uses a generative model to search for alternative ingredients. This allows for searching or generating by referencing a vast number of samples, making it easy to identify many potential alternative ingredients. Consequently, it becomes easier to recognize dishes that can achieve sufficient taste and other qualities even when using alternative ingredients. As a result, users can more easily identify recipes that achieve a satisfactory taste and other qualities for themselves from among recipes that use alternative ingredients.

[0060] The alternative ingredient recognition unit of the present invention is not limited to this configuration, and may be configured to recognize alternative ingredients as long as it is capable of doing so. For example, the alternative ingredient recognition unit may be configured to recognize alternative ingredients using a generative model or the like located outside the system. Alternatively, the alternative ingredient recognition unit may be configured to recognize alternative ingredients based on correlation data representing the correlation between an ingredient and the role of that ingredient in a dish using that ingredient.

[0061] The alternative recipe recognition unit 14 (cooking point recognition unit) recognizes the desired dish based on the user's input, and based on the desired dish, excluded ingredients, and alternative ingredients, recognizes an alternative recipe, which is a recipe for making the desired dish using alternative ingredients instead of excluded ingredients.

[0062] In the recipe recognition system S, the alternative recipe recognition unit 14 recognizes alternative recipes by referring to multiple data representing recipes related to cooking and multiple data representing the characteristics of alternative ingredients. When the desired dish, excluded ingredients, and alternative ingredients are input, the system outputs data representing an alternative recipe using a generative model.

[0063] More specifically, the alternative recipe recognition unit 14 recognizes alternative recipes by using machine learning to determine the correlation between ingredients and recipes from data representing the desired dish, recipes with similar characteristics to the desired dish, and data representing the characteristics of ingredients, all stored in the recipe information storage unit 11. Based on the input desired dish, excluded ingredients, and alternative ingredients, it generates and outputs a recipe using the alternative ingredients instead of the excluded ingredients. For example, a so-called large-scale language model can be employed as this model.

[0064] Furthermore, in the recipe recognition system S, the alternative recipe recognition unit 14 is configured to recognize multiple alternative recipes in response to user instructions when recognizing alternative recipes. Specifically, if the user responds to the question "Please let us know if you have any other desired conditions" on the screen shown in Figure 3 with an instruction such as "Please suggest multiple recipes," the alternative recipe recognition unit 14 will recognize multiple alternative recipes.

[0065] This is to allow the user to select from multiple recipes, as will be described later, and to enable the user to actually use the recipe. However, the alternative recipe recognition unit of the present invention is not limited to this configuration, and may be configured to recognize only one alternative recipe.

[0066] Furthermore, in the recipe recognition system S, the alternative recipe recognition unit 14 generates an image of the desired dish made using the recognized alternative recipe. Specifically, the alternative recipe recognition unit 14 uses a model that generates image images from text data, inputs the text data of the alternative recipe, and generates an image of the desired dish made using the alternative recipe.

[0067] Furthermore, in the recipe recognition system S, the alternative recipe recognition unit 14 recognizes the alternative recipe again when it receives the results of the cooking parameter comparison from the cooking parameter recognition unit 16, which will be described later.

[0068] Specifically, the alternative recipe recognition unit 14 recognizes alternative recipes that reduce the differences in cooking parameters that were recognized as differing as a result of the comparison. If there are differences in the cooking process between the alternative recipe recognized this time and the alternative recipe recognized last time, it also recognizes those differences as cooking points.

[0069] Furthermore, the alternative recipe recognition unit of the present invention is not limited to this configuration, and may be configured to recognize alternative recipes.

[0070] Therefore, for example, the alternative recipe recognition unit may be configured to recognize alternative recipes using a generative model located outside the system. Alternatively, the alternative recipe recognition unit may be configured to recognize alternative recipes based on correlation data representing the correlation between ingredients and recipes, and furthermore, similar to the case where a generative model is used, it may be configured to recognize cooking points based on that correlation data.

[0071] Furthermore, the alternative recipe recognition unit may be configured to recognize only one alternative recipe. Also, the alternative recipe recognition unit may be configured not to generate an image of the desired dish made using the alternative recipe. Furthermore, the alternative recipe recognition unit may be configured not to recognize alternative recipes again based on the results of comparing cooking parameters. Finally, the alternative recipe recognition unit may be configured not to recognize cooking points.

