Information processing device, information processing method, and program

Abstract

To solve the problem that conventionally, when a measure for the result of inspection is to be proposed, the reason for the proposal could not have been presented.SOLUTION: The above problem can be solved by an information processing device 1 comprising: a user information reception unit 121 that receives user information including result information that specifies a user's inspection result; a measure determination unit 133 that acquires learning information based on two or more training data that are associated with a measure identifier for identifying a measure taken by the user, and that include first result information that specifies the inspection result before the measure is taken and second result information that specifies the inspection result of the result after the measure is taken, and acquires a measure identifier that corresponds to the result information that the user information possesses, using the learning information and the received user information; a reason information acquisition unit 134 that, using the received user information and one or more training data items that corresponds to the acquired measure identifier, acquires reason information regarding the reason why the measure is recommended; and an information output unit 141 that outputs the measure identifier and the reason information.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to an information processing apparatus and the like that propose countermeasures according to the test results of a user's living body.

Background Art

[0002] In the prior art, there has been a technique for outputting information indicating health foods recommended according to test results, information indicating lifestyle habits recommended according to test results, and the like (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the prior art, when proposing countermeasures against the test results, the basis for the proposal could not be presented.

Means for Solving the Problems

[0005] The first information processing device of the present invention comprises: a user information receiving unit that receives user information having one or more user attribute values ​​including result information that identifies the results of an examination concerning a user's biological body; a learning information storage unit that stores learning information based on two or more training data sets, each containing two or more user attribute values, which are associated with a treatment identifier that identifies the treatment taken by the user, and which include a first result information that identifies the examination results before the treatment is performed and a second result information that identifies the examination results after the treatment is performed; a treatment decision unit that uses the learning information and the user information received by the user information receiving unit to obtain a treatment identifier that identifies the treatment according to the result information held by the user information; a basis information acquisition unit that uses one or more training data sets corresponding to the treatment identifier acquired by the treatment decision unit and the user information received by the user information receiving unit to obtain basis information regarding the basis for recommending the treatment identified by the treatment identifier acquired by the treatment decision unit; and an information output unit that outputs the treatment identifier acquired by the treatment decision unit and the basis information acquired by the basis information acquisition unit.

[0006] This configuration allows for the presentation of the rationale behind any proposed actions taken in response to the test results.

[0007] Furthermore, the information processing device of the second invention, in contrast to the first invention, is an information processing device in which the basis information acquisition unit uses one or more training data corresponding to the action identifier acquired by the action decision unit and user information received by the user information reception unit to acquire basis information that includes one or more types of information, such as a basis level that identifies the degree of strength of the basis for recommending an action, and reason information that indicates the reason for recommending an action, and which includes effectiveness information regarding the effectiveness if the action is taken, or satisfaction information regarding the satisfaction after the action is taken.

[0008] This configuration allows for the presentation of appropriate justification for any proposed actions to be taken in response to the test results.

[0009] Furthermore, the information processing device of this third invention is an information processing device in which, with respect to the first or second invention, the basis information acquisition unit acquires basis information indicating that there is no basis if it is unable to acquire basis information.

[0010] This configuration allows for the clear indication that, when proposing actions to take in response to test results, there is no basis for those proposals.

[0011] Furthermore, the information processing device of the fourth invention further comprises a reward information acquisition unit that acquires reward information which is information that identifies a reward for recommending that a user take action regarding any one of the first to third inventions, and is information that corresponds to the basis information, and an information output unit that also outputs the reward information acquired by the reward information acquisition unit.

[0012] This configuration demonstrates that when suggested actions to be taken in response to the test results, it is possible to provide the user with an incentive to take those actions.

[0013] Furthermore, the information processing device of the fifth invention, in contrast to the fourth invention, is an information processing device in which the reward information acquisition unit acquires reward information that identifies the reward according to the basis level included in the basis information.

[0014] This configuration demonstrates that when suggested actions to be taken in response to the test results, it is possible to provide the user with an incentive to take those actions.

[0015] Furthermore, the information processing device of the sixth invention further comprises a classification unit that, for each of the first to fifth inventions, classifies two or more training data into two or more classes using effect information, which is information obtained using second result information possessed by two or more training data for each response identifier and is information relating to the effect of the response, and associates each of the two or more training data with a class identifier that identifies the class, and the response determination unit uses user information received by the user information receiving unit and learning information based on two or more training data to determine the class to which the user information belongs for each of the two or more response identifiers, and distinguishes and obtains response identifiers corresponding to classes with a large difference between the first result information and the second result information and response identifiers corresponding to classes with a small difference.

[0016] This configuration allows for the determination of appropriate actions to be taken in response to the test results.

[0017] Furthermore, the information processing device of the seventh invention, compared to the sixth invention, includes user satisfaction as a result of the action taken as training data, and the classification unit classifies two or more training data into two or more classes for each action identifier using the effect information and satisfaction level for two or more training data, and associates each of the two or more training data with a class identifier that identifies the class.

[0018] This configuration allows for more appropriate decisions to be made regarding the test results.

[0019] Furthermore, the information processing device of the eighth invention, compared to the sixth invention, has a basis level associated with each of the two or more classes classified by the classification unit, based on effect information for the training data corresponding to the class, a basis information acquisition unit acquires the basis level corresponding to the class, an effect information which is information acquired using the second result information possessed by the training data corresponding to the class determined by the action decision unit, and obtains effectiveness information using effect information which is information relating to the effect of the action, and obtains basis information which includes the basis level and reason information which is information including effectiveness information, as described in claim 6.

[0020] With such a configuration, when proposing a countermeasure against the result of an inspection, it is possible to present an appropriate basis for the proposal.

[0021] Further, in the information processing apparatus of the ninth invention, with respect to the eighth invention, the teacher data also has the satisfaction degree of the user as a result of taking the countermeasure, and the classification unit classifies two or more pieces of teacher data into two or more classes for each countermeasure identifier by using the effect information and the satisfaction degree for two or more pieces of teacher data, and associates them with class identifiers for identifying classes for two or more pieces of teacher data, and the basis information acquisition unit acquires satisfaction degree information by using the satisfaction degree of the teacher data corresponding to the class to which the user information belongs, and acquires basis information including reason information including the satisfaction degree information.

[0022] With such a configuration, when proposing a countermeasure against the result of an inspection, it is possible to present a more appropriate basis for the proposal.

[0023] Further, the information processing apparatus of the tenth invention further includes a learning information acquisition unit that acquires learning information by using two or more pieces of teacher data having one or more user attribute values satisfying a similarity condition among the one or more user attribute values included in the user information received by the user information reception unit, with respect to any one of the first to ninth inventions, and the learning information storage unit of the learning information storage unit stores the learning information acquired by the learning information acquisition unit.

[0024] With such a configuration, it is possible to determine a more appropriate countermeasure against the result of an inspection.

[0025] Further, the information processing apparatus of the eleventh invention, with respect to any one of the first to tenth inventions, the countermeasure is a challenge to ingest a product for a certain period or more for improving the inspection result, and the teacher data includes the degree of engagement in the challenge, which is the response information of a questionnaire regarding the challenge.

[0026] With such a configuration, when proposing a countermeasure against the result of an inspection, it is possible to present a basis for the proposal.

Advantages of the Invention

[0027] According to the information processing apparatus of the present invention, when proposing a countermeasure against the result of an inspection, the basis of the proposal can be presented.

