Output device, output method, and program

Abstract

[Problem] Conventionally, target information for specifying a good nutrient balance could not be presented using results of analyzing intestinal bacteria. [Solution] Target information for specifying a good nutrient balance can be presented using results of analyzing intestinal bacteria, through use of an output device 1 comprising a learning information storage unit 111 in which learning information acquired by using two or more pieces of training data having an intestinal bacteria analysis result and target information that is information for specifying a good nutrient balance are stored, a reception unit 32 that receives subject information including intestinal bacteria analysis results for one user, a prediction unit 131 that acquires target information for specifying a nutrient balance to be recommended to one user by using the subject information and the learning information, and an output unit 14 that outputs the target information acquired by the prediction unit 131.

Description

Output device, output method, and program

[0001] The present invention relates to an output device or the like that uses the analysis results of intestinal bacteria to acquire and output target information for identifying a nutrient balance that is favorable for a user.

[0002] Conventionally, there has been a program that uses time-series blood glucose level data and a meal record to suggest recommended meal patterns for the next meal (see Patent Document 1).

[0003] JP 2023-19885 A

[0004] It is known that each person's body responds differently to food, and that there is a nutritional balance that is right for each individual.

[0005] However, conventional technologies have not been able to provide information for identifying a good nutrient balance for each individual. In particular, it has been difficult to provide a nutritional balance for each individual that will prevent blood sugar levels from rising.

[0006] Furthermore, in the prior art, it was not possible to present the type of a user using the analysis results of the user's intestinal bacteria.

[0007] Furthermore, in the prior art, it was not possible to use the analysis results of the user's intestinal bacteria to suggest meals that are likely to raise blood sugar levels or meals that are unlikely to raise blood sugar levels for the user.

[0008] The output device of the first invention is an output device comprising: a learning information storage unit that stores learning information acquired using two or more teacher data items having intestinal bacteria analysis results and target information, which is information for identifying a good nutrient balance; a reception unit that receives target information including the intestinal bacteria analysis results of a single user; a prediction unit that uses the target information and the learning information to acquire target information that identifies the nutrient balance to be recommended to the single user; and an output unit that outputs the target information acquired by the prediction unit.

[0009] With this configuration, it is possible to present target information for identifying a good nutrient balance using the results of the intestinal bacteria analysis.

[0010] In addition, the output device of the second invention is an output device in which, compared to the first invention, the teacher data has one or more types of information selected from the user's health information corresponding to the intestinal bacteria analysis results and lifestyle information related to the user's lifestyle, and the receiving unit receives target information selected from the intestinal bacteria analysis results of a single user and one or more types of information selected from the user's health information and lifestyle information.

[0011] With this configuration, it is possible to present target information for identifying a better nutrient balance using the results of the intestinal bacteria analysis and one or more types of information from among health information and lifestyle information.

[0012] Furthermore, the output device of the third invention, compared to the first or second invention, further includes a candidate management unit in which two or more candidate pieces of information indicating candidates for information that identify a good nutrient balance are stored, the prediction unit obtains blood glucose levels for each of the two or more pieces of candidate information using the candidate information, target information, and learning information, and obtains candidate information corresponding to the blood glucose level that satisfies the selection conditions, and the output unit is an output device that outputs the candidate information obtained by the prediction unit.

[0013] With this configuration, it is possible to present target information for identifying a good nutrient balance from two or more candidates using the results of the intestinal bacteria analysis.

[0014] Furthermore, the output device of the fourth invention is an output device in which, compared to the first or second invention, the target information is a balance identifier that identifies the nutrient balance or balance information that indicates the nutrient balance, and the output unit outputs balance information identified by the balance identifier that is the target information, or balance information that is the target information.

[0015] With this configuration, balance information showing a good nutrient balance can be presented using the analysis results of intestinal bacteria.

[0016] In addition, the output device of the fifth invention, compared to any one of the first to fourth inventions, further includes a menu acquisition unit that acquires balance information indicating nutrient balance based on objective information and acquires menu information that identifies a meal menu corresponding to the balance information, and the output unit is an output device that outputs the menu information instead of or in addition to the objective information.

[0017] With this configuration, it is possible to identify a favorable nutrient balance using the results of the intestinal bacteria analysis and to suggest a meal menu based on that nutrient balance.

[0018] In addition, the output device of the sixth invention is an output device according to any one of the first to fifth inventions, further comprising a receiving unit that receives reference meal information relating to a reference meal, obtains balance information that identifies the nutrient balance corresponding to the reference meal information, and a difference obtaining unit that obtains difference information relating to the difference between the balance information and the balance information that is the target information obtained by the prediction unit, and the output unit outputs the difference information.

[0019] With this configuration, it is possible to present the user with difference information for improving the diet that the user has taken.

[0020] Furthermore, the output device of the seventh invention is an output device in which, compared to any one of the first to sixth inventions, the learning information is a learning model obtained by a machine learning learning process using two or more teacher data, and the prediction unit performs a machine learning prediction process using the target information and the learning model to obtain the target information.

[0021] With this configuration, it is possible to use the results of the intestinal bacteria analysis to present target information for identifying a good nutrient balance through machine learning.

[0022] The output device of the present invention is also an output device that includes a learning information storage unit that stores learning information acquired using two or more pieces of teacher data that include intestinal bacteria analysis results and objective information, which is information for identifying a good nutrient balance; a type management unit that stores two or more pieces of type information that include nutrient balance conditions, which are conditions related to the objective information for identifying the nutrient balance, and a type identifier that identifies the user's type; a reception unit that receives target information including the intestinal bacteria analysis results of a single user; a prediction unit that uses the target information and the learning information to acquire objective information that identifies the nutrient balance recommended to a single user; a type determination unit that acquires a type identifier corresponding to the objective information; and an output unit that outputs the type identifier.

[0023] With this configuration, the type of the user can be presented using the analysis results of the user's intestinal bacteria.

[0024] In addition, the output device of the present invention, in addition to the above invention, further includes a type information acquisition unit that acquires type information of a user including the type identifier acquired by the type determination unit, and the output unit is an output device that outputs the type information.

[0025] With this configuration, the type of the user can be presented using the analysis results of the user's intestinal bacteria.

[0026] In addition, the output device of the present invention is an output device in which, compared to the above invention, the receiving unit receives reference meal information related to a reference meal, the prediction unit acquires objective information corresponding to the type determined by the type determination unit, acquires balance information specifying the nutrient balance corresponding to the reference meal information, and further includes a difference acquisition unit that acquires difference information related to the difference between the balance information and the balance information that is the objective information acquired by the prediction unit, and the output unit outputs the difference information.

[0027] According to the present invention, the output device further includes: a learning information storage unit that stores second learning information acquired using two or more pieces of training data each having an intestinal bacteria analysis result, food information, and a blood glucose level; a menu management unit that stores two or more pieces of menu management information each having a food identifier that identifies a food and food information having information on one or more nutrients contained in the food; a second prediction unit that uses the intestinal bacteria analysis result of the user, the food information included in each of the two or more pieces of menu management information, and the second learning information to acquire blood glucose levels of the user when the user eats foods identified by each of the two or more pieces of food information; a recommendation determination unit that acquires recommendation information for foods using blood glucose levels for each of the two or more pieces of food information acquired by the second prediction unit; and an output unit that outputs the recommendation information acquired by the recommendation determination unit.

[0028] With this configuration, the analysis results of the user's intestinal bacteria can be used to present meals that are likely to raise blood sugar levels or meals that are unlikely to raise blood sugar levels for the user.

[0029] The output device according to the present invention can present target information for identifying a good nutrient balance using the analysis results of intestinal bacteria.

[0030] FIG. 1 is a conceptual diagram of an information system A in a first embodiment; FIG. 2 is a block diagram of the information system A; FIG. 3 is a flowchart illustrating an operation example of the output device 1; FIG. 4 is a flowchart illustrating a first example of the learning process; FIG. 5 is a flowchart illustrating a third example of the learning process; FIG. 6 is a flowchart illustrating a first example of the prediction process; FIG. 7 is a flowchart illustrating a second example of the prediction process; FIG. 8 is a flowchart illustrating a third example of the learning process; FIG. 9 is a flowchart illustrating a fourth example of the prediction process; FIG. 10 is a flowchart illustrating a fifth example of the prediction process;

[0031] Hereinafter, embodiments of an output device and the like will be described with reference to the drawings. Note that components with the same reference numerals in the embodiments perform similar operations, and therefore repeated description may be omitted.

[0032] (Embodiment 1) In this embodiment, an output device is described that uses the analysis results of a user's intestinal bacteria and learning information to acquire and output target information, which is information for identifying a good nutrient balance for the user.

[0033] In addition, in this embodiment, an output device is described that acquires and outputs target information using one or more types of information from the user's health information and the user's lifestyle information in addition to the analysis results of the user's intestinal bacteria.

[0034] In addition, in this embodiment, an output device is described that predicts blood glucose levels using the analysis results of the user's intestinal bacteria and two or more nutrient balance candidates, and obtains and outputs nutrient balance candidates corresponding to blood glucose levels that satisfy selection conditions.

[0035] In this embodiment, an output device that uses the acquired nutrient balance information to propose meal menus will be described.

[0036] Furthermore, in this embodiment, an output device that presents the difference between the acquired nutrient balance information and the nutrient balance obtained from the user's diet information will be described.

[0037] In this specification, information X being associated with information Y means that information Y can be obtained from information X, or information X can be obtained from information Y, and the method of association is not important. Information X and information Y may be linked, may exist in the same buffer, information X may be included in information Y, or information Y may be included in information X, etc.

[0038] Furthermore, in this specification, selecting information Z means obtaining information Z, obtaining a pointer to information Z, obtaining the ID of information Z, setting a flag on information Z, etc., and it is sufficient if information Z can be accessed.

[0039] 1 is a conceptual diagram of an information system A according to this embodiment. The information system A includes an output device 1 and one or more terminal devices 2.

[0040] The output device 1 is a device that acquires and outputs target information. The output device 1 is, for example, a cloud server or an ASP server, but the type is not important. The output device 1 may also be a standalone device. In other words, the output device 1 may also be a terminal. When the output device 1 is a terminal, the output device 1 is, for example, a so-called personal computer, a smartphone, or a tablet terminal, but the type is not important.

[0041] The terminal device 2 is a terminal used by a user. The terminal device 2 may be, for example, a personal computer, a smartphone, or a tablet terminal, but the type of the terminal device 2 is not important.

[0042] The output device 1 and the two or more terminal devices 2 are usually capable of communicating with each other via a network such as the Internet.

[0043] 2 is a block diagram of an information system A according to this embodiment. The output device 1 constituting the information system A includes a storage unit 11, a reception unit 12, a processing unit 13, and an output unit .

[0044] The storage unit 11 includes a learning information storage unit 111, a candidate management unit 112, and a menu management unit 113. The processing unit 13 includes a learning unit 130, a prediction unit 131, a menu acquisition unit 132, and a difference acquisition unit 133.

[0045] The terminal device 2 includes a terminal storage unit 21 , a terminal reception unit 22 , a terminal processing unit 23 , a terminal transmission unit 24 , a terminal reception unit 25 , and a terminal output unit 26 .

[0046] Various types of information are stored in the storage unit 11 constituting the output device 1. The various types of information include, for example, two or more pieces of teacher data (described later), learning information (described later), candidate information (described later), and menu management information (described later).

[0047] The learning information storage unit 111 stores learning information. The learning information is information acquired using two or more pieces of teacher data.

[0048] Teacher data is information for acquiring learning information. Teacher data is, for example, information for creating a learning model (described later), information for creating a correspondence table (described later), or information for creating an arithmetic formula (described later). Note that teacher data may also be information constituting learning information. Teacher data includes intestinal bacteria analysis results and objective information. It is preferable that teacher data includes one or more types of information selected from health information and lifestyle information in addition to the intestinal bacteria analysis results and objective information. The intestinal bacteria analysis results, health information, and lifestyle information contained in one piece of teacher data are information about the same person.

[0049] The intestinal bacteria analysis result (hereinafter also simply referred to as "analysis result") is usually information indicating the results of analyzing a person's intestinal bacteria. The analysis result is information regarding one or more bacteria present in the intestine. The analysis result is, for example, information regarding the amount of each bacterium. The analysis result is, for example, information regarding the proportion of each bacterium. The analysis result is, for example, information regarding the amount of a total of 41 bacteria, enterobacteria PCA0 to PCA40. The analysis result is, for example, information indicating the amount of Prevotella, information indicating the amount of Bifidobacterium, information indicating the amount of Megamonas, information indicating the amount of Ruminococcus, information indicating the amount of Akkermansia, and information indicating the amount of Fusobacterium. The analysis result is, for example, information obtained by compressing information indicating the amount of each enterobacteria by dimensionality reduction.