[0072] The specified parameter recognition unit 15 recognizes, based on user input, the specified parameter, which is a cooking parameter specified by the user, from among cooking parameters that relate to at least one of the taste, aroma, texture, appearance, and nutrition of the desired dish.

[0073] In the recipe recognition system S, the specified parameter recognition unit 15 recognizes the cooking parameters input via the input unit 20 of the user terminal 2 as specified parameters. In this embodiment, texture is assumed to be the specified parameter (see Figure 3).

[0074] The cooking parameter recognition unit 16 (cooking parameter comparison unit) recognizes the cooking parameters of the desired dish made by each of the multiple alternative recipes, and the cooking parameters of the desired dish made by the basic recipe, and compares these parameters.

[0075] In the recipe recognition system S, the cooking parameter recognition unit 16 refers to a plurality of data (the fourth data of the present invention) which are text representing cooking parameters for a dish made by a predetermined recipe, and when it receives data representing a plurality of recipes (the fifth data of the present invention), it uses a model (the second generation model of the present invention) that outputs data representing the cooking parameters for each of the dishes made by those plurality of recipes, and the result of the comparison of those cooking parameters (the sixth data of the present invention) to recognize the cooking parameters of the desired dish made by an alternative recipe, the cooking parameters of the desired dish made by a basic recipe, and the result of the comparison between them.

[0076] More specifically, the cooking parameter recognition unit 16 recognizes the cooking parameters and their comparison results using a model that uses machine learning to correlate recipes and cooking parameters from data representing the desired dish and dishes with similar characteristics stored in the recipe information storage unit 11, as well as data representing the cooking parameters corresponding to the dishes made by those recipes. Based on multiple input alternative recipes and a basic recipe, the model generates and outputs the cooking parameters of the desired dish made by those recipes and the results of their comparison. For example, a so-called large-scale language model can be employed as such a model.

[0077] In this case, the recipe recognition system S can output not only numerical data representing the differences in the specified parameters as a result of the comparison, but also text data representing the differences in the specified parameters according to the user's instructions (see Figure 3).

[0078] The digitized data is sent to the alternative recipe recognition unit 14, which, as described above, uses the transmission of that data as a trigger to perform alternative recipe recognition again.

[0079] Meanwhile, the text data is transmitted to the recipe presentation unit 17, which then presents the data to the user along with the corresponding alternative recipe.

[0080] Generally, cooking parameters are expressed in a variety of ways, and the effects they produce can differ from person to person, even with the same numerical value. Therefore, even if cooking parameters are presented using a uniform numerical representation, it may not be possible to adequately convey their meaning to the user.

[0081] Therefore, by including textual data of the differences in the comparison results, cooking parameters can be presented in a way that is easy for users to understand. As a result, users who are shown this data can intuitively understand the differences between the alternative recipe and the basic recipe.

[0082] The text data may simply point out the differences, or it may specifically indicate the advantages or disadvantages when compared to the desired dish made using the basic recipe.

[0083] Furthermore, the cooking parameter recognition unit of the present invention is not limited to this configuration, and may be configured to recognize cooking parameters, which are parameters relating to at least one of the taste, aroma, texture, appearance, and nutrition of a dish. Also, the cooking parameter comparison unit of the present invention is not limited to this configuration, and may be configured to compare the cooking parameters of a desired dish made using an alternative recipe with the cooking parameters of a desired dish made using a basic recipe.

[0084] Therefore, for example, the cooking parameter recognition unit and the cooking parameter comparison unit may be configured as independent processing units. In this case, at least one of the cooking parameter recognition unit and the cooking parameter comparison unit may be configured to recognize cooking parameters or comparison results based on correlation data rather than a generative model.

[0085] Furthermore, in this embodiment, a comparison is made between desired dishes made using multiple alternative recipes. However, the cooking parameter comparison unit of the present invention may only compare each of the desired dishes made using multiple alternative recipes with the desired dish made using the basic recipe.