Brief Description of the Drawings

[0028] [Figure 1] Conceptual diagram of information system A in Embodiment 1 [Figure 2] Block diagram of the same information system A [Figure 3] Flowchart for explaining an operation example of the same information processing apparatus 1 [Figure 4] Flowchart for explaining the first example of the same learning information creation process [Figure 5] Flowchart for explaining an example of the same classification process [Figure 6] Flowchart for explaining an example of the same learning process [Figure 7] Flowchart for explaining the second example of the same learning information creation process [Figure 8] Flowchart for explaining the third example of the same learning information creation process [Figure 9] Flowchart for explaining the fourth example of the same learning information creation process [Figure 10] Flowchart for explaining an example of the same output information acquisition process [Figure 11] Flowchart for explaining an example of the same countermeasure information acquisition process [Figure 12] Flowchart for explaining an example of the same countermeasure information acquisition process [Figure 13] Flowchart for explaining an example of the same countermeasure information acquisition process [Figure 14] Flowchart for explaining an example of the same countermeasure information acquisition process [Figure 15] Flowchart for explaining an example of the same basis information processing [Figure 16] Flowchart for explaining an example of the same reward information processing [Figure 17] Diagram showing the same teacher data management table [Figure 18] A diagram showing the same evidence level determination table. [Figure 19] This diagram illustrates the concept of classifying the same training data. [Figure 20] Figure showing an example of the same output. [Figure 21] Overview of the computer system [Figure 22] Block diagram of the computer system [Modes for carrying out the invention]

[0029] The embodiments of the information processing device, etc., will be described below with reference to the drawings. In the embodiments, components that are denoted by the same reference numerals perform the same operation, and therefore, further explanation may be omitted.

[0030] (Embodiment 1) This embodiment describes an information processing device that receives user test results, uses learned information to acquire and output appropriate actions (e.g., products) corresponding to the test results, and supporting information regarding the basis for recommending such actions.

[0031] Furthermore, in this embodiment, we will describe an information processing device that acquires and outputs reward information using the basis information.

[0032] Furthermore, in this embodiment, we will describe an information processing device that, after the user has taken action, uses the response information from a questionnaire regarding the action to acquire and output information regarding the appropriate action based on the inspection results and the basis for recommending that action.

[0033] In this embodiment, the association of information X 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 limited. Information X and information Y may be linked, may exist in the same buffer, may information X be contained in information Y, or information Y may be contained in information X, and so on.

[0034] Figure 1 is a conceptual diagram of information system A in this embodiment. Information system A comprises an information processing device 1 and one or more terminal devices 2.

[0035] Information Processing Device 1 is a device that outputs appropriate actions and supporting information based on the user's test results. Information Processing Device 1 is a so-called server, such as a cloud server or an ASP server, but the type is not specified. Information Processing Device 1 may also be a standalone device.

[0036] Terminal device 2 is a device used by the user. Terminal device 2 can be, for example, a personal computer, a smartphone or other multifunctional mobile phone, a mobile phone, or a tablet device, but the type is not limited. The user is a person who uses information processing device 1 or the administrator of information processing device 1.

[0037] The information processing device 1 and one or more terminal devices 2 can communicate with each other via a network such as the Internet or a LAN.

[0038] Figure 2 is a block diagram of information system A in this embodiment. The information processing device 1 comprises a storage unit 11, a reception unit 12, a processing unit 13, and an output unit 14. The storage unit 11 comprises a teacher data storage unit 111 and a learning information storage unit 112. The reception unit 12 comprises a user information reception unit 121. The processing unit 13 comprises a classification unit 131, a learning information acquisition unit 132, a response decision unit 133, a rationale information acquisition unit 134, and a reward information acquisition unit 135. The output unit 14 comprises an information output unit 141.

[0039] The terminal device 2 includes a terminal storage unit 21, a terminal receiving unit 22, a terminal processing unit 23, a terminal transmission unit 24, a terminal receiving unit 25, and a terminal output unit 26.

[0040] The storage unit 11, which constitutes the information processing device 1, stores various types of information. These various types of information include, for example, training data, learning information, survey information, response information, and various conditions, which will be described later. These various conditions include, for example, effectiveness conditions, satisfaction conditions, acquisition conditions, reward conditions, and information that associates class identifiers with justification levels, which will be described later.

[0041] The training data storage unit 111 stores one or more training data. Training data is information that forms the basis of learning information. Training data is associated with an action identifier that identifies the actions taken by the user (which may also be called the "subject"). Training data has two or more user attribute values. The two or more user attribute values ​​include first result information and second result information. The two or more user attribute values ​​include, for example, the user's gender, the user's age, the user's height, the user's weight, action information, lifestyle information, goals, goal achievement rate (which may also be called the To-Be achievement rate), and satisfaction level. Training data may be obtained directly from the user themselves, or it may be obtained indirectly from academic papers, journals, etc.

[0042] Training data preferably includes effect information. Effect information is information about the effectiveness of an action. Effect information is usually obtained using secondary result information. Effect information is, for example, information about the difference between the first result information and the second result information. Effect information is, for example, information indicating the degree of improvement, such as the difference between the first result information and the second result information, or the ratio of the difference between the first result information and the second result information to the first result information. Effect information is, for example, information about the difference between the second result information and the target. Effect information is, for example, information indicating the degree of improvement, such as the difference between the second result information and the target, or the ratio of the difference between the second result information and the target to the target, or the degree of achievement of the target. If training data includes effect information, training data does not need to include the first result information and the second result information. Effect information is information about the effect of an action taken by a user. Effect information refers to information about the difference between the first and second outcome information, such as the number of blood pressure decreases (increases), the number of blood glucose decreases (increases), information indicating whether or not there was an effect, the improvement rate of indoxyl sulfate measurements, the number of increases in gut health, the rate of increase in gut health, the amount of weight loss (increase), the amount of improvement in body composition, and the increase or decrease in height.

[0043] Actions taken by a user refer to information about the user's actions in response to the test results. Examples of actions taken by a user include products consumed by the user, services provided to the user, actions taken by the user, and challenges undertaken by the user. Challenges taken by a user include, for example, consuming a product for a specified period (e.g., eating meals using low-sodium product A for two weeks), enjoying a service provided by the user for a specified period, or taking a specified action for a specified period (e.g., walking for 30 minutes in one month, running for one hour every day for two weeks, quitting or limiting smoking or drinking for one month, etc.). The products consumed, services, and actions taken by the user are, for example, products intended to improve the test results.

[0044] The first result information identifies the test results before treatment, which are identified by the treatment identifier. The second result information identifies the test results after treatment, which are identified by the treatment identifier. Examples of the first and second result information include blood glucose levels, blood pressure, indoxyl sulfate measurement values, gut health, weight, body composition, and height. Examples of body composition include muscle mass (percentage), body fat mass (percentage), visceral fat mass (percentage), subcutaneous fat mass (percentage), bone density, BMI, body age, etc.

[0045] The examination is a biological examination of the user. The examination is performed using, for example, a specimen. The specimen is either a biological sample or an in vivo sample. Biological samples include, for example, urine, feces, blood, oral cells, saliva, hair, body hair, sebum, nails, skin fragments, semen, tears, sweat, breast milk, nasal mucus, sputum, tartar, and tongue coating. In vivo samples include, for example, photographic data of the subject, video data of the subject, audio data of the subject, and house dust from the subject's residence. The examination is not limited to, for example, a test using a test kit. For example, it may be a test using physical examination equipment such as a weighing scale, body composition analyzer, or height meter. The examination may also be, for example, a subjective stress level check or discomfort check using a questionnaire, or a cognitive function test. Examples of such tests include the Mini-Mental State Examination (MMSE) test, which assesses cognitive status (see URL: http: / / www.shizuokamind.org / wp-content / uploads / 2013 / 10 / MMSE.pdf), and the Kupperman Menopausal Symptom Index (KKSI) test, which assesses menopausal symptoms (see URL: https: / / ohana-clinic-kinoshitacho.com / wp-content / themes / ohana / download / kuppaman.pdf).

[0046] A test kit is, for example, an item for testing using substances exuded from a subject's body. Substances exuded from the body include, for example, urine, blood, feces, and other bodily fluids. A test kit is, for example, an item for testing to obtain indoxyl sulfate measurements. Examples of test kits include an equol test kit, a urine test to measure whether equol is produced from soy isoflavones; an intestinal environment test kit, a urine test to measure the health of the intestinal environment based on the amount of putrefactive substances derived from intestinal bacteria; an oxidative stress test kit, a urine test to measure DNA (8-OHdG) damaged by reactive oxygen species; a salt reduction test kit, a urine test to measure how much salt is consumed per day; and a urine test kit to measure the presence or absence of antibodies to Helicobacter pylori, which increases the risk of stomach cancer.