[0050] The target information is information for identifying a good nutrient balance for the user. The target information can be said to be information that the output device 1 wants to obtain. The target information is, for example, balance information or blood glucose level. The balance information is information that identifies the nutrient balance. The balance information is, for example, an identifier (usually an ID) that identifies information indicating the nutrient balance, or information indicating the nutrient balance. The target information may also be an identifier that identifies one of two or more types of nutrient balance.

[0051] Information indicating nutrient balance is, for example, information indicating the amount or proportion of protein, lipids, carbohydrates, sugars, and dietary fiber (e.g., g or %). Information indicating nutrient balance is, for example, information indicating the amount or proportion of each nutrient that makes up the three major nutrients (protein, lipids, carbohydrates) (e.g., g or %), or information indicating the amount or proportion of each nutrient that makes up the five major nutrients (protein, lipids, carbohydrates, minerals, vitamins) (e.g., g or %). Information indicating nutrient balance is, for example, the amount of energy (e.g., kcal). Information indicating nutrient balance is, for example, a vector (amount of energy, amount of carbohydrates, amount of lipids, amount of dietary fiber, amount of protein, amount of sugars).

[0052] Health information is information about a living organism. Health information can be said to be information related to the care of a living organism. The route or method of obtaining the health information is not important. Health information typically includes health checkup information. Health information also includes, for example, data that can be obtained by a device or data that can be measured at home (e.g., weight, blood pressure, pulse rate, etc.). Health checkup information is information about the results of a health checkup. Health information is, for example, information about blood test results, physical measurement results, or user attribute values. The health information may be, for example, age, sex, smoking status, height, weight, BMI, history of obesity, waist circumference, waist / hip ratio, hip circumference, blood pressure, systolic pulse, ALT (GPT), basophils, C-peptide, eosinophils, HbA1c, HDL, total hemoglobin (ctHb), insulin, IL-6, LDL, lymphocytes, MCH, MCV, monocytes, neutrophils, non-HDL, nucleated red blood cells, Ht (hematocrit), hematocrit, platelets, red blood cell count, red blood cell volume particle size distribution width, thyroid-stimulating hormone, total cholesterol, total cholesterol to HDL ratio, triglycerides, white blood cell count, or baseline blood glucose level.

[0053] Lifestyle information is information about the user's life. The lifestyle information is, for example, information about sleep, information about meals, or information about activities. The sleep information is, for example, sleep time, sleep efficiency, or the time from waking up to going to bed. The meal information is, for example, the amount of carbohydrates ingested, the amount of dietary fiber ingested, or the amount of carbohydrates in the previous meal. The activity information is, for example, the time spent exercising or the time spent engaged in vigorous activity. The lifestyle information may be information about life at two or more points in time. That is, the lifestyle information is, for example, a collection of information about sleep at two or more points in time, information about meals at two or more points in time, or information about activities at two or more points in time. The lifestyle information is, for example, the sleep time for each day for two or more days, the total sleep time for two or more days, the amount of carbohydrates ingested at each of two or more meals, the total amount of carbohydrates ingested at two or more meals, the time spent exercising for each day for two or more days, or the total time spent exercising for two or more days.

[0054] The learning information is information used to acquire target information, such as a learning model, a correspondence table, or an arithmetic formula.

[0055] A learning model is information obtained by a machine learning learning process using two or more pieces of training data. Here, the training data is information in which the analysis results are used as explanatory variables and the objective information is used as the objective variable. The training data may also be information in which one or more types of information selected from health information and lifestyle information and the analysis results are used as explanatory variables and the objective information is used as the objective variable. A learning model is information used in a machine learning prediction process. A learning model may also be called a learner, classifier, classification model, etc. Any machine learning algorithm may be used, such as deep learning, random forest, decision tree, SVM, SVR, etc. Furthermore, for machine learning, various machine learning functions such as the TensorFlow (registered trademark) library, the random forest module of the R language, and TinySVM, as well as various existing libraries, can be used.

[0056] The correspondence table is a table that manages the correspondence between explanatory variables and objective information. The correspondence table has two or more pieces of correspondence information, and the correspondence information is information that indicates the correspondence between explanatory variables and objective information. The explanatory variables are, for example, analysis results. The explanatory variables include, for example, one or more types of information from health information and lifestyle information, and the analysis results.

[0057] The arithmetic formula is an arithmetic formula for substituting one or more parameters contained in the analysis results to obtain a predicted blood glucose level. For example, the arithmetic formula is an arithmetic formula for substituting one or more parameters contained in the analysis results and one or more types of information selected from health information and lifestyle information to output a predicted blood glucose level.

[0058] The candidate management unit 112 stores two or more pieces of candidate information. The candidate information is a candidate for information identifying a favorable nutrient balance. The candidate information is typically balance information. Here, balance information typically refers to information indicating nutrient balance or an identifier that identifies the information indicating nutrient balance. The candidate information may include, for example, information indicating the amount of energy and the amount or proportion of each nutrient constituting the three major nutrients (protein, lipids, and carbohydrates), information indicating the amount of energy and the amount or proportion of each nutrient constituting the five major nutrients (protein, lipids, carbohydrates, minerals, and vitamins), information indicating the amount of energy to be ingested in a unit period (e.g., one day, one meal) and the amount or proportion of each nutrient constituting the three major nutrients, or information indicating the amount of energy to be ingested in a unit period (e.g., one day) and the amount or proportion of each nutrient constituting the five major nutrients. The amount or proportion of carbohydrates may also include the amount or proportion of sugars contained in carbohydrates and the amount or proportion of dietary fiber contained in carbohydrates.

[0059] The menu management unit 113 stores two or more pieces of menu management information. The menu management information is information that manages menu information corresponding to balance information. The menu information is information that specifies a meal menu. A menu may be one or more dishes, one or more foods, or one or more dishes and one or more foods. In other words, the menu information may be information that specifies an item before cooking, information that specifies an item after cooking, or information that specifies an item before cooking and an item after cooking.

[0060] Menu management information is, for example, food information. Food information is information that identifies a food. Food information includes, for example, two or more pieces of nutrient information. Nutrient information is information about the energy and nutrients contained in a food. Nutrient information is, for example, the amount of energy contained in a food and the amount of one or more nutrients. Food information includes, for example, the name of a food and an image of the food.

[0061] The reception unit 12 receives various information and instructions. The reception unit 12 receives various information and instructions from, for example, the terminal device 2. The reception unit 12 receives various information and instructions from, for example, a user. The various information and instructions are, for example, target information, meal information, and learning instructions. The meal information here is, for example, reference meal information.

[0062] Here, "reception" typically refers to the reception of information transmitted via a wired or wireless communication line, but may also be a concept that includes the reception of information input from an input device such as a keyboard, mouse, or touch panel, or the reception of information read from a recording medium such as an optical disk, magnetic disk, or semiconductor memory.

[0063] The target information is information about a target user to whom the purpose information is presented. The target information includes an analysis result of the user. In addition to the analysis result, the target information may include one or more types of information of the user's health information and lifestyle information.

[0064] Dietary information is information related to a meal. Dietary information is, for example, information specifying a meal, dietary balance information, or an image of a meal (dish). Dietary balance information includes, for example, the amount or proportion of each of two or more nutrients ingested through a meal. Each of the two or more nutrients is, for example, carbohydrates, lipids, dietary fiber, sugar, and protein. Such balance information includes, for example, the amount of energy.

[0065] The reference meal information is reference meal information. The reference meal information is usually information that serves as a base when obtaining difference information. The reference meal information is, for example, information about a meal consumed by a user, information about a meal desired by a user, or information about a meal consumed by an average user.

[0066] The processing unit 13 performs various types of processing, such as processing performed by the learning unit 130, the prediction unit 131, the menu acquisition unit 132, or the difference acquisition unit 133.

[0067] The learning unit 130 acquires learning information using two or more pieces of teacher data in the storage unit 11. The two or more pieces of teacher data are stored in an external device (not shown) and may be information temporarily stored in the storage unit 11.

[0068] The learning information acquired by the learning unit 130 is, for example, a learning model, a correspondence table, or an arithmetic expression. Each of these will be explained below. When the learning information is a learning model

[0069] The learning unit 130, for example, provides two or more pieces of teacher data from the storage unit 11 to a machine learning learning module, executes the module, and acquires a learning model. The learning model is provided to a machine learning prediction module together with the received target information, and is used to acquire target information. (1-1) When the target information is a blood glucose level and balance information is used

[0070] The learning unit 130 performs machine learning learning processing using, for example, two or more sets of teacher data in which the balance information and the target information are used as explanatory variables and the blood glucose level is used as the objective variable, and acquires a learning model. (1-2) In the case where the objective information is the blood glucose level and the balance information is not used

[0071] The learning unit 130 performs machine learning learning processing using, for example, two or more pieces of training data in which the target information is an explanatory variable and the blood glucose level is an objective variable, and acquires a learning model. (1-3) When the objective information is a balance identifier that identifies the nutrient balance

[0072] The learning unit 130 performs a machine learning learning process using, for example, two or more sets of training data in which the target information is an explanatory variable and the balance identifier is a response variable, and acquires a learning model. (1-4) When the response information is information indicating nutrient balance

[0073] The learning unit 130 performs machine learning learning processing using, for example, two or more pieces of training data in which the target information is an explanatory variable and a vector representing information indicating nutrient balance is an objective variable, and acquires a learning model. Such vectors are, for example, (amount of energy, amount of carbohydrates, amount of fat, amount of dietary fiber, amount of protein). (2) When the learning information is a correspondence table (2-1) When one piece of training data is one piece of correspondence information

[0074] For example, the learning unit 130 vectorizes explanatory variables contained in the teacher data for each of two or more pieces of teacher data in the storage unit 11. Next, the learning unit 130 acquires correspondence information containing the vector and objective information contained in the teacher data for each of two or more pieces of teacher data. Next, the learning unit 130 creates a correspondence table containing the acquired two or more pieces of correspondence information. Note that the vectors are, for example, (amount of enterobacteria PCA0, amount of enterobacteria PCA1, ..., amount of PCA40, height, weight, ..., blood pressure, ..., red blood cell count, ...). The objective information is, for example, information indicating blood glucose level, balance identifier, or nutrient balance. (2-2) When one piece of correspondence information is created using one or two or more explanatory variables paired with the same objective information

[0075] For example, the learning unit 130 vectorizes the explanatory variables of the teacher data for each of two or more teacher data in the storage unit 11. Next, for example, for each of two or more different pieces of objective information, the learning unit 130 acquires a representative vector of the vectors of one or more explanatory variables paired with the same piece of objective information. Next, for example, the learning unit 130 acquires correspondence information having a representative vector and the objective information for each of two or more different pieces of objective information. Next, the learning unit 130 constructs a correspondence table having the acquired two or more pieces of correspondence information. (3) When the learning information is an arithmetic formula

[0076] For example, the learning unit 130 vectorizes the explanatory variables of the teacher data for each of the two or more teacher data in the storage unit 11. Next, for each of the two or more teacher data, the learning unit 130 acquires an arithmetic formula that takes two or more elements of the vector obtained from the explanatory variables as input and outputs the blood glucose level, which is the target information. Note that the learning unit 130 acquires an arithmetic formula using the least squares method for each of the two or more teacher data, that takes two or more elements of the vector obtained from the explanatory variables as input and outputs the blood glucose level, which is the target information. The method of acquiring the arithmetic formula is not important.