[0086] Furthermore, in the recipe recognition system S, the specified parameters (answers to the question "Please tell us what you consider important") are recognized on the screen shown in Figure 3, and in response to the question "Please tell us if there are any other desired conditions," instructions such as "Please tell us specifically what differences there would be if other ingredients were used instead of eggs" are given. As a result, the cooking parameter recognition unit 16 recognizes and compares the cooking parameters, and accordingly, the alternative recipe recognition unit 14 recognizes the alternative recipe again and also recognizes additional cooking points.

[0087] However, if no instructions are given for the specified parameters (for example, if the user does not answer the question), the cooking parameter recognition unit 16 will not perform recognition and comparison, nor will the alternative recipe recognition unit 14 perform further recognition and additional recognition.

[0088] The recipe presentation unit 17 presents the recognized alternative recipes to the user via the user terminal 2.

[0089] As shown in Figure 4, in the recipe recognition system S, the recipe presentation unit 17 presents the recognized alternative recipes in a selectable format, and when a specific alternative recipe is selected (specifically, when the "View Details" button in the upper left corner of the alternative recipe is pressed), the details of that alternative recipe are displayed, as shown in Figure 5.

[0090] Each of the multiple alternative recipes presented at this time is accompanied by a comparison of the cooking parameters of the desired dish made using that alternative recipe with the cooking parameters of the desired dish made using other alternative recipes and the cooking parameters of the desired dish made using the basic recipe. Specifically, this includes text such as "Point! Better texture than other recipes," and "Point! Richer flavor than other recipes."

[0091] When using substitute ingredients, differences inevitably arise between dishes made with those substitutes and those made with the original ingredients. These differences can manifest in various ways, including taste, aroma, texture, appearance, and nutritional value, and the importance placed on each difference varies from user to user.

[0092] Therefore, the recipe recognition system S is configured to present, along with each of the multiple alternative recipes, the results of a comparison between the desired dish made by each alternative recipe and the desired dish made by the basic recipe, for cooking parameters, which are parameters related to at least one of the following: taste, aroma, texture, appearance, and nutrition.

[0093] With this configuration, users can easily understand the cooking parameters of the dishes produced by each alternative recipe. This allows users to easily identify recipes using alternative ingredients that will produce a taste and other qualities that are satisfactory to them.

[0094] Furthermore, at this time, among the multiple alternative recipes, the alternative recipe with the smallest difference between the specified parameters of the desired dish made by the alternative recipe and the specified parameters of the desired dish made by the basic recipe is presented in a higher position than the other alternative recipes.

[0095] By configuring the system to highlight alternative recipes that show small differences in specified parameters (i.e., cooking parameters that the user considers important), users can more easily select a recipe from among multiple alternative recipes based on the cooking parameters they value. This makes it even easier for users to identify recipes using alternative ingredients that can achieve a taste and other qualities that meet their needs.

[0096] Furthermore, the method of emphasis is not limited to presenting items in a higher-ranking manner, as in this embodiment; any method that is easily recognizable to the user is acceptable. For example, adding a "Recommended!" label to a recipe would also suffice.

[0097] Furthermore, as shown in Figure 5, each of the multiple alternative recipes presented at this time is accompanied by cooking points—points for minimizing the differences in cooking parameters that were identified as differing as a result of the comparison. Specifically, these are the text entries in the "Cooking Points" column.

[0098] Some users may want to use specific substitute ingredients. However, using those specific substitute ingredients may result in a greater difference between dishes using other substitute ingredients and dishes that do not.

[0099] Therefore, by configuring the system to present cooking tips to the user to minimize the differences caused by using substitute ingredients, users will be able to identify recipes that use the specific substitute ingredients they wish to use and that achieve a satisfactory taste and other qualities for them.

[0100] Furthermore, in the recipe recognition system S, the recipe presentation unit 17 also attaches an image of the alternative recipe to the content of the alternative recipe.

[0101] It should be noted that the recipe presentation unit of the present invention is not limited to this configuration, and may be configured to present alternative recipes. For example, the recipe presentation unit may present only one alternative recipe. Also, if the recipe presentation unit presents multiple alternative recipes, it may change the order of presentation based on other criteria (e.g., by cost of ingredients). Furthermore, the recipe presentation unit does not have to present information other than the content of the alternative recipes, or it may present information other than the information shown in this embodiment (e.g., cost of ingredients) along with the alternative recipes.