[0047] The test kit may be information for testing that uses information about the subject (e.g., rights information). Information about the subject may be, for example, the subject's voice data or image data, and the image data may be still images or videos. Non-physical materials may include both image and voice data. Information about the subject's health status can be obtained as a test result from the subject's voice data or image data. Such information about health status may be, for example, the degree of frailty or the degree of depression, but the content of the information is not restricted.

[0048] Activity information refers to information about the user's progress toward the challenge. Activity information may include, for example, the degree to which the user is committed to the challenge. Activity information may also include, for example, the completion rate and the number of days spent taking action. The completion rate may be, for example, the percentage of days spent taking action within a predetermined period. Activity information may include not only information about direct efforts toward the challenge but also information about incidental efforts. Information about incidental efforts may include, for example, information indicating that snacking was stopped, information indicating that sleep time was increased, information specifying sleep time, information indicating that exercise time was increased, information specifying exercise time, information indicating that smoking was stopped, information specifying the number of cigarettes smoked, information indicating that alcohol consumption was stopped, information specifying the amount of alcohol consumed, etc.

[0049] Lifestyle information refers to information about a user's lifestyle. Examples of lifestyle information include whether or not they smoke, whether or not they drink alcohol, the amount of cigarettes smoked per day, the frequency and amount of alcohol consumed, whether or not they exercise, and the amount of exercise they do over a specified period (e.g., one day).

[0050] A goal is a user's goal, and is usually a goal related to test results. A goal is a goal with a basis in fact. For example, a goal can be said to represent an ideal state based on research papers or statistical results. Examples of goals include target blood pressure values ​​and target weight values. Other target values ​​may also be set corresponding to primary and secondary result information, such as target values ​​for blood glucose levels, blood pressure, indoxyl sulfate measurement values, and gut health.

[0051] The To-Be achievement rate is the degree to which a user has progressed from their current state (As-Is) to their ideal state (To-Be). In other words, it is the percentage of the user's personal goal (for example, a systolic blood pressure of "120") that has been achieved.

[0052] Satisfaction level is the degree of satisfaction after an action has been taken. It typically represents the level of satisfaction with the action taken. For example, satisfaction can be expressed as a numerical value from 1 to 5, or as one of three levels: "satisfied," "average," or "not satisfied."

[0053] Furthermore, user attribute values ​​other than the first and second result information included in the training data are, for example, response information corresponding to a survey conducted with the user. This response information includes, for example, the user's gender, age, height, weight, activity information, lifestyle information, goals, To-Be achievement rate, and satisfaction level.

[0054] The learning information storage unit 112 stores learning information based on two or more training data sets. The learning information storage unit 112 may include, for example, a learning device (described later), a correspondence table (described later), and a training data set. The process of creating the learning information in the learning information storage unit 112 may be performed by the learning information acquisition unit 132 (described later), or by an external learning device (not shown). The training data set is a collection of two or more training data sets.

[0055] The learner is, for example, data created through a machine learning learning process using two or more training data sets. This learning process may be performed by the learning information acquisition unit 132 described later, or by an external learning device (not shown). If an external learning device performs the learning process, the training data storage unit 111 is not necessary. The learner may also be called a classifier, predictor, learning model, model, etc.

[0056] The correspondence table contains two or more correspondence information entries. For example, the correspondence information may show the relationship between one or more user attribute values ​​and a corresponding identifier. Alternatively, the correspondence information may show the relationship between one or more user attribute values ​​and a class identifier. Class identifiers will be discussed later.

[0057] The reception unit 12 receives various types of information and instructions. These types of information and instructions include, for example, user information, which will be described later.

[0058] Here, "reception" typically refers to the reception of information transmitted from terminal device 2 via a wired or wireless communication line. However, "reception" may also be a concept that includes the reception of information input from input devices such as keyboards, mice, and touch panels, as well as the reception of information read from recording media such as optical discs, magnetic discs, and semiconductor memory.

[0059] The user information receiving unit 121 receives user information for a single user. User information has one or more user attribute values. User attribute values ​​include result information. Result information is information that identifies the results of a single user's biological tests. Result information may include, for example, blood glucose levels, blood pressure, indoxyl sulfate measurement values, and gut health status. The user information receiving unit 121 typically receives user information to obtain treatment identifiers and rationale information. The result information contained in the user information received by the user information receiving unit 121 is information about test results before treatment is performed.

[0060] The processing unit 13 performs various processes. These processes include, for example, those performed by the classification unit 131, the learning information acquisition unit 132, the response decision unit 133, the rationale information acquisition unit 134, and the reward information acquisition unit 135.

[0061] The processing unit 13 may configure the output information using the basis information acquired by the basis information acquisition unit 134. The processing unit 13 may also configure the output information using the basis information acquired by the basis information acquisition unit 134 and the reward information acquired by the reward information acquisition unit 135. The output information is the information that is output.

[0062] The classification unit 131 classifies two or more training data from the training data storage unit 111 into two or more classes and associates each of the two or more training data with a class identifier. The class identifier is information that identifies a class. For example, the class identifier may be one of "Class 1", "Class 2", "Class 3", or "Class 4".

[0063] The classification unit 131 classifies two or more training data into two or more classes for each response identifier using effect information, and associates each of the two or more training data with a class identifier.

[0064] The classification unit 131, for example, obtains effect information regarding the difference between the first result information and the second result information of each of the two or more training data sets for each response identifier. Next, the classification unit 131 classifies the effect information corresponding to each of the two or more training data sets into two or more classes, and classifies the two or more training data sets into two or more classes according to the classification of the effect information, and associates each of the two or more training data sets with a class identifier. For example, the classification unit 131 classifies the two or more training data into three classes, such as class 1 for training data where "effect information <= threshold 1", class 2 for training data where "threshold 1 < effect information <= threshold 2", and class 3 for training data where "threshold 2 < effect information". The number of classes is not limited.

[0065] The classification unit 131, for example, classifies two or more training data into two or more classes for each treatment identifier, using the effect information and satisfaction level of each of the two or more training data, and associates each of the two or more training data with a class identifier that identifies the class.

[0066] The classification unit 131 classifies training data of two or more into four classes, for example, assigning training data where "effect information <= threshold 1 and satisfaction <= threshold a" to "Class 1," training data where "effect information <= threshold 1 and threshold a < satisfaction" to "Class 2," training data where "threshold 1 < effect information and satisfaction <= threshold a" to "Class 3," and training data where "threshold 1 < effect information and threshold a < satisfaction" to "Class 4." The number of classes is not limited.

[0067] Furthermore, the classification unit 131 may, for example, classify each of the two or more training data into two or more classes using a known cluster analysis algorithm. The known cluster analysis algorithms may be hierarchical methods such as k-means, nearest neighbor method, centroid method, or group average method, or non-hierarchical methods such as k-means or hypervolume method. Classifying two or more training data into two or more classes means associating each of the two or more training data with a class identifier.

[0068] The learning information acquisition unit 132 acquires learning information using two or more training data from the training data storage unit 111. It is preferable for the learning information acquisition unit 132 to acquire learning information using the results of the classification performed by the classification unit 131. The learning information is information used to acquire a corresponding identifier. The learning information may also be information to which the class identifier of the class to which the user belongs has been acquired before acquiring the corresponding identifier. The learning information may also be, for example, a learner or a correspondence table.

[0069] The learning information acquisition unit 132 preferably acquires learning information using two or more training data sets that have one or more user attribute values ​​satisfying similarity conditions to one or more user attribute values ​​contained in the user information received by the user information reception unit 121. The similarity conditions are, for example, that the similarity is equal to or greater than a threshold, or that the similarity is greater than a threshold. The similarity is the similarity between a vector whose elements are one or more user attribute values ​​contained in the user information received by the user information reception unit 121 and a vector whose elements are one or more user attribute values ​​in the training data. The algorithm for calculating the similarity between the two vectors is publicly known, so its explanation is omitted.