[0077] The prediction unit 131 acquires objective information using the received target information and learning information of a single user. For example, the prediction unit 131 acquires objective information that specifies the nutrient balance to be recommended to the single user using the target information and learning information. There are various possible methods for the prediction unit 131 to acquire objective information using the target information and learning information. The various methods will be described below. (1) When candidate information is not used (1-1) When learning information is a learning model

[0078] The prediction unit 131 performs machine learning prediction processing using the target information of one user and the learning model in the learning information storage unit 111 to acquire target information. More specifically, the prediction unit 131 provides the target information of one user and the learning model to a machine learning prediction module, executes the prediction module, and acquires target information. (1-2) When the learning information is a correspondence table (1-2-1) When one piece of correspondence information with the highest similarity is selected

[0079] The prediction unit 131 acquires a vector having, as elements, each of two or more pieces of information contained in the target information of one user. Next, the prediction unit 131 calculates the similarity between the vector and a vector included in each of two or more pieces of correspondence information contained in the correspondence table. Next, the prediction unit 131 acquires the target information contained in the correspondence information corresponding to the vector with the greatest similarity. (1-2-2) When correspondence information satisfying the selection condition is selected

[0080] The prediction unit 131 acquires a vector having, as elements, each of two or more pieces of information contained in the target information of one user. Next, the prediction unit 131 calculates the similarity between the vector and a vector included in each of two or more pieces of correspondence information contained in the correspondence table. Next, the prediction unit 131 selects one or more pieces of correspondence information whose similarity satisfies the selection condition. Next, the prediction unit 131 acquires target information contained in each of the one or more pieces of correspondence information. Next, the prediction unit 131 acquires a representative value of the acquired one or more pieces of target information. The representative value of the target information is, for example, the most common piece of target information, the average value of the target information, or the median value of the target information. (1-3) When the learning information is an arithmetic formula

[0081] The prediction unit 131 provides two or more pieces of information contained in the target information of one user to an arithmetic expression, executes the arithmetic expression, and obtains the blood glucose level, which is the target information.

[0082] When the prediction unit 131 uses candidate information, the prediction unit 131 acquires blood glucose levels for each of two or more pieces of candidate information using the candidate information, target information, and learning information. Next, the prediction unit 131 acquires candidate information corresponding to blood glucose levels that satisfy a selection condition. The selection condition is usually that the blood glucose level is the lowest value, but for example, the blood glucose level may be equal to or less than a threshold value. When the selection condition is that the blood glucose level is equal to or less than a threshold value, the prediction unit 131 may acquire two or more pieces of candidate information.

[0083] Hereinafter, the process of acquiring blood glucose levels differs depending on the learning information, but the process of acquiring candidate information corresponding to blood glucose levels that satisfy the selection conditions is the same. Therefore, the process of acquiring blood glucose levels will be explained below for each of the three types of learning information. (2-1) When the learning information is a learning model

[0084] The prediction unit 131 provides the candidate information, the target information, and the learning model for each of two or more pieces of candidate information to a machine learning prediction module, executes the prediction module, and obtains the blood glucose level. In this case, the machine learning algorithm may be any of deep learning, random forest, decision tree, SVR, etc.

[0085] The learning model here is information acquired by the learning unit 130 using two or more pieces of training data with the balance information and the target information as explanatory variables and the blood glucose level as the objective variable. (2-2) When the learning information is a correspondence table (2-2-1) When the correspondence information with the highest similarity is adopted

[0086] For each of two or more pieces of candidate information, the prediction unit 131 constructs a vector whose elements are each of two or more pieces of information contained in the candidate information and the target information. The prediction unit 131 calculates the similarity between the vector for each of two or more pieces of candidate information and the vector contained in each of two or more pieces of corresponding information in the correspondence table, and obtains the blood glucose level contained in the corresponding information with the greatest similarity. (2-2-2) When corresponding information satisfying the selection conditions is adopted

[0087] For each of two or more pieces of candidate information, the prediction unit 131 constructs a vector whose elements are the two or more pieces of information contained in the candidate information and the target information. The prediction unit 131 calculates the similarity between the vector for each of two or more pieces of candidate information and the vector contained in each of two or more pieces of corresponding information in the correspondence table, and acquires the blood glucose level contained in one or more pieces of corresponding information whose similarity satisfies the selection condition. Next, the prediction unit 131 acquires a representative value (e.g., average value, median) of the acquired one or more blood glucose levels. Note that the selection condition here is, for example, that the similarity is equal to or greater than a threshold value. (2-3) When the learning information is an arithmetic formula

[0088] For each of the two or more pieces of candidate information, the prediction unit 131 provides the two or more pieces of information contained in the candidate information and the target information to an arithmetic expression, executes the arithmetic expression, and obtains the blood glucose level.

[0089] The menu acquisition unit 132 acquires balance information indicating the nutrient balance based on the objective information, and acquires menu information specifying a meal menu corresponding to the balance information from the menu management unit 113 .

[0090] The difference acquisition unit 133 acquires balance information that specifies the nutrient balance corresponding to the received reference dietary information. Next, the difference acquisition unit 133 acquires difference information regarding the difference between the balance information and the balance information acquired by the prediction unit 131. The difference information is, for example, information indicating the difference in the amount or proportion of each of two or more nutrients. The difference information is, for example, information indicating the difference in energy amount. The difference information is, for example, food information that specifies foods that fill the difference in the amount or proportion of each of two or more nutrients, or menu information that specifies a meal menu that fills the difference in the amount or proportion of each of two or more nutrients. The difference information is, for example, food information that specifies foods that fill the difference in energy amount, or menu information that specifies a meal menu that fills the difference in energy amount.

[0091] The difference acquisition unit 133 acquires, for example, the amount of energy corresponding to each of one or more dish names and the amounts of two or more nutrients corresponding to each of one or more dish names contained in the reference meal information from a recipe database in the storage unit 11 or an external device (not shown). The recipe database is a database that stores the amount of energy and the amounts of two or more nutrients in pairs with each of one or more dish names.

[0092] Furthermore, the difference acquisition unit 133 provides, for example, a dish image, which is reference dietary information, and a learning model to a machine learning prediction module to acquire the amount of energy and the amounts of two or more nutrients corresponding to the dish image. Note that this learning model is information acquired by performing a machine learning learning process using two or more pieces of training data containing the dish image, the energy amount, and the amounts of two or more nutrients. The prediction module is a module that uses the dish image and the learning model to output the amount of energy and the amounts of two or more nutrients of the dish.

[0093] The received reference meal information may be the amount of each of two or more nutrients or the name of a menu item. When the reference meal information is one or more menu items, the difference acquisition unit 133 acquires the amount of each nutrient corresponding to each of the one or more menu items and acquires balance information using the amount of each nutrient. Next, the difference acquisition unit 133 acquires difference information regarding the difference between the balance information and the balance information acquired by the prediction unit 131.

[0094] The output unit 14 outputs the objective information acquired by the prediction unit 131. The objective information is, for example, balance information or blood glucose level. Note that when the prediction unit 131 acquires a balance identifier, the output unit 14 may acquire information indicating the nutrient balance identified by the balance identifier from the storage unit 11 or a device (not shown), and output the information indicating the nutrient balance.

[0095] The output unit 14 outputs, for example, the candidate information acquired by the prediction unit 131. The candidate information is balance information. Such candidate information can be said to be target information.

[0096] The output unit 14 outputs, for example, balance information identified by a balance identifier that is objective information, or balance information that is objective information.

[0097] The output unit 14 outputs, for example, the menu information acquired by the menu acquisition unit 132 instead of or in addition to the objective information.

[0098] The output unit 14 outputs the difference information acquired by the difference acquisition unit 133, for example, instead of or in addition to the objective information.

[0099] Various types of information are stored in the terminal storage unit 21 of the terminal device 2. The various types of information include, for example, a user identifier.

[0100] The terminal receiving unit 22 receives various types of information, instructions, etc. Examples of the various types of information, instructions, etc. include subject information, meal information, and study instructions.

[0101] Here, acceptance is a concept that includes acceptance of information input from input devices such as a keyboard, mouse, or touch panel, receipt of information transmitted via a wired or wireless communication line, and acceptance of information read from recording media such as an optical disk, magnetic disk, or semiconductor memory.

[0102] The means for inputting various information and instructions may be any means, such as a touch panel, keyboard, mouse, or menu screen.

[0103] The terminal processing unit 23 performs various types of processing. For example, the various types of processing are processing to convert received information, instructions, etc. into information, instructions, etc. with a structure to be transmitted. For example, the various types of processing are processing to convert received information into information with a structure to be output.

[0104] The terminal transmitting unit 24 transmits various types of information, instructions, etc. to the output device 1. The various types of information, instructions, etc. are, for example, object information, meal information, and study instructions.

[0105] The terminal receiving unit 25 receives various types of information, such as objective information, menu information, or difference information.

[0106] The terminal output unit 26 outputs various types of information, such as objective information, menu information, or difference information.

[0107] Here, output is a concept that includes displaying on a display, projection using a projector, printing on a printer, sound output, transmission to an external device, storage on a recording medium, and delivery of processing results to other processing devices or other programs.

[0108] The storage unit 11, the learning information storage unit 111, the candidate management unit 112, the menu management unit 113, and the terminal storage unit 21 are preferably non-volatile recording media, but may also be realized by volatile recording media.

[0109] There is no restriction on the process by which information is stored in the storage unit 11 etc. For example, information may be stored in the storage unit 11 etc. via a recording medium, information transmitted via a communication line etc. may be stored in the storage unit 11 etc., or information input via an input device may be stored in the storage unit 11 etc.

[0110] The reception unit 12 is preferably realized by a wireless or wired communication means, but may also be realized by a means for receiving broadcasts, a device driver for an input means such as a touch panel or keyboard, or control software for a menu screen.

[0111] The processing unit 13, the prediction unit 131, the menu acquisition unit 132, and the device processing unit 23 can typically be realized by a processor, memory, etc. The processing procedures of the processing unit 13, etc. are typically realized by software, and the software is recorded on a recording medium such as a ROM. However, they may also be realized by hardware (dedicated circuitry). The processor may be a CPU, MPU, GPU, etc., and the type does not matter.

[0112] The output unit 14 is usually realized by a wireless or wired communication means, but may also be realized by driver software for an output device such as a display or speaker, or by a combination of driver software for an output device and the output device.

[0113] The terminal reception unit 22 can be realized by a device driver for an input means such as a touch panel or a keyboard, or control software for a menu screen.

[0114] The terminal transmitting unit 24 is usually realized by a wireless or wired communication means, but may also be realized by a broadcasting means.

[0115] The terminal receiving unit 25 is usually realized by a wireless or wired communication means, but may also be realized by a means for receiving broadcasts.

[0116] The terminal output unit 26 may or may not be considered to include an output device such as a display, a speaker, etc. The terminal output unit 26 may be realized by driver software for an output device, or by a combination of driver software for an output device and the output device, etc.

[0117] Next, an example of the operation of the output device 1 will be described with reference to the flowchart of FIG.

[0118] (Step S301) The receiving unit 12 determines whether or not a learning instruction has been received. If a learning instruction has been received, the process proceeds to step S302, and if not, the process proceeds to step S303.

[0119] (Step S302) The learning unit 130 performs a learning process. Return to step S301. An example of the learning process will be described with reference to the flowcharts of FIGS.

[0120] (Step S303) The reception unit 12 determines whether or not target information, etc. has been received. If target information, etc. has been received, the process proceeds to step S304, and if not, the process returns to step S301. Note that the target information, etc. refers to the target information, or the target information and meal information.

[0121] (Step S304) The prediction unit 131 performs a prediction process to obtain target information. Examples of the prediction process will be described with reference to the flowcharts of FIGS.

[0122] (Step S305) The output unit 14 outputs the target information acquired in step S304. The information output here is preferably balance information indicating nutrient balance or blood glucose level. If a balance identifier is acquired in step S304, the prediction unit 131 preferably acquires balance information indicating the nutrient balance identified by the balance identifier, and the output unit 14 preferably outputs the balance information indicating the nutrient balance.

[0123] (Step S306) The processing unit 13 determines whether or not to output menu information. If the menu information is to be output, the process proceeds to step S307, and if not, the process returns to step S301.

[0124] (Step S307) The menu acquisition unit 132 acquires balance information indicating the acquired nutrient balance.

[0125] (Step S308) The menu acquisition unit 132 performs a process of acquiring menu information corresponding to the balance information acquired in step S307. An example of such menu acquisition process will be described with reference to the flowchart in Fig. 12. Note that the menu acquisition unit 132 may acquire menu information paired with the balance information acquired in step S307 from the menu management unit 113.

[0126] (Step S309) The output unit 14 outputs the menu information acquired in step S308. The output unit 14 transmits the menu information to the terminal device 2, for example.

[0127] (Step S310) The difference acquisition unit 133 determines whether or not the reception unit 12 also received meal information in step S303. If meal information has also been received, the process proceeds to step S311; if meal information has not also been received, the process returns to step S301.