[0102] [Processing performed in each processing unit] Next, with reference to Figures 2 to 6, the processes performed by the recipe recognition system S (i.e., the recipe recognition method) will be explained.

[0103] In this process, first, the recipe recognition system S recognizes the desired dish based on the user's input (Figure 6 / STEP01).

[0104] Specifically, as shown in Figure 3, the recipe recognition system S recognizes the dish entered via the input unit 20 of the user terminal 2 as the desired dish. In this embodiment, the desired dish is assumed to be "pudding," which is entered in response to the question, "What dish would you like to make?"

[0105] Next, the basic recipe recognition unit 10 recognizes the basic recipe, which is the recipe for the desired dish, based on the recognized desired dish (Figure 6 / STEP02).

[0106] Specifically, the basic recipe recognition unit 10 searches for and recognizes a basic recipe from the recipe information storage unit 11.

[0107] Next, the excluded ingredient recognition unit 12 recognizes the excluded ingredients, which are ingredients to be excluded from the basic recipe, based on the user's input (Figure 6 / STEP03).

[0108] Specifically, as shown in Figure 3, the excluded ingredient recognition unit 12 recognizes ingredients entered via the input unit 20 of the user terminal 2 as excluded ingredients. In this embodiment, "egg," entered in response to the question "Please tell me which ingredients you would like to exclude," is considered an excluded ingredient.

[0109] Next, the substitute ingredient recognition unit 13 recognizes substitute ingredients that can be used in place of excluded ingredients in the basic recipe, based on the basic recipe and excluded ingredients (Figure 6 / STEP04).

[0110] Specifically, the substitute ingredient recognition unit 13 recognizes substitute ingredients using a generation model (first generation model of the present invention) that outputs data representing substitute ingredients (third data of the present invention) when it receives data including the desired dish and excluded ingredients (second data of the present invention) by referring to a plurality of data representing the characteristics of ingredients (first data of the present invention). In this embodiment, it is assumed that, in place of the excluded ingredient egg, a combination of silken tofu and agar, a combination of yogurt and gelatin, etc., are recognized as substitute ingredients.

[0111] Next, the recipe recognition system S determines, based on the user's input, whether or not multiple recipes have been requested (Figure 6 / STEP05).

[0112] Specifically, as shown in Figure 3, the recipe recognition system S determines whether the instruction (text) entered via the input unit 20 of the user terminal 2 contains an instruction requesting multiple alternative recipes (for example, text such as "Please suggest multiple recipes").

[0113] If no multiple recipes are requested (the answer is NO in STEP05), the alternative recipe recognition unit 14 recognizes one alternative recipe based on the desired dish, excluded ingredients, and alternative ingredients, which is a recipe for making the desired dish using the alternative ingredients instead of the excluded ingredients (Figure 6 / STEP06).

[0114] Specifically, the alternative recipe recognition unit 14 recognizes one alternative recipe by referring to multiple data representing recipes related to cooking and multiple data representing the characteristics of alternative ingredients, and by using a generative model that outputs data representing an alternative recipe when the desired dish, excluded ingredients, and alternative ingredients are input.

[0115] On the other hand, if multiple recipes are requested (if the answer is YES in STEP 05), the alternative recipe recognition unit 14 recognizes multiple alternative recipes based on the desired dish, excluded ingredients, and alternative ingredients (Figure 6 / STEP 07).

[0116] Specifically, the generative model described above is used to recognize multiple alternative recipes.

[0117] Next, the specified parameter recognition unit 15 determines whether or not a specified parameter has been specified based on the user's input (Figure 6 / STEP08).

[0118] Specifically, the designated parameter recognition unit 15 recognizes the cooking parameters entered via the input unit 20 of the user terminal 2 as designated parameters, as shown in Figure 3. In this embodiment, "texture," entered in response to the question "What points are important to you?", is assumed to be the designated parameter.

[0119] If a specified parameter is specified (YES in STEP08), the cooking parameter recognition unit 16 (cooking parameter comparison unit) recognizes the cooking parameters of the desired dish made by each of the multiple alternative recipes, and the cooking parameters of the desired dish made by the basic recipe, and compares those parameters (Figure 6 / STEP09).