[0070] The following describes an example of a specific algorithm for obtaining training information. (1) Learning information for obtaining class identifiers (1-1) When the learning information is a learning device (1-1-1) When the learner is a learner that outputs one class identifier from among the candidates for class identifier (e.g., multi-class classification)

[0071] The learning information acquisition unit 132 acquires two or more training data from the training data storage unit 111 for each corresponding identifier. Next, for each corresponding identifier, the learning information acquisition unit 132 performs machine learning training using one or more user attribute values ​​from each of the two or more training data as explanatory variables and the class identifier corresponding to each of the two or more training data as the objective variable, acquires a learner, associates it with the corresponding identifier, and stores it in the learning information storage unit 112. The class identifier of the training data is information acquired by the classification unit 131.

[0072] The algorithms used for machine learning learning processes include deep learning, decision trees, random forests, SVR, etc., but are not limited to those. Similarly, the algorithms used for machine learning prediction processes, which will be discussed later, also include deep learning, decision trees, random forests, SVR, etc., but are not limited to those. Furthermore, various machine learning functions and existing libraries can be used for machine learning, such as the TensorFlow library, fastText, tinySVM, and the R language's random forest module. Note that modules can also be called programs, software, functions, methods, etc.

[0073] Furthermore, the learning information acquisition unit 132 does not need to use all of the user attribute values ​​that make up the training data when creating the learning device; it may use only some of the user attribute values. (1-1-2) When the learner is a binary classification learner

[0074] The learning information acquisition unit 132 acquires two or more training data from the training data storage unit 111 for each corresponding identifier. Next, for each corresponding identifier and each class identifier, the learning information acquisition unit 132 uses the training data corresponding to the class identifier of interest as a positive example and the training data not corresponding to the class identifier as a negative example, performs machine learning training, acquires a learner, associates the corresponding identifier with the class identifier, and stores it in the learning information storage unit 112.

[0075] The learning information acquisition unit 132 uses one or more user attribute values ​​from each of the two or more training data sets as explanatory variables and the class identifiers corresponding to each of the two or more training data sets as the target variable to perform machine learning training and acquire a binary classification learner. (1-2) When the learning information is a correspondence table (1-2-1) When one correspondence information corresponds to one training data.

[0076] The learning information acquisition unit 132 acquires two or more training data from the training data storage unit 111 for each corresponding identifier. Next, the learning information acquisition unit 132 constructs a vector for each corresponding identifier, with one or more user attribute values ​​from each of the two or more training data as elements. Then, for each corresponding identifier, the learning information acquisition unit 132 constructs a correspondence table having two or more correspondence pieces of information, each containing the vector and a class identifier corresponding to the training data, and stores the correspondence table in the learning information storage unit 112, associating it with the corresponding identifier. (1-2-2) When one correspondence information corresponds to one class identifier

[0077] The learning information acquisition unit 132 acquires two or more training data from the training data storage unit 111 for each corresponding identifier. Next, the learning information acquisition unit 132 constructs a vector for each corresponding identifier and class identifier, with one or more user attribute values ​​from each of the one or more training data. Next, the learning information acquisition unit 132 acquires a representative vector that represents one or more vectors for each corresponding identifier and class identifier. Next, the learning information acquisition unit 132 constructs a correspondence table for each corresponding identifier and class identifier, having two or more correspondence information items that have a representative vector and a class identifier, and stores the correspondence table in the learning information storage unit 112, associating it with the corresponding identifier. (2) Learning information for obtaining the corresponding identifier

[0078] If the learning information is information for obtaining a corresponding identifier, there is no need to classify the training data, and the classification unit 131 is unnecessary. (2-1) When the learning information is a learning device (2-1-1) When using effect information

[0079] The learning information acquisition unit 132 acquires effect information corresponding to each training data from two or more training data held in the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data whose effect information satisfies predetermined effect conditions. Next, the learning information acquisition unit 132 uses one or more user identifiers from each of the acquired training data as explanatory variables and the corresponding counter identifiers for each of the acquired training data as the objective variable, performs machine learning training, acquires a learner, and stores it in the learning information storage unit 112.

[0080] The effectiveness criteria are conditions used to determine whether the effect information is equal to or greater than a predetermined effect. For example, the effectiveness criteria may be that the effect information is above a threshold (e.g., "blood glucose level decrease of 10 or more", "systolic blood pressure decrease of 20 or more", "target achievement rate of 80% or more") or greater than a threshold (e.g., "blood glucose level decrease of 10 or more", "systolic blood pressure decrease of 20 or more", "target achievement rate of 70%") or a specific value (e.g., "effective", "significant improvement"). (2-1-2) When using effectiveness information and satisfaction level

[0081] The learning information acquisition unit 132 acquires effect information and satisfaction levels for each training data from each of the two or more training data held by the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data in which the effect information satisfies predetermined effect conditions and the satisfaction level satisfies predetermined satisfaction conditions. Next, the learning information acquisition unit 132 uses one or more user identifiers from each of the acquired training data as explanatory variables and the corresponding response identifiers for each of the training data as the objective variable, performs machine learning training, acquires a learner, and stores it in the learning information storage unit 112.

[0082] Furthermore, the effectiveness information for training data refers to the effectiveness information obtained from the information contained in the training data, or the effectiveness information contained in the training data. In addition, the satisfaction condition is a condition for determining that satisfaction is high, for example, that satisfaction is above or above a threshold or greater than a threshold (satisfied). (2-2) When the learning information is a correspondence table (2-2-1) When using effect information (2-2-1-1) When one correspondence information corresponds to one training data.

[0083] The learning information acquisition unit 132 acquires effect information corresponding to each training data from two or more training data held in the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data whose effect information satisfies predetermined effect conditions. Next, the learning information acquisition unit 132 constructs a vector whose elements are one or more user identifiers held in each of the acquired training data. Then, the learning information acquisition unit 132 constructs a correspondence table having two or more correspondence information, each having the vector and a corresponding handling identifier for the training data, and stores the correspondence table in the learning information storage unit 112. (2-2-1-2) When one correspondence information corresponds to one correspondence identifier

[0084] The learning information acquisition unit 132 acquires effect information corresponding to each training data from two or more training data held in the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data whose effect information satisfies predetermined effect conditions. Next, the learning information acquisition unit 132 constructs a vector for each response identifier, with one or more user attribute values ​​from each of the one or more training data as elements. Next, the learning information acquisition unit 132 acquires a representative vector for each response identifier, representing one or more vectors. Next, the learning information acquisition unit 132 constructs a correspondence table having two or more correspondence information items having a representative vector and a response identifier, and stores the correspondence table in the learning information storage unit 112. (2-2-2) When using effectiveness information and satisfaction level (2-2-2-1) When one correspondence information corresponds to one training data.

[0085] The learning information acquisition unit 132 acquires effect information and satisfaction levels corresponding to each training data from two or more training data held by the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data where the effect information satisfies predetermined effect conditions and the satisfaction level satisfies predetermined satisfaction conditions. Next, the learning information acquisition unit 132 constructs a vector whose elements are one or more user identifiers held by each of the acquired training data. Then, the learning information acquisition unit 132 constructs a correspondence table having two or more correspondence information, each having the vector and a corresponding handling identifier for the training data, and stores the correspondence table in the learning information storage unit 112. (2-2-2-2) When one correspondence information corresponds to one correspondence identifier

[0086] The learning information acquisition unit 132 acquires effect information and satisfaction levels corresponding to each training data from two or more training data held by the training data storage unit 111. Next, the learning information acquisition unit 132 acquires one or more training data where the effect information satisfies predetermined effect conditions and the satisfaction level satisfies predetermined satisfaction conditions. Next, the learning information acquisition unit 132 constructs a vector for each response identifier, with one or more user attribute values ​​from each of the one or more training data as elements. Next, the learning information acquisition unit 132 acquires a representative vector for each response identifier, which represents one or more vectors. Next, the learning information acquisition unit 132 constructs a correspondence table having two or more correspondence information sets, each containing a representative vector and a response identifier, and stores the correspondence table in the learning information storage unit 112.

[0087] The action determination unit 133 acquires learning information and uses the learning information and user information received by the user information reception unit 121 to acquire an action identifier that identifies one or more actions corresponding to the result information contained in the user information.