[0128] (Step S311) The difference acquisition unit 133 acquires difference information. An example of the difference acquisition process will be described with reference to the flowchart of FIG.

[0129] (Step S312) The output unit 14 outputs the difference information acquired in step S311. The process returns to step S301. The output unit 14 transmits the difference information to the terminal device 2, for example.

[0130] In the flowchart of FIG. 3, the output unit 14 may output two or more types of information together.

[0131] In the flowchart of FIG. 3, the process ends when the power is turned off or an interrupt occurs to end the process.

[0132] Next, a first example of the learning process in step S302 will be described with reference to the flowchart in Fig. 4. The first example of the learning process is a case where a learning model is obtained.

[0133] (Step S401) The learning unit 130 assigns 1 to a counter i.

[0134] (Step S402) The learning unit 130 determines whether or not the i-th teacher data exists in the storage unit 11. If the i-th teacher data exists, the process proceeds to step S403; if not, the process proceeds to step S405.

[0135] (Step S403) The learning unit 130 acquires the i-th teacher data from the storage unit 11. Next, the learning unit 130 converts the i-th teacher data into, for example, a vector having a structure to be provided to a learning module, and adds it to a buffer (not shown).

[0136] (Step S404) The learning unit 130 increments the counter i by 1. The process returns to step S402.

[0137] (Step S405) The learning unit 130 provides a set of two or more vectors (for example, a file) in a buffer (not shown) to the learning module in the storage unit 11.

[0138] (Step S406) The learning unit 130 executes the learning module.

[0139] (Step S407) The learning unit 130 acquires a learning model that is a result of the execution of the learning module in step S406, and stores the learning model in the learning information storage unit 111. The process returns to the upper level process.

[0140] Next, a second example of the learning process in step S302 will be described with reference to the flowchart in Fig. 5. The second example of the learning process is a case where a correspondence table having one piece of correspondence information corresponding to one piece of training data is obtained.

[0141] (Step S501) The learning unit 130 assigns 1 to a counter i.

[0142] (Step S502) The learning unit 130 determines whether or not the i-th teacher data exists in the storage unit 11. If the i-th teacher data exists, the process proceeds to step S503; if not, the process returns to the upper level process.

[0143] (Step S503) The learning unit 130 uses the i-th teacher data to construct correspondence information and add it to the correspondence table. Note that the correspondence information includes, for example, correspondence information contained in the teacher data and objective information contained in the teacher data.

[0144] (Step S504) The learning unit 130 increments the counter i by 1. The process returns to step S402.

[0145] Next, a third example of the learning process in step S302 will be described with reference to the flowchart in Fig. 6. In this third example of the learning process, one or more pieces of teacher data corresponding to the same target information are used to construct a single piece of correspondence information, and a correspondence table is obtained. In the flowchart in Fig. 6, it is preferable to sort two or more pieces of teacher data using the target information as a key as an initial process.

[0146] (Step S601) The learning unit 130 assigns 1 to a counter i.

[0147] (Step S602) The learning unit 130 determines whether or not the i-th target information of different information exists. If the i-th target information of different information exists, the process proceeds to step S603; if not, the process returns to the upper level process.

[0148] (Step S603) The learning unit 130 assigns 1 to a counter j.

[0149] (Step S604) The learning unit 130 determines whether the jth element, which is information excluding the target information in the training data, exists. If the jth element exists, the process proceeds to step S605; if not, the process proceeds to step S609. The jth element may be, for example, the amount of energy or the amount of each of two or more nutrients.

[0150] (Step S605) The learning unit 130 acquires information on the j-th element from all training data that includes the i-th target information. The j-th element is

[0151] (Step S606) The learning unit 130 obtains a representative value of the two or more pieces of information obtained in step S605. The representative value of the two or more pieces of information is, for example, the most common piece of information or an average value.

[0152] (Step S607) The learning unit 130 assigns the representative value acquired in step S606 as information on the j-th element paired with the i-th target information.

[0153] (Step S608) The learning unit 130 increments the counter j by 1. The process returns to step S604.

[0154] (Step S609) The learning unit 130 acquires correspondence information indicating the correspondence between the i-th target information and the representative vector, and stores the information in a correspondence table. Note that the information on the elements constituting the representative vector is the information acquired in step S607.

[0155] (Step S610) The learning unit 130 increments the counter i by 1. The process returns to step S602.

[0156] As described above, the learning unit 130 may use the least squares method or the like to obtain an arithmetic formula that takes two or more elements of a vector obtained from information excluding blood glucose levels, which is the target information in the training data, as input and outputs the blood glucose levels, which is the target information.

[0157] Next, a first example of the prediction process in step S304 will be described with reference to the flowchart in Fig. 7. The first example of the prediction process is a case where target information is obtained by machine learning prediction processing.

[0158] (Step S701) The prediction unit 131 acquires the received target information.

[0159] (Step S702) The prediction unit 131 acquires the learning model from the learning information storage unit 111.

[0160] (Step S703) The prediction unit 131 provides the prediction module in the storage unit 11 with the correspondence information acquired in step S701 and the learning model acquired in step S702.

[0161] (Step S704) The prediction unit 131 executes the prediction module.

[0162] (Step S705) The prediction unit 131 acquires the objective information, which is the execution result of the prediction module in step S704, and returns to the upper-level processing.

[0163] Next, a second example of the prediction process in step S304 will be described with reference to the flowchart in Fig. 8. The second example of the prediction process is a case where target information is obtained using a correspondence table.

[0164] (Step S801) The prediction unit 131 acquires the received target information.

[0165] (Step S802) The prediction unit 131 constructs a vector having two or more pieces of information included in the correspondence information acquired in step S801 as elements.

[0166] (Step S803) The prediction unit 131 assigns 1 to a counter i.

[0167] (Step S804) The prediction unit 131 determines whether the i-th correspondence information exists in the correspondence table of the learning information storage unit 111. If the i-th correspondence information exists, the process proceeds to step S805; if not, the process proceeds to step S808.

[0168] (Step S805) The prediction unit 131 acquires the vector included in the i-th correspondence information.

[0169] (Step S806) The prediction unit 131 calculates the similarity between the vector acquired in step S802 and the vector acquired in step S805, and associates the similarity with the i-th correspondence information.

[0170] (Step S807) The prediction unit 131 increments the counter i by 1. The process returns to step S804.

[0171] (Step S808) The prediction unit 131 acquires target information included in the correspondence information corresponding to a similarity that satisfies a selection condition, such as a similarity that is in the top N or higher (N is a natural number equal to or greater than 1), a similarity that is equal to or greater than a threshold.

[0172] (Step S809) The prediction unit 131 determines whether the number of pieces of target information acquired in step S808 is equal to or greater than 2. If it is equal to or greater than 2, the process proceeds to step S810, and if it is not equal to or greater than 2, the process returns to the upper level process.

[0173] (Step S810) The prediction unit 131 acquires a representative of the two or more pieces of target information acquired in step S808. The process returns to the upper level process. The representative of the two or more pieces of target information may be, for example, the target information with the largest number, the average value, or the median value.

[0174] Next, a third example of the prediction process in step S304 will be described with reference to the flowchart in Fig. 9. The third example of the prediction process is a case where target information is obtained by machine learning prediction processing using candidate information.

[0175] (Step S901) The prediction unit 131 acquires the received target information.

[0176] (Step S902) The prediction unit 131 assigns 1 to a counter i.

[0177] (Step S903) The prediction unit 131 determines whether or not the i-th candidate information exists in the candidate management unit 112. If the i-th candidate information exists, the process proceeds to step S904; if not, the process proceeds to step S909.

[0178] (Step S904 ) The prediction unit 131 acquires the i-th candidate information from the candidate management unit 112 .

[0179] (Step S905) The prediction unit 131 acquires a learning model from the learning information storage unit 111. The prediction unit 131 acquires information to be provided to the prediction module using the learning model, the correspondence information acquired in step S901, and the candidate information acquired in step S904.

[0180] (Step S906) The prediction unit 131 provides the information acquired in step S905 to the prediction module and executes the prediction module.

[0181] (Step S907) The prediction unit 131 acquires the blood glucose level that is the execution result of the prediction module in step S906, and associates the blood glucose level with the i-th candidate information.

[0182] (Step S908) The prediction unit 131 increments the counter i by 1. The process returns to step S903.

[0183] (Step S909) The prediction unit 131 acquires one or more pieces of candidate information paired with a blood glucose level that meets a selection condition. The selection condition may be, for example, the lowest blood glucose level, a blood glucose level that is equal to or lower than the bottom N (N is a natural number of 2 or more), a blood glucose level that is equal to or lower than a threshold value, or a blood glucose level that is smaller than a threshold value.

[0184] (Step S910) The prediction unit 131 determines whether the number of pieces of candidate information selected in step S909 is two or more, or one. If the number is two or more, the process proceeds to step S911, and if the number is one, the process returns to the upper process.

[0185] (Step S911) The prediction unit 131 acquires a representative of the two or more pieces of candidate information selected in step S909. The process returns to the upper level process. Note that the representative of the two or more pieces of candidate information is, for example, the candidate information with the largest number of pieces of candidate information.

[0186] Next, a fourth example of the prediction process in step S304 will be described with reference to the flowchart in Fig. 10. The fourth example of the prediction process is a case where target information is obtained from a correspondence table using candidate information.

[0187] (Step S1001) The prediction unit 131 acquires the received target information.

[0188] (Step S1002) The prediction unit 131 assigns 1 to a counter i.

[0189] (Step S1003) The prediction unit 131 determines whether or not the i-th candidate information exists in the candidate management unit 112. If the i-th candidate information exists, the process proceeds to step S1004; if not, the process proceeds to step S1013.

[0190] (Step S1004 ) The prediction unit 131 acquires the i-th candidate information from the candidate management unit 112 .

[0191] (Step S1005) The prediction unit 131 constructs a vector whose elements are two or more pieces of information contained in the target information acquired in step S1001 and one or more pieces of information contained in the i-th candidate information acquired in step S1004.

[0192] (Step S1006) The prediction unit 131 assigns 1 to a counter j.

[0193] (Step S1007) The prediction unit 131 determines whether the jth correspondence information exists in the correspondence table of the learning information storage unit 111. If the jth correspondence information exists, the process proceeds to step S1008; if not, the process proceeds to step S1011.

[0194] (Step S1008) The prediction unit 131 acquires the vector contained in the j-th correspondence information in the correspondence table.

[0195] (Step S1009) The prediction unit 131 obtains the similarity between the vector obtained in step S1005 and the vector obtained in step S1008, and associates the similarity with the j-th correspondence information.

[0196] (Step S1010) The prediction unit 131 increments the counter j by 1. The process returns to step S1007.

[0197] (Step S1011) The prediction unit 131 acquires blood glucose levels from corresponding information associated with similarities that meet the selection conditions. The acquired blood glucose levels are associated with candidate information. The selection conditions here include, for example, a top N similarity (N is a natural number equal to or greater than 1), a similarity equal to or greater than a threshold, or a similarity greater than a threshold.

[0198] (Step S1012) The prediction unit 131 increments the counter i by 1. The process returns to step S1003.

[0199] (Step S1013) The prediction unit 131 acquires one or more pieces of candidate information corresponding to blood glucose levels that meet the selection conditions.

[0200] (Step S1014) The prediction unit 131 determines whether the number of pieces of candidate information selected in step S1013 is two or more, or one. If the number is two or more, the process proceeds to step S1015; if the number is one, the process returns to the upper process.

[0201] (Step S1015) The prediction unit 131 acquires a representative of the two or more pieces of candidate information selected in step S1014. The process returns to the upper level process. Note that the representative of the two or more pieces of candidate information is, for example, the candidate information with the largest proportion.

[0202] Next, a fifth example of the prediction process in step S304 will be described with reference to the flowchart in Fig. 11. The fifth example of the prediction process is a case where target information is obtained by an arithmetic expression using candidate information.

[0203] (Step S1101) The prediction unit 131 acquires the received target information.

[0204] (Step S1102) The prediction unit 131 assigns 1 to a counter i.

[0205] (Step S1103) The prediction unit 131 determines whether or not the i-th candidate information exists in the candidate management unit 112. If the i-th candidate information exists, the process proceeds to step S1104; if not, the process proceeds to step S1109.

[0206] (Step S1104 ) The prediction unit 131 acquires the i-th candidate information from the candidate management unit 112 .