[0120] Specifically, the cooking parameter recognition unit 16, by referring to a plurality of data (the fourth data of the present invention) which are text representing cooking parameters for a dish made by a predetermined recipe, and inputting data representing a plurality of recipes (the fifth data of the present invention), uses a model (the second generation model of the present invention) that outputs data representing the cooking parameters for each of the dishes made by those plurality of recipes, and the result of the comparison of those cooking parameters (the sixth data of the present invention) to recognize the cooking parameters of the desired dish made by the alternative recipe, the cooking parameters of the desired dish made by the basic recipe, and the result of the comparison between them.

[0121] Next, the alternative recipe recognition unit 14 recognizes alternative recipes that reduce the differences in cooking parameters that were identified as differing as a result of the comparison. If there are differences in the cooking process between the alternative recipe recognized this time and the alternative recipe recognized last time, it also recognizes those differences as cooking points (Figure 6 / STEP 10).

[0122] If no multiple recipes are requested and only one alternative recipe is recognized (when the process in STEP06 is executed), if the specified parameter is not specified (when the result is NO in STEP08), or if an alternative recipe is recognized again (when the process in STEP10 is executed), the recipe presentation unit 17 presents the alternative recipe to the user via the output unit 21 of the user terminal 2 (Figure 6 / STEP11), and then terminates the current process.

[0123] Specifically, if one alternative recipe is recognized, the recipe presentation unit 17 will present that one alternative recipe. If no specified parameters are provided, the recipe presentation unit 17 will present the recognized recipes in a predetermined order set by the system administrator (for example, by ingredient cost).

[0124] Furthermore, if an alternative recipe is recognized again, the recipe presentation unit 17 will present, as shown in Figure 4, alternative recipes in which the difference between the specified parameters of the desired dish made by the alternative recipe and the specified parameters of the desired dish made by the basic recipe is small, in order to rank them higher than other alternative recipes.

[0125] At this time, as shown in Figure 5, the recipe presentation unit 17 presents not only the recipe itself, but also information related to the recipe (for example, the cooking points mentioned above, an image of the desired dish made by the alternative recipe generated by the generative model, etc.).

[0126] As explained above, the recipe recognition method implemented by the recipe recognition system S employs a combination of multiple ingredients as substitute ingredients. This allows for more flexible reproduction of the taste and other characteristics of a specific ingredient than when a single ingredient is used as a substitute ingredient.

[0127] Therefore, this recipe recognition method can flexibly reproduce the taste and other characteristics of specific ingredients, enabling the recognition of dishes that can achieve sufficient taste and other characteristics even when using substitute ingredients. Consequently, users can become familiar with recipes for dishes that can achieve sufficient taste and other characteristics even when using substitute ingredients.

[0128] [Other embodiments] Although the illustrated embodiments have been described above, the present invention is not limited to these embodiments.

[0129] For example, the above embodiment utilizes multiple generative models. However, the recipe recognition system of the present invention is not limited to such a configuration. For example, the recipe recognition system of the present invention may combine some of these multiple generative models into a single model, or it may use a combination of processing by generative models and processing that does not use generative models (for example, processing based on correlation data) instead of processing performed by those generative models.

[0130] Furthermore, in the above embodiment, for example, the recipe presentation unit 17 presents the recognized recipe to the user. However, the present invention is not limited to this configuration. Specifically, when transmitting the recognized recipe to another system (such as an automated cooking system), or when publishing it on a webpage, in a book, etc., it is not always necessary to present the recognized recipe to the user.

[0131] Furthermore, in the above embodiment, for example, the case in which the recipe recognition system S is a single computer system was described. However, the present invention also includes a recipe recognition program for causing any one or more computers to execute the aforementioned recipe recognition method, and a recording medium for which the program is recorded and which can be read by a computer used by a user or the like. [Explanation of Symbols]

[0132] 1...Server, 10...Basic recipe recognition unit, 11...Recipe information storage unit, 12...Excluded ingredient recognition unit, 13...Alternative ingredient recognition unit, 14...Alternative recipe recognition unit (cooking point recognition unit), 15...Specified parameter recognition unit, 16...Cooking parameter recognition unit (Cooking parameter comparison unit), 17...Recipe presentation unit, 2...User terminal, 20...Input unit, 21...Output unit, S...Recipe recognition system.