[0088] The response determination unit 133, for example, uses the user information received by the user information reception unit 121 and the learning information to determine the class to which the user information belongs for each of the two or more response identifiers, and distinguishes and obtains response identifiers corresponding to classes with a large difference between the first result information and the second result information (classes with a large response effect) and response identifiers corresponding to classes with a small difference (classes with a small response effect). Distinguishing and obtaining them means, for example, sorting the response identifiers in descending order of difference, or obtaining only the response identifiers corresponding to differences that satisfy the effect conditions.

[0089] The response determination unit 133 obtains a response identifier, for example, through machine learning prediction processing. The response determination unit 133 obtains a response identifier, for example, using a correspondence table. An example of the processing of the response determination unit 133 is described below. When learning information is a learning device (1-1) When the learner is a learner that obtains a class identifier (1-1-1) When there is only one learning device

[0090] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as an element. The response determination unit 133 also obtains a learner corresponding to the response identifier from the learning information storage unit 112. Next, for each response identifier, the response determination unit 133 provides the vector and the learner to a machine learning prediction processing module, executes the module, and obtains a class identifier.

[0091] Next, the action decision unit 133 associates the acquired class identifier with the action identifier and stores it in a buffer (not shown). The action decision unit 133 may also acquire one or more action identifiers corresponding to class identifiers that satisfy the acquisition conditions (for example, identifiers for classes with high effectiveness and high satisfaction) and store them in a buffer (not shown). The acquisition conditions are the conditions for acquiring the action identifier. The acquisition conditions are effect conditions based on effect information. The acquisition conditions may also be effect conditions and satisfaction conditions. For example, the effect condition is that the effect information is above or above a threshold or greater than a threshold. The satisfaction condition is a condition based on satisfaction, for example, that the satisfaction level is above or above a threshold or greater than a threshold. For example, the acquisition condition is that the effect indicated by the effect information is the highest and the satisfaction level is the highest. (1-1-2) When the learner is a binary classification learner with 2 or more classes

[0092] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as an element. The response determination unit 133 also obtains learners corresponding to the response identifier and each class identifier from the learning information storage unit 112. Next, the response determination unit 133 provides the constructed vector and the obtained learners for each response identifier and class identifier to a machine learning prediction processing module, executes the module, and obtains a prediction result indicating whether or not it belongs to the class identified by the class identifier. The prediction result may include a score.

[0093] Next, the action decision unit 133 obtains one or more class identifiers corresponding to the prediction result, which includes the information that it "belongs to a class".

[0094] Next, the action determination unit 133 associates the acquired class identifier with each action identifier and stores them in a buffer (not shown). The action determination unit 133 may also acquire one or more action identifiers corresponding to class identifiers that satisfy the acquisition conditions and store them in the buffer (not shown). The acquisition conditions are, for example, that the effect information corresponding to the class identifier is equal to or greater than a threshold (has a significant effect). (1-2) When the learner is a learner that obtains a corresponding identifier (1-2-1) When there is only one learning device

[0095] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as an element. The response determination unit 133 also obtains a learner from the learning information storage unit 112. Next, the response determination unit 133 provides the vector and the learner to a machine learning prediction processing module, executes the module, and obtains the response identifier. (1-2-2) When the learner is a binary classification learner with 2 or more options

[0096] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as an element. The response determination unit 133 also obtains a learner from the learning information storage unit 112 for each response identifier. Next, the response determination unit 133 provides the vector and the learner to a machine learning prediction processing module for each response identifier, executes the module, and obtains a prediction result that includes information indicating whether or not it belongs to the response identified by the response identifier. Finally, the response determination unit 133 obtains a response identifier corresponding to the prediction result that includes information indicating that it "belongs to the response identified by the response identifier". (2) When the learning information is a correspondence table (2-1) When the correspondence table is a correspondence table from which class identifiers are obtained

[0097] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as its elements. Next, the response determination unit 133 determines the correspondence information having the vector that most closely approximates the said vector, and obtains the class identifier from the said correspondence information.

[0098] Next, the action determination unit 133 associates the acquired class identifier with the action identifier and stores it in a buffer (not shown). The action determination unit 133 may also acquire one or more action identifiers corresponding to class identifiers that satisfy the acquisition conditions and store them in the buffer (not shown). The acquisition conditions are, for example, that the effect information corresponding to the class identifier is equal to or greater than a threshold (has a significant effect). (2-2) When the correspondence table is a correspondence table from which a corresponding identifier is obtained

[0099] The response determination unit 133 obtains one or more attribute values ​​from the user information received by the user information reception unit 121 for each response identifier. Next, the response determination unit 133 constructs a vector with each of these one or more attribute values ​​as its elements. Next, the response determination unit 133 determines the corresponding information that has the vector most similar to the vector and obtains the response identifier from the corresponding information.

[0100] The action determination unit 133 may also obtain the action identifiers for all two or more candidate actions from the storage unit 11.

[0101] The rationale information acquisition unit 134 acquires rationale information for each of the one or more action identifiers acquired by the action decision unit 133, using one or more training data corresponding to the action identifier and user information received by the user information reception unit 121. The rationale information is information relating to the basis for recommending the action identified by the action identifier acquired by the action decision unit 133.

[0102] The evidence information acquisition unit 134 acquires, for example, the evidence level corresponding to the class identifier assigned by the classification unit 131. The evidence information acquisition unit 134 acquires effectiveness information using, for example, the effectiveness information for the training data corresponding to the class identifier assigned by the classification unit 131, and acquires the evidence level and reason information which includes the effectiveness information. It is preferable for the evidence information acquisition unit 134 to acquire evidence information which includes one or more types of information from the evidence level and reason information.

[0103] The basis information acquisition unit 134 preferably acquires satisfaction information using the satisfaction level of the training data corresponding to the class to which the user information belongs, and acquires basis information including reason information that includes the satisfaction level information.

[0104] The term "basis" can also be called "evidence." "Basis information" can also be called "evidence information." Basis information, for example, includes one or more pieces of information from the following categories: evidence level and reason information. Reason information includes one or more pieces of information from the following categories: effectiveness information and satisfaction information.

[0105] The evidence level is information that identifies the degree of strength of the evidence recommending an action. The evidence level can be one of "1," "2," or "3," but it is acceptable as long as the information is ordered, such as "A," "B," or "C."

[0106] Reason information is information about the reasons for recommending an action. Effectiveness information is information about the effectiveness of taking an action. Effectiveness information includes, for example, the average effectiveness score when an action is taken, the percentage of users whose effectiveness score is above a threshold when an action is taken (percentage of training data), and the number of users whose effectiveness score is above a threshold when an action is taken. Satisfaction information is information about satisfaction after taking an action. Satisfaction information includes the average satisfaction score after an action is taken, the percentage of users whose satisfaction score is above a threshold when an action is taken (training data may also be used), and the number of users whose satisfaction score is above a threshold when an action is taken.

[0107] The basis information acquisition unit 134 acquires a basis level for each of the one or more response identifiers acquired by the response decision unit 133, using one or more training data corresponding to the response identifier and user information received by the user information reception unit 121. The basis information acquisition unit 134 acquires a class identifier that is paired with one or more training data corresponding to the response identifier acquired by the response decision unit 133. Next, the basis information acquisition unit 134 acquires a basis level corresponding to the class identifier. In this case, the basis level is associated with the class identifier and stored in the storage unit 11.

[0108] The evidence information acquisition unit 134 acquires effectiveness information for each countermeasure identifier, for example, using one or more training data corresponding to the countermeasure identifier acquired by the countermeasure decision unit 133. The evidence information acquisition unit 134 acquires effect information for each of the one or more training data corresponding to each countermeasure identifier acquired by the countermeasure decision unit 133. Next, the evidence information acquisition unit 134 acquires the number of effective individuals, which is the number of effective individuals that satisfy the effect condition, from the acquired effect information for each countermeasure identifier. Next, the evidence information acquisition unit 134 acquires effectiveness information, which is the percentage of effective individuals, for example, using the total number of training data corresponding to the countermeasure identifier and the number of effective individuals. Note that effectiveness information may also be the number of effective individuals, etc. The effect condition is information that shows a high effect in the effect information, for example, the effect information is above a threshold, the effect information is greater than a threshold, or the effect information is "effective".