[0207] (Step S1105) The prediction unit 131 acquires an arithmetic expression from the learning information storage unit 111. The prediction unit 131 provides the arithmetic expression with the correspondence information acquired in step S1101 and the candidate information acquired in step S1104.

[0208] (Step S1106) The prediction unit 131 executes the arithmetic expression formed in step S1105, obtains the blood glucose level, and associates the blood glucose level with the i-th candidate information.

[0209] (Step S1107) The prediction unit 131 increments the counter i by 1. The process returns to step S1103.

[0210] (Step S1108) The prediction unit 131 acquires one or more pieces of candidate information paired with a blood glucose level that meets a selection condition. The selection condition may be, for example, the lowest blood glucose level, a blood glucose level that is equal to or lower than the bottom N (N is a natural number of 2 or more), a blood glucose level that is equal to or lower than a threshold value, or a blood glucose level that is smaller than a threshold value.

[0211] (Step S1109) The prediction unit 131 determines whether the number of pieces of candidate information selected in step S1108 is two or more, or one. If the number is two or more, the process proceeds to step S1110; if the number is one, the process returns to the upper process.

[0212] (Step S1110) The prediction unit 131 acquires a representative of the two or more pieces of candidate information selected in step S1108. The process returns to the upper level process. Note that the representative of the two or more pieces of candidate information is, for example, the candidate information with the largest number of pieces of candidate information.

[0213] Next, an example of the menu acquisition process in step S308 will be described with reference to the flowchart of FIG.

[0214] (Step S1201) The menu obtaining unit 132 assigns 1 to a counter i.

[0215] (Step S1202) The menu acquisition unit 132 determines whether or not the ith food information of the selection candidate exists in the menu management unit 113. If the ith food information exists, the process proceeds to step S1203; if not, the process proceeds to step S1207. The ith food information of the selection candidate is at least food information that has not been selected.

[0216] (Step S1203) The menu obtaining unit 132 obtains one or more pieces of nutrient information included in the i-th food information.

[0217] (Step S1204) The menu acquisition unit 132 uses the acquired balance information (two or more pieces of nutrient information) and the nutrient information acquired in step S1203 to determine whether the nutrient information acquired in step S1203 meets the adoption conditions. If the adoption conditions are met, the process proceeds to step S1205; if not, the process proceeds to step S1206. Note that the adoption conditions may be, for example, that the one or more pieces of nutrient information acquired in step S1203 do not exceed the nutrient information (e.g., the amount of energy, the amount of each nutrient) contained in the acquired balance information.

[0218] (Step S1205) The menu acquisition unit 132 acquires the i-th food item information and temporarily stores it in a buffer (not shown).

[0219] (Step S1206) The menu obtaining unit 132 increments the counter i by 1. The process returns to step S1202.

[0220] (Step S1207) The menu acquisition unit 132 determines whether two or more pieces of food information exist in the buffer (not shown). If two or more pieces of food information exist, the process proceeds to step S1208; if not, the process proceeds to step S1211.

[0221] (Step S1208) The menu obtaining unit 132 uses two or more pieces of food information stored in a buffer (not shown) to create a selection screen for the user to select a food item.

[0222] (Step S1209) The menu acquisition unit 132 outputs the selection screen constructed in step S1208.

[0223] (Step S1210) The menu acquisition unit 132 determines whether or not the selection of food information from the user has been accepted on the selection screen presented to the user. If the selection of food information has been accepted, the process proceeds to step S1211; if not, the process returns to step S1210. The menu acquisition unit 132 temporarily stores the selected food information in a buffer (not shown).

[0224] (Step S1211) The menu acquisition unit 132 acquires information about one food item selected by the user and temporarily stores the information in a buffer (not shown).

[0225] (Step S1212) The menu acquisition unit 132 determines whether or not to acquire the next food item information. If the next food item information is to be acquired, the process proceeds to step S1213. If the next food item information is not to be acquired, the process proceeds to step S1216.

[0226] Whether or not to acquire the next food item information is determined in advance, for example. For example, it is determined in advance that food item information on two types of food items, a staple food and a side dish food, is to be acquired.

[0227] (Step S1213) The menu acquisition unit 132 acquires nutrient information contained in the food information temporarily accumulated in step S1210.

[0228] (Step S1214) The menu obtaining unit 132 obtains nutrient information of the difference between the balance information obtained in step S303 and the nutrient information obtained in step S121.

[0229] (Step S1215) The menu acquisition unit 132 performs a menu acquisition process using the nutrient information of the difference acquired in step S1214.

[0230] (Step S1216) Menu acquisition unit 132 uses one or more pieces of food information temporarily stored in a buffer (not shown) to create menu information to be output, and then returns to the upper-level process.

[0231] Next, an example of the difference acquisition process in step S311 will be described with reference to the flowchart of FIG.

[0232] (Step S1301) The difference acquisition unit 133 acquires balance information corresponding to the received diet information. The balance information here is the amount of energy and the amounts of each of two or more nutrients.

[0233] (Step S1302) The difference acquiring unit 133 assigns 1 to a counter i.

[0234] (Step S1303) The difference acquisition unit 133 determines whether the i-th item constituting the difference information exists. If the i-th item exists, the process proceeds to step S1304; if not, the process proceeds to step S1308. The item may be, for example, energy or two or more nutrients.

[0235] (Step S1304) The difference acquisition unit 133 acquires the ith item information included in the acquired object information. The ith item information is, for example, the amount of energy and the amount of each of two or more nutrients.

[0236] (Step S1305) The difference acquisition unit 133 acquires the ith item information corresponding to the meal information, which is the information acquired in step S1301.

[0237] (Step S1306) The difference obtaining unit 133 obtains the difference between the two pieces of item information, associates it with the i-th item, and temporarily stores it in a buffer (not shown).

[0238] (Step S1307) The difference acquisition unit 133 increments the counter i by 1. The process returns to step S1303.

[0239] (Step S1308) The difference acquiring unit 133 generates difference information having information on the difference corresponding to each of two or more items stored in a buffer (not shown).

[0240] A specific example of the operation of the information system A in this embodiment will be described below.

[0241] Currently, a learning model is stored in the learning information storage unit 111. The learning model is information acquired by the learning unit 130 through a machine learning learning process using two or more pieces of training data. Note that the training data here is data in which the intestinal bacteria analysis results (amount of intestinal bacteria PCA0, amount of intestinal bacteria PCA1, ..., amount of PCA40) and balance information (amount of energy, amount of carbohydrates, amount of lipids, amount of dietary fiber, amount of sugar, amount of protein) are used as explanatory variables, and blood glucose level is used as a response variable.

[0242] It is also assumed that two or more pieces of candidate information are stored in the candidate management unit 112. Each of the two or more pieces of candidate information here has a structure of (amount of energy, amount of carbohydrates, amount of lipids, amount of dietary fiber, amount of sugar, amount of protein).

[0243] In this situation, assume that user X inputs his / her intestinal bacteria analysis results (amount of PCA0, PCA1, ..., PCA40) and meal information into terminal device 2 on June 23rd. The meal information includes the character strings of each dish that make up the meal: "grilled lean beef, potatoes, cod roe macaroni, boiled broccoli, grilled king oyster mushrooms, tofu, miso soup (onion and seaweed), white rice with small plums, and sesame salt," as well as an image of the meal. Terminal device 2 then accepts the analysis results and the meal information, associates them with user identifier "X," and transmits them to output device 1.

[0244] Next, the receiving unit 12 of the output device 1 receives the analysis results and diet information associated with the user identifier “X” from the terminal device 2 .

[0245] Next, the prediction unit 131 of the output device 1 obtains a predicted blood glucose level for each of two or more pieces of candidate information through machine learning prediction processing using the candidate information, the analysis results, and the learning model. Next, the prediction unit 131 obtains a blood glucose level that meets the selection condition "the blood glucose level is the lowest value" and obtains candidate information corresponding to the blood glucose level. Assume that the candidate information (amount of energy, amount of carbohydrates, amount of lipids, amount of dietary fiber, amount of sugar, amount of protein) is (612, C1, L1, D1, S1, P1). Furthermore, it can be said that the selected candidate information is ideal balance information for user X.

[0246] It is also assumed that the difference acquisition unit 133 uses the received dietary information to acquire balance information (560, C2, L2, D2, S2, P2) corresponding to the dietary information.

[0247] Next, the difference acquisition unit 133 acquires difference information (-52, -30.5, +3.4, -2.2, -26.1, +3.4), which is a set of differences between each item information contained in the balance information corresponding to the dietary information and each item information contained in the ideal balance information.

[0248] Next, the processing unit 13 composes the target information to be output. Here, the target information is assumed to be the amount of energy "612 kcal" contained in the ideal balance information and a radar chart obtained from the information on the five nutrients contained in the ideal balance information.

[0249] The processing unit 13 also generates difference information to be output, which is (-52, -30.5, +3.4, -2.2, -26.1, +3.4).

[0250] Furthermore, the processing unit 13 uses the acquired various information to compose screen information to be transmitted to the terminal device 2. Next, the output unit 14 transmits the screen information to the terminal device 2 of the user X.

[0251] Next, user X's terminal device 2 receives the screen information, creates a screen from the screen information, and outputs it. An example of such output is shown in FIG. 14. FIG. 14 includes dietary information 1401, balance information 1402 acquired from dietary information 1401, difference information 1403, and ideal balance information 1404. The ideal balance information 1404 includes a radar chart that also includes a graph showing the nutrient balance of the current meal. Needless to say, the output format of the information on the five nutrients included in the ideal balance information is not limited to a radar chart, and may include a graph showing the amount of each nutrient or a table showing the amount of each nutrient.

[0252] As described above, according to this embodiment, it is possible to present target information for identifying a good nutrient balance using the analysis results of intestinal bacteria.

[0253] Furthermore, according to this embodiment, it is possible to present target information for identifying a better nutrient balance using the results of the intestinal bacteria analysis and one or more types of information selected from health information and lifestyle information.

[0254] Furthermore, according to this embodiment, it is possible to present target information for identifying a good nutrient balance from two or more pieces of candidate information using the analysis results of intestinal bacteria.

[0255] Furthermore, according to this embodiment, it is possible to identify a favorable nutrient balance using the results of the analysis of intestinal bacteria, and to suggest a meal menu based on the nutrient balance.

[0256] Furthermore, according to this embodiment, it is possible to present target information for identifying a good nutrient balance by machine learning using the results of the analysis of intestinal bacteria.

[0257] The processing in this embodiment may be implemented by software. This software may be distributed by software download, etc. Furthermore, this software may be recorded on a recording medium such as a CD-ROM and distributed. This also applies to other embodiments in this specification. The software implementing the output device 1 in this embodiment is the following program. That is, this program causes a computer that can access a learning information storage unit that stores learning information acquired using two or more pieces of training data containing intestinal bacteria analysis results and target information, which is information for identifying a good nutrient balance, to function as: a receiving unit that receives target information including the intestinal bacteria analysis results of a single user; a prediction unit that uses the target information and the learning information to acquire target information that identifies a nutrient balance to be recommended to the single user; and an output unit that outputs the target information acquired by the prediction unit.

[0258] Second Embodiment In the present embodiment, an output device capable of presenting the type of a user using the analysis results of the user's intestinal bacteria will be described.

[0259] In this embodiment, an output device that can present meals that tend to raise blood sugar levels or meals that tend not to raise blood sugar levels for a user using the analysis results of the user's intestinal bacteria will be described.

[0260] The conceptual diagram of the information system B in this embodiment is the same as that in FIG. 1 except for the reference numerals of the output devices.

[0261] 15 is a block diagram of an information system B according to this embodiment. The information system B includes an output device 3 and one or more terminal devices 2.

[0262] The output device 3 includes a storage unit 31, a reception unit 32, a processing unit 33, and an output unit 34. The storage unit 31 includes a learning information storage unit 311, a type management unit 312, a candidate management unit 112, and a menu management unit 313. The processing unit 33 includes a learning unit 330, a prediction unit 331, a type determination unit 332, a type information acquisition unit 333, a second prediction unit 334, a recommendation determination unit 335, and a difference acquisition unit 336.

[0263] Various types of information are stored in the storage unit 31 constituting the output device 3. The various types of information include, for example, two or more pieces of teacher data, learning information, second learning information (described later), type information (described later), menu management information, two or more pieces of type detailed information, or two or more pieces of advice information.