Claims

1. A recipe recognition method that recognizes recipes that use other ingredients instead of specific ingredients, The basic recipe recognition unit recognizes a basic recipe, which is the recipe for the desired dish, based on the user's input. The excluded ingredient recognition unit recognizes excluded ingredients, which are ingredients to be excluded from the basic recipe, based on the user's input. The alternative ingredient recognition unit recognizes, based on the basic recipe and the excluded ingredients, an alternative ingredient that can be used in place of the excluded ingredients in the basic recipe, The alternative recipe recognition unit includes the step of recognizing an alternative recipe, which is a recipe for making the desired dish using the alternative ingredients instead of the excluded ingredients, based on the desired dish, excluded ingredients, and alternative ingredients. The recipe recognition method is characterized in that the substitute ingredients consist of a combination of multiple ingredients.

2. In the recipe recognition method described in claim 1, The recipe recognition method is characterized in that the substitute ingredient recognition unit recognizes the substitute ingredient using a first generation model that, when it receives second data including the desired dish and the excluded ingredients by referring to a plurality of first data representing the characteristics of the ingredients, outputs third data representing the substitute ingredient.

3. In the recipe recognition method described in claim 1, The recipe presentation unit provides the alternative recipe, The cooking parameter recognition unit recognizes cooking parameters, which are parameters relating to at least one of the following aspects of a dish: taste, aroma, texture, appearance, and nutrition. The cooking parameter comparison unit includes the step of comparing the cooking parameters of the desired dish made by the alternative recipe with the cooking parameters of the desired dish made by the basic recipe, The alternative recipe recognition unit recognizes a plurality of alternative recipes, The recipe recognition method is characterized in that the recipe presentation unit presents, along with each of the multiple alternative recipes, the results of comparing the cooking parameters of the desired dish made by the alternative recipe with the cooking parameters of the desired dish made by the basic recipe.

4. In the recipe recognition method described in claim 3, The specified parameter recognition unit includes the step of recognizing a specified parameter, which is a cooking parameter specified by the user, from among the cooking parameters, based on the user's input. The recipe recognition method is characterized in that the recipe presentation unit presents, with emphasis, an alternative recipe among a plurality of alternative recipes in which the difference between the specified parameters in the desired dish made by the alternative recipe and the specified parameters in the desired dish made by the basic recipe is small, compared to other alternative recipes.

5. In the recipe recognition method described in claim 3, The cooking point recognition unit includes the step of recognizing cooking points that are cooking points for reducing the differences in cooking parameters that have been recognized as being different as a result of the comparison, The recipe recognition method is characterized in that the recipe display unit displays the cooking points along with the alternative recipe.

6. In the recipe recognition method described in claim 3, The cooking parameter recognition unit and the cooking parameter comparison unit, upon inputting a plurality of fourth data which are text representing the cooking parameters for a dish made by a predetermined recipe, receive a plurality of fifth data representing a plurality of recipes, and use a second generation model that outputs a sixth data representing the cooking parameters for each of the plurality of recipes and the result of the comparison of the cooking parameters, to recognize the cooking parameters in the desired dish made by the alternative recipe, the cooking parameters in the desired dish made by the basic recipe, and the result of the comparison. The recipe recognition method is characterized in that the sixth data includes text representing the result of the comparison.

7. A recipe recognition system that recognizes recipes that use other ingredients instead of specific ingredients, A recipe recognition system characterized by being configured to perform the recipe recognition method described in any one of claims 1 to 6.

8. A recipe recognition device that recognizes recipes using other ingredients instead of specific ingredients, A recipe recognition device characterized by being configured to perform the recipe recognition method described in any one of claims 1 to 6.

9. A recipe recognition program that causes a computer to execute a recipe recognition method that recognizes recipes using other ingredients instead of specific ingredients, A recipe recognition program characterized by causing the computer to execute the recipe recognition method described in any one of claims 1 to 6.

10. A recording medium characterized by recording the recipe recognition program described in claim 9, and the recipe recognition program being readable by the computer.

Images