[0109] The evidence information acquisition unit 134 acquires satisfaction information for each response identifier, for example, using one or more training data corresponding to the response identifier acquired by the response decision unit 133. The evidence information acquisition unit 134 acquires the satisfaction levels of each of the one or more training data corresponding to each response identifier acquired by the response decision unit 133. Next, the evidence information acquisition unit 134 acquires the number of satisfied people, which is the number of satisfaction levels that meet the satisfaction criteria, from among the acquired satisfaction levels, for each response identifier. Next, the evidence information acquisition unit 134 acquires satisfaction information, which is the percentage of satisfaction, using the total number of training data corresponding to the response identifier and the number of satisfied people. Note that satisfaction information may also be the number of satisfied people, etc.

[0110] The rationale information acquisition unit 134 acquires effectiveness information and satisfaction information for each response identifier using one or more training data corresponding to the response identifier acquired by the response decision unit 133. It is preferable that the rationale information acquisition unit 134 then constructs reason information that includes the acquired effectiveness information and the acquired satisfaction information.

[0111] The evidence acquisition unit 134 is preferably capable of acquiring recommendation information to strongly recommend actions to action identifiers that satisfy the goal achievement conditions. The goal achievement conditions are that the degree of goal achievement is equal to or greater than a threshold. The goal achievement rate is the degree of achievement of the goal, which is a user attribute value. The degree of goal achievement may be, for example, a representative value (e.g., mean, median) of goal achievement rates of 1 or more, the number of users who achieved the goal, or the percentage of users who achieved the goal.

[0112] The basis information acquisition unit 134 is preferable to acquire basis information indicating that there is no basis for a countermeasure identifier for which basis information could not be obtained.

[0113] The evidence acquisition unit 134 typically associates the acquired evidence information with the corresponding counter-identifier and temporarily stores it in a buffer (not shown).

[0114] The reward information acquisition unit 135 acquires reward information, which is information that identifies the reward for encouraging the user to take action and corresponds to the rationale information. Reward information may include, for example, information recommending the purchase of a product or service corresponding to the action. Reward information may also include, for example, information indicating the discount rate for the product or service corresponding to the action.

[0115] The reward information acquisition unit 135 acquires reward information that identifies the reward according to the rationale level included in the rationale information. For example, the reward information acquisition unit 135 acquires reward information that identifies a higher reward the lower the rationale level included in the rationale information. In this case, the reward indicated by the reward information is a reward for the user's efforts to adopt a low rationale level. Also, for example, the reward information acquisition unit 135 acquires reward information that identifies a higher reward the higher the rationale level included in the rationale information. In this case, it is when the user is desired to take an effort with a high rationale level. The relationship between such rationale levels and reward information may be set by the administrator of the information processing device 1.

[0116] The reward information acquisition unit 135 acquires, for example, reward information corresponding to the acquired justification level from the storage unit 11.

[0117] The output unit 14 outputs various types of information. These types of information include, for example, a response identifier, justification information, and reward information.

[0118] Here, output usually refers to transmission to terminal device 2, but it may also be a concept that includes display on a screen, projection using a projector, printing with a printer, sound output, storage on a recording medium, and delivery of processing results to other processing devices or other programs.

[0119] The information output unit 141 outputs the response identifier acquired by the response decision unit 133 and the basis information acquired by the basis information acquisition unit 134. It is preferable that the information output unit 141 also outputs the reward information acquired by the reward information acquisition unit 135.

[0120] Various types of information are stored in the terminal storage unit 21, which constitutes the terminal device 2. These types of information include, for example, a user identifier. The user identifier may also be the ID of the terminal device 2, etc.

[0121] The terminal reception unit 22 receives various types of information and instructions. These types of information and instructions include, for example, user information. The means of inputting these types of information and instructions can be anything, such as a microphone, touch panel, keyboard, mouse, or menu screen.

[0122] The terminal processing unit 23 performs various processes. These processes include, for example, converting instructions and information received by the terminal receiving unit 22 into a data structure for transmission. Other processes include, for example, converting information received by the terminal receiving unit 25 into a data structure for output.

[0123] The terminal transmission unit 24 transmits various types of information and instructions to the information processing device 1. These types of information and instructions include, for example, user information.

[0124] The terminal receiving unit 25 receives various types of information from the information processing device 1. These types of information include, for example, a response identifier, justification information, and reward information.

[0125] The terminal output unit 26 outputs various types of information. These types of information include, for example, a response identifier, justification information, and reward information.

[0126] The storage unit 11, the teacher data storage unit 111, the learning information storage unit 112, and the terminal storage unit 21 are preferably made of non-volatile recording media, but can also be made of volatile recording media.

[0127] The process by which information is stored in the storage unit 11, etc. is not relevant. 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.

[0128] The reception unit 12 and the user information reception unit 121 are preferably implemented by wireless or wired communication means, but may also be implemented by means of receiving broadcasts, device drivers for input means such as touch panels and keyboards, or control software for menu screens.

[0129] The processing unit 13, learning information acquisition unit 132, classification unit 131, response decision unit 133, rationale information acquisition unit 134, reward information acquisition unit 135, and terminal processing unit 23 can typically be implemented using a processor, memory, etc. The processing procedures of the processing unit 13, etc., are usually implemented in software, and this software is recorded on a recording medium such as ROM. However, it may also be implemented in hardware (dedicated circuitry). The processor can be a CPU, MPU, GPU, etc., and the type is not limited.

[0130] The output unit 14, the information output unit 141, and the terminal transmission unit 24 are usually implemented by wireless or wired communication means, but may also be implemented by broadcasting means.

[0131] The terminal reception unit 22 can be implemented using device drivers for input means such as touch panels and keyboards, or control software for menu screens, etc.

[0132] The terminal receiving unit 25 is usually implemented by wireless or wired communication means, but it may also be implemented by means of receiving broadcasts.

[0133] The terminal output unit 26 may or may not be considered to include output devices such as a display or speakers. The terminal output unit 26 can be implemented using driver software for an output device, or driver software for an output device and an output device.

[0134] Next, we will explain an example of the operation of information system A. First, we will explain an example of the operation of information processing device 1 using the flowchart in Figure 3.

[0135] (Step S301) The reception unit 12 determines whether it has received one or more training data. If it has received one or more training data, it proceeds to step S302; if it has not received any training data, it proceeds to step S303. Note that reception here refers to, for example, reception from terminal device 2.

[0136] (Step S302) The processing unit 13 stores one or more training data received in step S301 in the training data storage unit 111. The process returns to step S301.

[0137] (Step S303) The processing unit 13 determines whether or not to create learning information. If it decides to create learning information, it proceeds to step S304; if it decides not to create learning information, it proceeds to step S305. The processing unit 13 decides to create learning information, for example, when the receiving unit 12 receives a learning information creation instruction. The processing unit 13 also decides to create learning information, for example, when a predetermined time has arrived. The processing unit 13 also decides to create learning information, for example, when the number of training data stored in the training data storage unit 111 exceeds a threshold. However, the conditions under which the processing unit 13 decides to create learning information are not specified.

[0138] (Step S304) The learning information acquisition unit 132 performs the learning information creation process. The process returns to step S301. An example of the learning information creation process will be explained using the flowcharts in Figures 4, 7, 8, and 9.

[0139] (Step S305) The user information receiving unit 121 determines whether or not it has received user information. If it has received user information, it proceeds to step S306; if it has not received user information, it returns to step S301. Note that reception here refers to, for example, reception from terminal device 2.

[0140] (Step S306) The processing unit 13 performs output information acquisition processing. An example of the output information acquisition processing will be explained using the flowchart in Figure 10.

[0141] (Step S307) The information output unit 141 outputs the output information acquired in step S306. The process returns to step S301. Note that the output here is, for example, transmitted to terminal device 2.

[0142] In the flowchart in Figure 3, processing is terminated by power-off or processing termination interrupts.