[0264] The type detail information is information that explains the type, which will be described later, in detail. The type detail information is information that corresponds to the type identifier, which will be described later. The advice information is information that gives advice to the user according to the user's type. The advice information is, for example, information about foods that are recommended to the user. The advice information is information that corresponds to the type identifier, which will be described later. The content of the advice does not matter. The advice is, for example, advice about diet, but it may also be advice about exercise or sleep. The type detail information and advice information may be stored in the type management unit 312.

[0265] The learning information storage unit 311 stores learning information. The learning information storage unit 311 stores second learning information. The learning information is the same as the learning information described in the first embodiment.

[0266] The second learning information is information acquired using two or more pieces of training data including the intestinal bacteria analysis results, food information, and blood glucose levels. The second learning information is preferably information acquired by the learning unit 330 described below. The second learning information is, for example, a learning model, a correspondence table, or an arithmetic formula.

[0267] Two or more pieces of type management information are stored in the type management section 312. For example, 18 pieces of type management information are stored in the type management section 312.

[0268] The type management information is information about the user's type. The type management information is information for acquiring a type identifier corresponding to the objective information for specifying nutrient balance. The type management information is, for example, a type correspondence table, a type condition table, or a type learning model.

[0269] The type is usually a type related to the user's diet. The type is preferably a type of diet that is preferred by the user. The type is preferably a type of diet that is less likely to raise blood sugar levels for the user. The type may also be referred to as a class, division, classification, etc.

[0270] The type correspondence table has two or more types of correspondence information each having purpose information for specifying a nutrient balance and a type identifier.

[0271] The type condition table has nutrient balance conditions for each of two or more type identifiers. The nutrient balance conditions are conditions related to the target information that specifies the nutrient balance. For example, the nutrient balance conditions include information indicating the range of amounts of each of two or more nutrients, or information indicating the range of proportions of each of two or more nutrients.

[0272] The type learning model is a learning model obtained by performing a machine learning learning process using two or more pieces of training data having objective information for specifying nutrient balance and a type identifier. The learning model, together with information for specifying nutrient balance, is provided to a module that performs machine learning prediction processing. The module is then executed to obtain a type identifier.

[0273] The type identifier is information that identifies the type of the user. The type identifier is, for example, a type ID or type name. The type identifier is, for example, "A", "B", "C", etc.

[0274] The menu management unit 313 stores two or more items of menu management information. The menu management information here is information that identifies nutrients contained in food. The menu management information here includes food information. The menu management information here includes a food identifier and food information. The food identifier is information that identifies a food. The food identifier is, for example, a food ID or a food name. The food information includes information on one or more nutrients contained in a food.

[0275] The reception unit 32 receives various information and instructions. The reception unit 12 receives various information and instructions, for example, from the terminal device 2. The reception unit 12 receives various information and instructions, for example, from a user. The various information and instructions include, for example, reference meal information, learning instructions, second learning instructions, typing output instructions, and food suggestion instructions.

[0276] The learning instruction is an instruction to acquire learning information. The second learning instruction is an instruction to acquire second learning information. The typing output instruction is an instruction to output typing information. The food suggestion instruction is an instruction to output suggestion information.

[0277] The processing unit 33 performs various types of processing. The various types of processing are, for example, processing performed by the learning unit 330, the prediction unit 331, the type determination unit 332, the type information acquisition unit 333, the second prediction unit 334, or the recommendation determination unit 335. Note that the various types of processing are performed on digital signals.

[0278] The learning unit 330 acquires learning information using two or more pieces of teacher data. This process of the learning unit 330 is the same as the process of the learning unit 130.

[0279] The learning unit 330 acquires second learning information using two or more pieces of training data. The training data here includes the intestinal bacteria analysis results, food information, and blood glucose levels. The intestinal bacteria analysis results and food information are explanatory variables, and the blood glucose levels are objective variables. It is preferable that the training data here include one or two types of information selected from health information and lifestyle information in addition to the intestinal bacteria analysis results and objective information. The health information and lifestyle information are explanatory variables.

[0280] The above-mentioned training data is usually stored in the storage unit 31. Alternatively, the training data may be stored in an external device (not shown) and may be information temporarily stored in the storage unit 31. (1) When the second learning information is a learning model

[0281] The learning unit 330 acquires two or more pieces of teacher data from the storage unit 31. The learning unit 130 provides the two or more pieces of teacher data to a machine learning learning module, executes the module, and acquires a second learning model, which is second learning information. The second learning model is provided to a machine learning prediction module together with the received target information, and is used to acquire blood glucose levels. (2) When the second learning information is a correspondence table

[0282] The learning unit 330 vectorizes the explanatory variables of the teacher data for each of the two or more teacher data in the storage unit 31. Next, the learning unit 130 acquires, for example, correspondence information for each of the two or more teacher data, including the vector and the blood glucose level, which is the target information of the teacher data. Next, the learning unit 130 constructs, for example, a second correspondence table including the acquired two or more pieces of correspondence information. Note that the vector is, for example, (amount of enterobacteria PCA0, amount of enterobacteria PCA1, ..., amount of PCA40, height, weight, ..., blood pressure, ..., red blood cell count, ...). (3) When the second learning information is an arithmetic formula

[0283] For example, the learning unit 330 vectorizes the explanatory variables of the teacher data for each of the two or more teacher data in the storage unit 31. Next, the learning unit 330 acquires a second arithmetic formula for each of the two or more teacher data, which has two or more elements of the vector obtained from the explanatory variables as input and outputs the blood glucose level. Note that the learning unit 330 acquires the second arithmetic formula for each of the two or more teacher data, which has two or more elements of the vector obtained from the explanatory variables as input and outputs the blood glucose level, using the least squares method, for example. The method for acquiring the second arithmetic formula is not important.

[0284] The prediction unit 331 acquires the objective information corresponding to the type determined by the type determination unit 332, which will be described later.

[0285] The prediction unit 331 acquires, for example, the purpose information paired with the type determined by the type determination unit 332 from a type correspondence table having two or more pieces of type correspondence information, each of which is a pair of type and purpose information.

[0286] The prediction unit 331 acquires goal information using, for example, the target information, type, and learning information of a single user that has been received. The process by which the prediction unit 331 acquires goal information using the learning information is the same as the process by the prediction unit 131, except that the type determined by the type determination unit 332 is also used. The prediction unit 331 acquires goal information that specifies the nutrient balance to be recommended to the single user, for example, using the target information, type, and learning information of the single user. There are various possible methods for the prediction unit 331 to acquire goal information using the target information, type, and learning information. The method of acquiring such goal information is the same as the process by the prediction unit 131 described in embodiment 1, except that the type is also used, and therefore a description thereof will be omitted. The processing of the prediction unit 331 is similar to the processing of the prediction unit 131 in the following cases: (1) when candidate information is not used; (1-1) when the learning information is a learning model; (1-2) when the learning information is a correspondence table; (1-3) when the learning information is an arithmetic formula; (2) when candidate information is used; (2-1) when the learning information is a learning model; (2-2) when the learning information is a correspondence table; (2-3) when the learning information is an arithmetic formula.

[0287] Furthermore, the prediction unit 331 may perform all of the processing that the prediction unit 131 performs without using the type determined by the type determination unit 332 .

[0288] The type determination unit 332 acquires a type identifier corresponding to the objective information that specifies the nutrient balance. The type determination unit 332 acquires a type identifier corresponding to the objective information that specifies the nutrient balance using the type management information of the type management unit 312. (1) When the type management information is a type correspondence table

[0289] The type determination unit 332 obtains, from the type correspondence table, a type identifier that is paired with the objective information that has the highest similarity to the objective information that specifies the nutrient balance obtained by the prediction unit 131. (2) When the type management information is a type condition table

[0290] The type determination unit 332 acquires, from the type condition table, a type identifier that is paired with the nutrient balance condition satisfied by the objective information that specifies the nutrient balance acquired by the prediction unit 131. (3) When the type management information is a type learning model

[0291] The type determination unit 332 provides the objective information for identifying the nutrient balance acquired by the prediction unit 131 and the type learning model to a prediction module that performs machine learning prediction processing, executes the prediction module, and acquires a type identifier.

[0292] The type information acquisition unit 333 acquires type information of one user including the type identifier acquired by the type determination unit 332. For example, the type information acquisition unit 333 acquires type detail information paired with the type identifier acquired by the type determination unit 332 from the storage unit 31, and acquires type information having the type detail information and the type identifier. For example, the type information acquisition unit 333 acquires advice information paired with the type identifier acquired by the type determination unit 332 from the storage unit 31, and acquires type information having the advice information and the type identifier. Note that the type information may have a type identifier, one or more type detail information pieces, and one or more pieces of advice information.

[0293] The second prediction unit 334 uses the intestinal bacteria analysis results of the single user, the food information contained in each of the two or more pieces of menu management information, and the second learning information to obtain, for each of the two or more pieces of menu management information, the blood glucose level of the single user when eating a food identified by the food information contained in the menu management information. Note that, for each of the two or more pieces of menu management information, "for each of the two or more pieces of menu management information" refers to each piece of food information contained in each of the two or more pieces of menu management information.

[0294] The second prediction unit 334 obtains the blood glucose level of the user, for example, by the following (1), (2), or (3). The second prediction unit 334 associates the blood glucose level obtained for each piece of menu management information with the food information contained in the menu management information. (1) When the second learning information is a learning model

[0295] The second prediction unit 334 provides the intestinal bacteria analysis result of one user accepted by the accepting unit 32, the food information contained in the menu management information, and the second learning model to a machine learning prediction module for each of two or more pieces of menu management information in the menu management unit 113, executes the prediction module, and obtains the blood glucose level. (2) When the second learning information is a correspondence table

[0296] The second prediction unit 334 vectorizes the intestinal bacteria analysis result of one user received by the reception unit 32 and the food information contained in the menu management information for each of two or more pieces of menu management information in the menu management unit 113. For each piece of menu management information, the second prediction unit 334 calculates the similarity between the vector and the vector contained in each piece of correspondence information in the second correspondence table. For each piece of menu management information, the second prediction unit 334 obtains from the second correspondence table the blood glucose level paired with the vector with the greatest similarity. (3) When the second learning information is an arithmetic formula

[0297] The second prediction unit 334 substitutes the intestinal bacteria analysis results of a user accepted by the acceptance unit 32 and the food information contained in the menu management information into a second arithmetic formula for each of two or more menu management information in the menu management unit 113, executes the second arithmetic formula, and obtains the blood glucose level.

[0298] The recommendation determination unit 335 obtains recommendation information using the blood glucose levels for each of the two or more pieces of food information obtained by the second prediction unit 334.

[0299] The recommendation information is information related to food recommendations. For example, the recommendation information is information that suggests meals that are less likely to raise blood sugar levels for the user. For example, the recommendation information is information that notifies the user of meals that are likely to raise blood sugar levels for the user.

[0300] The recommendation determination unit 335 uses the blood glucose levels obtained by the second prediction unit 334 for each of two or more menu management information to obtain recommendation information including food names contained in the menu management information that are paired with one or more blood glucose levels that satisfy the selection conditions.

[0301] The selection condition is, for example, that the blood glucose level is equal to or higher than a threshold value, or that the blood glucose level is equal to or lower than the top N (N is a natural number equal to or greater than 1) highest blood glucose level. In such cases, the recommendation determination unit 335 acquires recommendation information including the names of foods that are not recommended to the user.

[0302] The selection condition is, for example, that the blood glucose level is equal to or less than a threshold value, or is within the top N lowest blood glucose levels (N is a natural number equal to or greater than 1). In such cases, the recommendation determination unit 335 acquires recommendation information including the names of foods recommended to the user.

[0303] The difference acquisition unit 336 acquires balance information that specifies the nutrient balance corresponding to the received reference diet information. Next, the difference acquisition unit 336 acquires difference information regarding the difference between the balance information and the balance information acquired by the prediction unit 331.

[0304] The output unit 34 outputs various types of information. Here, output usually means transmission to the terminal device 2, but may also be a concept including display on a display, projection using a projector, printing on a printer, sound output, transmission to an external device, storage in a recording medium, and delivery of processing results to another processing device or another program.

[0305] The output unit 34, for example, outputs the type identifier acquired by the type determination unit 332. The output unit 34, for example, outputs the type information acquired by the type information acquisition unit 333. The output unit 34, for example, outputs the recommendation information acquired by the recommendation determination unit 335. The output unit 34, for example, outputs information specifying the nutrient balance acquired by the prediction unit 331. The output unit 34, for example, outputs the difference information acquired by the difference acquisition unit 336.

[0306] The storage unit 31, the learning information storage unit 311, the type management unit 312, and the menu management unit 313 are preferably non-volatile recording media, but may also be realized as volatile recording media.

[0307] There is no restriction on the process by which information is stored in the storage unit 31 etc. For example, information may be stored in the storage unit 31 etc. via a recording medium, information transmitted via a communication line etc. may be stored in the storage unit 31 etc., or information input via an input device may be stored in the storage unit 31 etc.

[0308] The processing unit 33, learning unit 330, prediction unit 331, type determination unit 332, type information acquisition unit 333, second prediction unit 334, recommendation determination unit 335, and difference acquisition unit 336 can typically be realized by a processor, memory, or the like. The processing procedures of the processing unit 33, etc. are typically realized by software, and the software is recorded on a recording medium such as a ROM. However, they may also be realized by hardware (dedicated circuit). The processor may be a CPU, MPU, GPU, or the like, and the type does not matter.

[0309] The output unit 34 is usually realized by a wireless or wired communication means, but may also be realized by driver software for an output device such as a display or speaker, or by a combination of driver software for an output device and the output device.

[0310] Next, an example of the operation of the output device 3 will be described with reference to the flowchart of FIG.

[0311] (Step S1601) The receiving unit 32 determines whether or not a second learning instruction has been received. If a learning instruction has been received, the process proceeds to step S1602, and if not, the process proceeds to step S1603.

[0312] (Step S1602) The learning unit 330 performs a second learning process. Return to step S1601. An example of the second learning process will be described with reference to the flowchart in FIG.

[0313] (Step S1603) The reception unit 32 determines whether or not a typing instruction has been received. If a typing instruction has been received, the process proceeds to step S1604, and if not, the process proceeds to step S1609.

[0314] (Step S1604) The prediction unit 131 acquires target information included in the typing instruction. The target information includes an analysis result of the user. In addition to the analysis result, the target information may include one or more types of information of the user's health information and lifestyle information.

[0315] (Step S1605) The prediction unit 131 performs a prediction process using the target information. Examples of the prediction process have been described using the flowcharts in Figures 7, 8, 9, 10, and 11. Note that the target information acquired in this prediction process is information indicating nutrient balance.

[0316] (Step S1606) The type determination unit 332 determines the type of the user using the information indicating the nutrient balance. An example of such type determination processing will be described with reference to the flowchart in FIG.

[0317] (Step S1607) The type information acquisition unit 333 acquires the type identifier acquired in step S1606. The type information acquisition unit 333 acquires the type detail information and advice information paired with the type identifier from the storage unit 31. The type information acquisition unit 333 composes type information having the type identifier, the type detail information, and the advice information.

[0318] (Step S1608) The output unit 34 outputs the type information. The process returns to step S1601.

[0319] (Step S1609) The reception unit 32 determines whether or not a food suggestion instruction has been received. If a food suggestion instruction has been received, the process proceeds to step S16104; if not, the process returns to step S1601.

[0320] (Step S1610) The second prediction unit 334 acquires target information included in the food suggestion instruction.

[0321] (Step S1611) The second prediction unit 334 performs second prediction processing using the target information. A first example of the second prediction processing will be described with reference to the flowchart in FIG.

[0322] (Step S1612) The recommendation determination unit 335 performs a recommendation determination process. An example of the recommendation determination process will be described with reference to the flowchart in FIG.

[0323] (Step S1613) The output unit 34 outputs the recommendation information. The process returns to step S1601.

[0324] 16, the output device 3 may receive a learning instruction, and the learning unit 330 may acquire learning information. This process is the same as the process of the learning unit 130.

[0325] In the flowchart of FIG. 16, the output unit 34 may acquire both the type information and the recommendation information and output both the type information and the recommendation information.

[0326] Furthermore, in the flowchart of FIG. 16, the process ends when the power is turned off or an interrupt occurs to end the process.

[0327] Next, a first example of the second learning process in step S1602 will be described with reference to the flowchart in Fig. 17. The first example of the second learning process is a case where a learning model is acquired.

[0328] (Step S1701) The learning unit 330 assigns 1 to a counter i.

[0329] (Step S1702) The learning unit 330 determines whether or not the i-th teacher data exists in the storage unit 31. If the i-th teacher data exists, the process proceeds to step S1703; if not, the process proceeds to step S1705.

[0330] (Step S1703) The learning unit 330 acquires the i-th teacher data from the storage unit 31. Next, the learning unit 330 converts the i-th teacher data into, for example, a vector having a structure to be provided to a learning module, and adds it to a buffer (not shown).

[0331] (Step S1704) The learning unit 330 increments the counter i by 1. The process returns to step S1702.

[0332] (Step S1705) The learning unit 330 provides a set of two or more vectors (for example, a file) in a buffer (not shown) to the learning module in the storage unit 11.

[0333] (Step S1706) The learning unit 330 executes the learning module.

[0334] (Step S1707) The learning unit 330 acquires a second learning model that is a result of the execution of the learning module in step S1706, and stores the second learning model in the learning information storage unit 311. The process returns to the upper level process.

[0335] 17, the learning unit 330 may use two or more pieces of teacher data to construct the second learning information, which is a correspondence table. Also, the learning unit 330 may use two or more pieces of teacher data to construct the second learning information, which is an arithmetic formula.

[0336] Next, a first example of the type determination process in step S1606 will be described with reference to the flowchart in FIG.

[0337] (Step S1801) The type determination unit 332 acquires the balance information acquired in S1605.

[0338] (Step S1802) The type determining unit 332 assigns 1 to a counter i.

[0339] (Step S1803) The type determination unit 332 determines whether the i-th type correspondence information exists in the type correspondence table. If the i-th type correspondence information exists, the process proceeds to step S1804; if not, the process proceeds to step S1807.

[0340] (Step S1804) The type determination unit 332 acquires balance information contained in the i-th type correspondence information.

[0341] (Step S1805) The type determination unit 332 calculates the similarity between the balance information acquired in step S1801 and the balance information acquired in step S184, and associates the calculated similarity with the i-th type correspondence information. Note that the balance information is usually a vector.

[0342] (Step S1806) The type determination unit 332 increments the counter i by 1. The process returns to step S1803.

[0343] (Step S1807) The type determination unit 332 acquires the type identifier contained in the type correspondence information associated with the maximum similarity, and returns to the upper level processing.

[0344] Next, a first example of the second prediction process in step S1611 will be described with reference to the flowchart in FIG.

[0345] (Step S1901) The second prediction unit 334 acquires the second learning model from the learning information storage unit 311.

[0346] (Step S1902) The second prediction unit 334 assigns 1 to a counter i.

[0347] (Step S1903) The second prediction unit 334 determines whether or not the i-th menu management information exists in the menu management unit 313. If the i-th menu management information exists, the process proceeds to step S1904; if not, the process returns to the upper level process.

[0348] (Step S1904) The second prediction unit 334 acquires food information contained in the i-th menu management information.

[0349] (Step S1905) The second prediction unit 334 acquires the intestinal bacteria analysis results of the user included in the food suggestion instruction. The second prediction unit 334 acquires food information and a second learning model. The second prediction unit 334 provides the intestinal bacteria analysis results, food information, and the second learning model to a prediction module that performs machine learning prediction processing, executes the prediction module, and acquires the blood glucose level. The second prediction unit 334 associates the blood glucose level with the food information included in the i-th menu management information. Note that the user's intestinal bacteria analysis results may be the user's intestinal bacteria analysis results alone, or may include the user's intestinal bacteria analysis results as well as one or more types of information selected from health information and lifestyle information.

[0350] (Step S1906) The second prediction unit 334 increments the counter i by 1. The process returns to step S1903.

[0351] Next, an example of the recommendation determination process in step S1612 will be described with reference to the flowchart in FIG.

[0352] (Step S2001) The recommendation determination unit 335 sorts food information using blood glucose levels as a key.

[0353] (Step S2002) The recommendation determination unit 335 acquires one or more pieces of food information that satisfy a first selection condition. The first selection condition may be, for example, that the food has a blood glucose level that is within the top N (N is a natural number equal to or greater than 1) or that the blood glucose level is equal to or higher than a threshold.

[0354] (Step S2003) The recommendation determination unit 335 acquires one or more pieces of food information that satisfy a second selection condition. The second selection condition may be, for example, that the blood glucose level is within the top N (N is a natural number equal to or greater than 1) or that the blood glucose level is equal to or less than a threshold.

[0355] (Step S2004) The recommendation determination unit 335 generates recommendation information including one or more pieces of food information acquired in step S2002 and one or more pieces of food information acquired in step S2003, and then returns to the upper processing.

[0356] A specific example of the operation of the information system B in this embodiment will be described below.

[0357] The type management unit 312 of the output device 3 stores the type management table shown in FIG. 21 . The type management table here is a table for managing type management information. The type management table here is a type correspondence table. The type management table manages records having "ID," "energy (kcal)," "protein (g)," "fat (g)," "carbohydrates (g)," "sugars (g)," "dietary fiber (g)," "fat / carbohydrates," and "meal pattern." "ID" is information for identifying a record. "Energy (kcal)," "protein (g)," "fat (g)," "carbohydrates (g)," "sugars (g)," "dietary fiber (g)," and "fat / carbohydrates" are balance information corresponding to a type. "Meal pattern" is a type identifier. Here, the type identifier is a number from "1" to "18." Note that type identifiers "1" to "18" can be considered to be types "A" to "R," respectively.

[0358] The type management unit 312 also stores a type relationship information management table shown in Fig. 22. The type relationship information management table is a table for managing records having "type identifier," "type detailed information," and "advice information."

[0359] The menu management unit 313 stores the menu management table shown in FIG. 23. The menu management table manages records having "ID," "food name," "energy (kcal)," "protein (g)," "fat (g)," "carbohydrates (g)," "sugars (g)," "dietary fiber (g)," and "fat / carbohydrates." "ID" is information that identifies the record. "Food name" is a food identifier. "Energy (kcal)" is the calorie content of the food, while "protein (g)," "fat (g)," "carbohydrates (g)," "sugars (g)," "dietary fiber (g)," and "fat / carbohydrates" are the amounts of nutrients contained in the food, respectively.

[0360] It is also assumed that the learning information storage unit 311 stores the above-mentioned learning model and the second learning model.

[0361] Furthermore, it is assumed that two or more pieces of candidate information are stored in the candidate management unit 112. Here, each of the two or more pieces of candidate information has a structure of (amount of energy, amount of carbohydrates, amount of lipids, amount of dietary fiber, amount of sugar, amount of protein).

[0362] In the above situation, two specific examples of the operation of the output device 3 will be described using the processing concept diagram of Fig. 24. Specific Example 1 is an example in which the output device 3 outputs type information. Specific Example 2 is an example in which the output device 3 outputs recommendation information.

[0363] (Specific Example 1) It is assumed that the receiving unit 32 of the output device 3 receives a typing instruction including target information of user Y from the terminal device 2. It is assumed that the target information here includes, for example, the analysis results of user Y, health information of user Y, and lifestyle information of user Y (2401). It is also possible that the typing instruction includes target information, for example, because the target information can be acquired based on a user identifier included in the typing instruction.

[0364] Next, the prediction unit 131 acquires the target information included in the typing instruction. Next, for each of two or more pieces of candidate information, the prediction unit 131 provides the candidate information (2402), the target information (2401), and the learning model to a prediction module that performs machine learning prediction processing, executes the prediction module, and acquires a predicted blood glucose level. Note that the blood glucose level is the postprandial blood glucose level when the food corresponding to the candidate information is eaten. Furthermore, the blood glucose level is associated with the candidate information.

[0365] Next, the prediction unit 131 acquires a blood glucose level that meets the selection condition "blood glucose level is the lowest value" and acquires candidate information corresponding to the blood glucose level. The selected candidate information (2403) can be said to be ideal balance information for user Y. The ideal balance information is information on nutritional balance that is unlikely to raise the user's blood glucose level.