[0143] Next, we will explain the first example of the learning information creation process in step S304 using the flowchart in Figure 4. The first example of the learning information creation process is an example of obtaining a learner to obtain a class identifier.

[0144] (Step S401) The processing unit 13 assigns 1 to counter i.

[0145] (Step S402) The processing unit 13 determines whether the i-th handling identifier exists. If the i-th handling identifier exists, the process proceeds to step S403; otherwise, the process returns to the higher-level processing unit.

[0146] (Step S403) The classification unit 131 classifies two or more training data corresponding to the i-th response identifier. An example of such classification processing will be explained using the flowchart in Figure 5.

[0147] (Step S404) The learning information acquisition unit 132 performs a learning process using the results of the classification process in step S403 and acquires a learner. An example of the learning process will be explained using the flowchart in Figure 6.

[0148] (Step S405) The learning information acquisition unit 132 associates the learning device acquired in step S404 with the i-th correspondence identifier and stores it in the learning information storage unit 112.

[0149] (Step S406) The processing unit 13 increments counter i by 1. The process returns to step S402.

[0150] Next, an example of the classification process in step S403 will be explained using the flowchart in Figure 5.

[0151] (Step S501) The classification unit 131 retrieves all training data corresponding to the target treatment identifier (the i-th treatment identifier in S402) from the training data storage unit 111.

[0152] (Step S502) The classification unit 131 assigns 1 to counter i.

[0153] (Step S503) The classification unit 131 determines whether or not the i-th training data exists among the training data obtained in step S501. If the i-th training data exists, the unit proceeds to step S504; otherwise, it returns to the higher-level processing.

[0154] (Step S504) The classification unit 131 obtains effect information for the i-th training data. The classification unit 131 obtains, for example, the first result information and the second result information of the i-th training data, and obtains effect information regarding the difference between the first result information and the second result information. The classification unit 131 obtains, for example, the effect information of the i-th training data.

[0155] (Step S505) The classification unit 131 obtains the satisfaction level of the i-th training data.

[0156] (Step S506) The classification unit 131 uses the effect information obtained in step S504 and the satisfaction level obtained in step S505 to obtain a class identifier corresponding to the effect information and the satisfaction level.

[0157] (Step S507) The classification unit 131 associates the i-th training data with the class identifier obtained in step S506.

[0158] (Step S508) The classification unit 131 increments counter i by 1. Return to step S503.

[0159] In the flowchart of Figure 5, the classification unit 131 determined the class of the training data using the effect information and satisfaction level. However, the classification unit 131 may also determine the class of the training data using only one of the effect information or satisfaction level.

[0160] Furthermore, in the flowchart of Figure 5, the classification unit 131 may classify two or more training data using the known cluster analysis algorithm described above.

[0161] Next, an example of the learning process in step S404 will be explained using the flowchart in Figure 6.

[0162] (Step S601) The learning information acquisition unit 132 assigns 1 to counter i.

[0163] (Step S602) The learning information acquisition unit 132 determines whether or not a class identifier for the i-th class classified by the classification unit 131 in step S403 exists. If the i-th class identifier exists, the process proceeds to step S603; otherwise, it returns to the higher-level processing.

[0164] (Step S603) The learning information acquisition unit 132 acquires one or more positive examples. A positive example is training data associated with the i-th class identifier.

[0165] (Step S604) The learning information acquisition unit 132 acquires one or more negative examples. Note that negative examples are training data that are not associated with the i-th class identifier. Training data that are not associated with the i-th class identifier are usually training data that are associated with class identifiers other than the i-th class identifier.

[0166] (Step S605) The learning information acquisition unit 132 uses the one or more positive examples obtained in step S603 and the one or more negative examples obtained in step S604 to perform machine learning learning and acquire a learner. This learner is a learner that determines whether or not an object belongs to the i-th class and is a learner that performs binary classification.

[0167] (Step S606) The learning information acquisition unit 132 associates the target identification identifier with the i-th class identifier and stores the learner acquired in step S605 in the learning information storage unit 112.

[0168] (Step S607) The learning information acquisition unit 132 increments counter i to 1. Return to step S602.

[0169] In the flowchart of Figure 6, the learning information acquisition unit 132 may use two or more training data corresponding to the target identification identifier of interest, use one or more user attribute values ​​from each training data as explanatory variables, and the class identifier as the target variable, perform machine learning training, acquire a learner, associate it with the target identification identifier of interest, and store the acquired learner in the learning information storage unit 112. In this case, there is one learner corresponding to a single target identification identifier. Such a learner is a learner for predicting one of the class identifiers.

[0170] Next, a second example of the learning information creation process in step S304 will be explained using the flowchart in Figure 7. The second example of the learning information creation process is an example of obtaining a learner for obtaining a response identifier. The learner obtained in the second example is a learner for determining whether or not to perform a response identified by the corresponding response identifier, and is obtained for each response. Such a learner is a learner that performs binary classification.

[0171] (Step S701) The learning information acquisition unit 132 assigns 1 to counter i.

[0172] (Step S702) The learning information acquisition unit 132 determines whether the i-th correspondence identifier exists. If the i-th correspondence identifier exists, the process proceeds to step S703; otherwise, it returns to the higher-level process.

[0173] (Step S703) The learning information acquisition unit 132 acquires two or more training data corresponding to the i-th response identifier from the training data storage unit 111.

[0174] (Step S704) The learning information acquisition unit 132 acquires one or more positive examples from the two or more training data acquired in step S703. A positive example is training data that meets the positive example criteria.

[0175] The positive example condition is a condition for deciding whether action should be taken. The positive example condition may be one or two of the following: meeting the effectiveness condition and meeting the satisfaction condition. For example, the positive example condition may be based on one or more types of information, such as effectiveness information and satisfaction. For example, the positive example condition may be "effectiveness information is above or above a threshold," "satisfaction is above or above a threshold," or "effectiveness information is above or above a threshold AND satisfaction is above or above a threshold."

[0176] (Step S705) The learning information acquisition unit 132 acquires one or more negative examples from the two or more training data acquired in step S703. Negative examples are training data that do not meet the positive example conditions.

[0177] (Step S706) The learning information acquisition unit 132 uses the one or more positive examples acquired in step S704 and the one or more negative examples acquired in step S705 to perform machine learning learning and acquire a learner.

[0178] (Step S707) The learning information acquisition unit 132 stores the learning device acquired in step S706 in the learning information storage unit 112, associating it with the target identification identifier.

[0179] (Step S708) The learning information acquisition unit 132 increments counter i by 1. Return to step S702.

[0180] Next, a third example of the learning information creation process in step S304 will be explained using the flowchart in Figure 8. The third example of the learning information creation process is an example of obtaining a single learner for obtaining a corresponding identifier. Such a learner is usually a learner for performing multi-class classification.

[0181] (Step S801) The learning information acquisition unit 132 acquires one or more training data that matches the positive example condition from the training data storage unit 111.

[0182] (Step S802) The learning information acquisition unit 132 provides the one or more training data acquired in step S801 to the machine learning learning module and acquires a learner.

[0183] (Step S803) The learning information acquisition unit 132 stores the learner acquired in step S802 in the learning information storage unit 112. It then returns to the higher-level processing.

[0184] Next, the fourth example of the learning information creation process in step S304 will be explained using the flowchart in Figure 9. The fourth example of the learning information creation process creates a correspondence table for each corresponding identifier. In the flowchart in Figure 9, the explanation for the steps that are the same as those in the flowchart in Figure 4 will be omitted.

[0185] (Step S901) The learning information acquisition unit 132 assigns 1 to counter j.

[0186] (Step S902) The learning information acquisition unit 132 determines whether or not a j-th class identifier exists that corresponds to the i-th handling identifier. If a j-th class identifier exists, the unit proceeds to step S903; otherwise, it proceeds to step S907.

[0187] (Step S903) The learning information acquisition unit 132 acquires one or more training data corresponding to the i-th correspondence identifier and the j-th class identifier.

[0188] (Step S904) The learning information acquisition unit 132 acquires a representative vector of the vector composed of one or more training data points.