[0366] Next, the type determination unit 332 acquires a vector of ideal balance information (energy (kcal), protein (g), lipid (g), carbohydrate (g), sugar (g), dietary fiber (g), lipid / carbohydrate). Next, the type determination unit 332 calculates the similarity between the vector (energy (kcal), protein (g), lipid (g), carbohydrate (g), sugar (g), dietary fiber (g), lipid / carbohydrate) composed of each record in FIG. 21 and the vector of user Y's ideal balance information. Next, the type determination unit 332 acquires, from FIG. 21, a type identifier paired with the vector composed of each record with the greatest similarity. Here, it is assumed that the type determination unit 332 acquires a type identifier "7." Note that here, the type identifiers "1," "2," ..., "7," ..., "18" are respectively rephrased as type identifiers "A," "B," ..., "G," ..., "R."

[0367] The type information acquisition unit 333 acquires the type detail information and advice information paired with the type identifier "G" from Fig. 22. Next, the type information acquisition unit 333 creates type information (2404) having the type identifier, the type detail information, and the advice information.

[0368] Next, the output unit 34 transmits the type information to the terminal device 2. Note that the output unit 34 may also transmit information about the ideal balance for the user Y, which information has been acquired by the output device 3.

[0369] Next, the terminal device 2 receives and outputs the type information. Examples of such output are shown in Figs. 25, 26, and 27.

[0370] FIG. 25 has a type identifier (Type "G") (2501) and type detail information (2502). FIG. 26 shows advice information paired with the type identifier "G". FIG. 27 graphically shows the nutritional composition corresponding to each of the 18 type identifiers. In FIG. 27, the type identifier "G" of user Y is made to stand out compared to the other type identifiers (2701). The information constituting FIG. 27 is stored in the storage unit 31.

[0371] The terminal device 2 may also receive and output the ideal balance information for the user Y acquired by the output device 3. An example of such an output is shown in Fig. 28. In Fig. 28, the gray area indicates the range recommended in general nutritional science.

[0372] (Specific Example 2) It is assumed that the reception unit 32 of the output device 3 receives a food suggestion instruction including target information of the user Y from the terminal device 2.

[0373] Next, the second prediction unit 334 acquires target information included in the food suggestion instruction. Then, the second prediction unit 334 performs the second prediction process using the target information as follows. Note that the target information here includes, for example, the analysis result of user Y, health information of user Y, and lifestyle information of user Y (2401).

[0374] The second prediction unit 334 acquires the second learning model from the learning information storage unit 311. Next, for each piece of menu management information in the menu management unit 313, the second prediction unit 334 provides the food information (2402) and target information (2401) contained in the menu management information and the second learning model to a prediction module that performs machine learning prediction processing, executes the prediction module, acquires the blood glucose level, and associates the blood glucose level with the menu management information.

[0375] Next, the recommendation determination unit 335 sorts the food information contained in the menu management information using blood glucose levels as a key. Next, the recommendation determination unit 335 acquires 15 pieces of food information that satisfy a first selection condition (here, for example, "top 15 food items from highest blood glucose levels"). The recommendation determination unit 335 also acquires 10 pieces of food information that satisfy a second selection condition (here, for example, "top 10 food items from lowest blood glucose levels"). Next, the recommendation determination unit 335 uses the 15 pieces of food information that satisfy the first selection condition to construct first recommendation information (2404) that includes food information on foods that user Y should avoid. The recommendation determination unit 335 also uses the 15 pieces of food information that satisfy the second selection condition to construct second recommendation information (2404) that includes food information on foods to be recommended to user Y. Next, the output unit 34 transmits the first recommendation information and the second recommendation information to the terminal device 2.

[0376] The terminal device 2 receives and outputs the first recommendation information and the second recommendation information. An example of the first recommendation information is 2901 in Fig. 29, and an example of the second recommendation information is 2902 in Fig. 29.

[0377] As described above, according to this embodiment, the type of a user can be presented using the analysis results of the user's intestinal bacteria.

[0378] Furthermore, according to this embodiment, the analysis results of the user's intestinal bacteria can be used to present meals that tend to raise blood sugar levels or meals that tend not to raise blood sugar levels for the user.

[0379] The software for realizing the output device 3 in this embodiment is the following program: That is, this program causes a computer that can access a learning information storage unit that stores learning information acquired using two or more pieces of teacher data that include intestinal bacteria analysis results and objective information, which is information for identifying a favorable nutrient balance, to function as a type management unit that stores two or more pieces of type management information that include type identifiers that identify a user type corresponding to the objective information for identifying the nutrient balance, a reception unit that receives target information including the intestinal bacteria analysis results of a single user, a prediction unit that uses the objective information and the learning information to acquire objective information that identifies a nutrient balance recommended to the single user, a type determination unit that acquires a type identifier corresponding to the objective information, and an output unit that outputs the type identifier.

[0380] The software that realizes the output device 3 is the following program. In other words, this program causes a computer that can access a learning information storage unit that stores second learning information acquired using two or more pieces of training data that include intestinal bacteria analysis results, food information, and blood glucose levels, and a menu management unit that stores two or more pieces of menu management information that include food identifiers that identify foods and food information that includes one or more nutrients contained in the foods, to function as: a second prediction unit that acquires the blood glucose level of the user when the user eats a food identified by each of the two or more pieces of food information, using the intestinal bacteria analysis results of the user, the food information included in each of the two or more pieces of menu management information, and the second learning information; a recommendation determination unit that acquires recommendation information regarding food recommendations, using the blood glucose levels for each of the two or more pieces of food information acquired by the second prediction unit; and an output unit that outputs the recommendation information acquired by the recommendation determination unit.

[0381] 30 shows the appearance of a computer that executes the programs described in this specification to realize the output device 1, output device 3, etc. of the various embodiments described above. The above-described embodiments can be realized by computer hardware and a computer program executed thereon. FIG. 30 is an overview of this computer system 300, and FIG. 31 is a block diagram of the system 300.

[0382] In FIG. 30, a computer system 300 includes a computer 301 including a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.

[0383] 31, in addition to a CD-ROM drive 3012, the computer 301 includes an MPU 3013, a bus 3014 connected to the CD-ROM drive 3012 etc., a ROM 3015 for storing programs such as a boot-up program, a RAM 3016 connected to the MPU 3013 for temporarily storing instructions for application programs and providing temporary storage space, and a hard disk 3017 for storing application programs, system programs, and data. Although not shown here, the computer 301 may further include a network card for providing connection to a LAN.

[0384] A program that causes the computer system 300 to execute the functions of the output device 1 and the like of the above-described embodiment may be stored on a CD-ROM 3101, inserted into the CD-ROM drive 3012, and then transferred to the hard disk 3017. Alternatively, the program may be transmitted to the computer 301 via a network (not shown) and stored on the hard disk 3017. The program is loaded into the RAM 3016 when executed. The program may also be loaded directly from the CD-ROM 3101 or the network.

[0385] The program does not necessarily include an operating system (OS) or a third-party program that causes the computer 301 to execute the functions of the output device 1 of the above-described embodiment. The program only needs to include instructions that call appropriate functions (modules) in a controlled manner to achieve the desired results. How the computer system 300 operates is well known, and a detailed description thereof will be omitted.

[0386] In addition, in the above program, the steps of transmitting information and receiving information do not include processing performed by hardware, such as processing performed by a modem or interface card in the transmission step (processing that can only be performed by hardware).

[0387] The computer that executes the program may be a single computer or a plurality of computers, that is, it may perform centralized processing or distributed processing.

[0388] Furthermore, in each of the above embodiments, it goes without saying that two or more communication means present in one device may be physically realized by one medium.

[0389] Furthermore, in each of the above embodiments, each process may be realized by centralized processing in a single device, or may be realized by distributed processing in a plurality of devices.

[0390] The present invention is not limited to the above-described embodiment, and various modifications are possible, and it goes without saying that these modifications are also included within the scope of the present invention.

[0391] As described above, the output device according to the present invention has the effect of being able to present target information for identifying a good nutrient balance using the analysis results of intestinal bacteria, and is useful as an output device, etc.

Claims

1. An output device comprising: a learning information storage unit in which learning information obtained using two or more teacher data having intestinal bacteria analysis results and object information, which is information for identifying a good nutrient balance, is stored; a reception unit that receives target information including intestinal bacteria analysis results of a single user; a prediction unit that uses the object information and the learning information to obtain object information for identifying a nutrient balance to be recommended to the single user; and an output unit that outputs the object information obtained by the prediction unit.

2. The output device of claim 1, wherein the teacher data includes one or more types of information selected from the group consisting of health information of a user corresponding to an intestinal bacteria analysis result and lifestyle information related to the user's life, and the receiving unit receives target information selected from the group consisting of the intestinal bacteria analysis result of the one user and one or more types of information selected from the group consisting of the health information of the one user and the lifestyle information of the one user.

3. An output device as described in claim 1 or claim 2, further comprising a candidate management unit in which two or more candidate pieces of candidate information indicating candidates for information that identifies a good nutrient balance are stored, wherein the prediction unit obtains a blood glucose level for each of the two or more pieces of candidate information using the candidate information, the target information and the learning information, and obtains candidate information corresponding to the blood glucose level that satisfies a selection condition, and the output unit outputs the candidate information obtained by the prediction unit.

4. An output device as described in claim 1 or claim 2, wherein the target information is a balance identifier that identifies the nutrient balance or balance information that indicates the nutrient balance, and the output unit outputs balance information identified by the balance identifier that is the target information, or the balance information that is the target information.

5. An output device as described in claim 1 or claim 2, further comprising a menu acquisition unit that acquires balance information indicating nutrient balance based on the objective information and acquires menu information specifying a meal menu corresponding to the balance information, wherein the output unit outputs the menu information instead of or in addition to the objective information.

6. An output device as described in claim 1 or claim 2, wherein the receiving unit receives reference dietary information relating to a reference diet, acquires balance information specifying a nutrient balance corresponding to the reference dietary information, and further includes a difference acquisition unit that acquires difference information relating to a difference between the balance information and the balance information that is the target information acquired by the prediction unit, and the output unit outputs the difference information.

7. An output device as described in claim 1 or claim 2, wherein the learning information is a learning model obtained by a machine learning learning process using the two or more teacher data, and the prediction unit performs a machine learning prediction process using the target information and the learning model to obtain the target information.

8. An output device as described in claim 1 or claim 2, further comprising: a type management unit in which two or more type management information having a type identifier that identifies a type of user corresponding to objective information specifying nutrient balance is stored; a type determination unit that refers to the type management unit and obtains a type identifier corresponding to the objective information obtained by the prediction unit; and an output unit that outputs the type identifier.

9. An output device according to claim 8, further comprising a type information acquisition unit that acquires type information of the one user including the type identifier acquired by the type determination unit, wherein the output unit outputs the type information.

10. The output device of claim 8, further comprising a difference acquisition unit that acquires objective information corresponding to the type determined by the type determination unit, acquires balance information specifying the nutrient balance corresponding to the objective information, and acquires difference information regarding the difference between the balance information and the balance information that is the objective information acquired by the prediction unit, and the output unit outputs the difference information.

11. The learning information storage unit stores second learning information acquired using two or more teacher data having intestinal bacteria analysis results, food information, and blood glucose levels; a menu management unit stores two or more menu management information having a food identifier for identifying a food and food information having information on one or more nutrients contained in the food; a second prediction unit acquires the blood glucose level of the one user when the one user eats a food identified by each of the two or more food information using the intestinal bacteria analysis result of the one user, the food information contained in each of the two or more menu management information, and the second learning information; and a recommendation determination unit acquires recommendation information regarding food recommendations using the blood glucose levels for each of the two or more food information acquired by the second prediction unit; and the output unit outputs the recommendation information acquired by the recommendation determination unit.

12. An output method realized by a learning information storage unit in which learning information obtained using two or more teacher data having intestinal bacteria analysis results and target information, which is information for identifying a good nutrient balance, is stored, a reception unit, a prediction unit, and an output unit, comprising: a reception step in which the reception unit receives target information including intestinal bacteria analysis results of a single user; a prediction step in which the prediction unit uses the target information and the learning information to obtain target information for identifying a nutrient balance to be recommended to the single user; and an output step in which the output unit outputs the target information obtained in the prediction step.

13. A program for causing a computer that can access a learning information storage unit in which learning information obtained using two or more pieces of teacher data having intestinal bacteria analysis results and target information, which is information for identifying a good nutrient balance, to function as: a reception unit that receives target information including intestinal bacteria analysis results of a single user; a prediction unit that uses the target information and the learning information to obtain target information for identifying a nutrient balance to be recommended to the single user; and an output unit that outputs the target information obtained by the prediction unit.

Images