[0189] (Step S905) The learning information acquisition unit 132 temporarily stores the representative vector obtained in step S904, associating it with the j-th class identifier.

[0190] (Step S906) The learning information acquisition unit 132 increments counter j by 1. Return to step S902.

[0191] (Step S907) The learning information acquisition unit 132 constructs a correspondence table having two or more correspondence information items, each having a representative vector and a class identifier, obtained in step S905, and stores the correspondence table in the learning information storage unit 112, associating it with the i-th correspondence identifier.

[0192] (Step S908) The learning information acquisition unit 132 increments counter i to 1. Return to step S402.

[0193] Next, an example of the output information acquisition process in step S306 will be explained using the flowchart in Figure 10.

[0194] (Step S1001) The action decision unit 133 acquires action information. An example of the action information acquisition process will be explained using the flowcharts in Figures 11, 12, 13, and 14.

[0195] (Step S1002) The evidence information acquisition unit 134 acquires evidence information. An example of evidence information processing will be explained using the flowchart in Figure 15.

[0196] (Step S1003) The reward information acquisition unit 135 acquires reward information. An example of reward information processing will be explained using the flowchart in Figure 16.

[0197] (Step S1004) The processing unit 13 constructs output information using the response information, rationale information, and reward information. It then returns to the higher-level processing unit.

[0198] Next, an example of the process for acquiring action information in step S1001 will be explained using the flowchart in Figure 11. In the flowchart in Figure 11, learners are used for each action and for each class.

[0199] (Step S1101) The handling decision unit 133 obtains one or more user attribute values ​​from the received user information.

[0200] (Step S1102) The action decision unit 133 assigns 1 to counter i.

[0201] (Step S1103) The response determination unit 133 determines whether the i-th response identifier exists. If the i-th response identifier exists, the process proceeds to step S1104; otherwise, it returns to the higher-level process.

[0202] (Step S1104) The action decision unit 133 assigns 1 to counter j.

[0203] (Step S1105) The action decision unit 133 determines whether the j-th class identifier exists. If the j-th class identifier exists, the unit proceeds to step S1106; otherwise, it proceeds to step S1111.

[0204] (Step S1106) The response determination unit 133 obtains a learner corresponding to the i-th response identifier and the j-th class identifier from the learning information storage unit 112.

[0205] (Step S1107) The response decision unit 133 uses the learner acquired in step S1106 and one or more user attribute values ​​acquired in step S1101 to perform machine learning prediction processing and obtain prediction results. The prediction results include information on whether or not the user belongs to the class identified by the j-th class identifier. Preferably, the prediction results include a score.

[0206] (Step S1108) The decision-making unit 133 proceeds to step S1109 if the prediction result obtained in step S1107 is "belongs to the class identified by the j-th class identifier", and to step S1110 if the prediction result is "does not belong to the class identified by the j-th class identifier".

[0207] (Step S1109) The response determination unit 133 temporarily stores the j-th class identifier in a buffer (not shown) in association with the i-th response identifier.

[0208] (Step S1110) The action decision unit 133 increments counter j by 1. Return to step S1105.

[0209] (Step S1111) The action decision unit 133 increments counter i by 1. Return to step S1103.

[0210] In the flowchart of Figure 11, a class identifier was determined for each action identifier. The action determination unit 133 may then store only the action identifiers corresponding to the determined class identifier in the storage unit 11, only if the determined class identifier satisfies the acquisition conditions.

[0211] Next, an example of the process for acquiring action information in step S1001 will be explained using the flowchart in Figure 12. In the flowchart in Figure 12, the explanation of the same steps as in the flowchart in Figure 11 will be omitted. Note that in the flowchart in Figure 12, a learning unit is used to determine whether or not to take action, and a learning unit is used for each action.

[0212] (Step S1201) The response determination unit 133 obtains the learner corresponding to the i-th response identifier from the learning information storage unit 112.

[0213] (Step S1202) The action decision unit 133 uses the learner acquired in step S1201 and one or more user attribute values ​​acquired in step S1101 to perform machine learning prediction processing and obtain prediction results. The prediction results include information on whether or not the action identified by the j-th action identifier is effective (whether or not it satisfies the condition). Preferably, the prediction results include a score.

[0214] (Step S1203) The action decision unit 133 proceeds to step S1204 if the prediction result in step S1202 includes "satisfy" (e.g., "1"), and to step S1207 if it includes "does not satisfy" (e.g., "0"). Note that a prediction result of "satisfying" indicates that the action identified by the action identifier should be taken.

[0215] (Step S1204) The action decision unit 133 obtains the score of the prediction process in step S1202.

[0216] (Step S1205) The action decision unit 133 determines whether the score obtained in step S1204 satisfies the conditions. If the conditions are met, the process proceeds to step S1206; otherwise, the process proceeds to step S1207. The conditions are that the score is high, for example, "the score is above the threshold" or "the score is greater than the threshold".

[0217] (Step S1206) The action decision unit 133 stores the j-th action identifier in the storage unit 11. It is preferable for the action decision unit 133 to also store the score obtained in step S1204.

[0218] (Step S1207) The action decision unit 133 increments counter i by 1. Return to step S1105.

[0219] Next, an example of the process for acquiring countermeasure information in step S1001 will be explained using the flowchart in Figure 13. In the flowchart in Figure 13, the explanation of the same steps as in the flowchart in Figure 11 will be omitted. Note that in the flowchart in Figure 13, a learner that outputs a countermeasure identifier is used.

[0220] (Step S1301) The action decision unit 133 acquires the learner from the learning information storage unit 112.

[0221] (Step S1302) The response determination unit 133 uses the learner acquired in step S1301 and one or more user attribute values ​​acquired in step S1101 to perform machine learning prediction processing and obtain the prediction result. The prediction result includes a response identifier. Preferably, the prediction result includes a score.

[0222] (Step S1303) The action decision unit 133 stores the action identifier obtained in step S1302. It returns to the higher-level processing. It is preferable that the action decision unit 133 also stores a score at this point.

[0223] Next, an example of the process for obtaining the handling information in step S1001 will be explained using the flowchart in Figure 14. In the flowchart in Figure 14, the explanation of the same steps as in the flowchart in Figure 11 will be omitted. Note that the flowchart in Figure 14 uses a correspondence table to obtain the class identifier.

[0224] (Step S1401) The response determination unit 133 constructs a vector whose elements are one or more user attribute values ​​obtained in step S1101. Next, the response determination unit 133 determines the corresponding information having the vector most similar to the said vector from the correspondence table in the learning information storage unit 112, which is the correspondence table paired with the i-th response identifier.

[0225] (Step S1402) The response determination unit 133 obtains the class identifier associated with the response information determined in step S1401.

[0226] (Step S1403) The action decision unit 133 determines whether the class identifier obtained in step S1402 satisfies the condition. If the condition is met, the process proceeds to step S1404; otherwise, the process proceeds to step S1405. The condition here is that the class identifier is one of one or more predetermined class identifiers (for example, an identifier of a class that satisfies the effect condition).

[0227] (Step S1404) The action determination unit 133 stores the class identifier obtained in step S1402 in pairs with the i-th action identifier.

[0228] (Step S1405) The action decision unit 133 increments counter i by 1. Return to step S1103.

[0229] In addition, a correspondence table may be used in the flowchart of Figure 14 to obtain the response identifier. In this case, in step S1402, the response determination unit 133 obtains and stores the response identifier contained in the response information determined in step S1401. Also, in this case, steps S1403 and S1404 are unnecessary.

[0230] Next, an example of the processing of the basis information in step S1002 will be explained using the flowchart in Figure 15.

[0231] (Step S1501) The basis information acquisition unit 134 assigns 1 to counter i.

[0232] (Step S1502) The basis information acquisition unit 134 determines whether or not the i-th counter identifier exists among the counter identifiers acquired in step S1002. If the i-th counter identifier exists, the process proceeds to step S1503; otherwise, it returns to the higher-level process.

[0233] (Step S1503) The basis information acquisition unit 134 acquires the source information. The source information is the information used when acquiring the basis level. The source information is, for example, the class identifier corresponding to the counter identifier and the score corresponding to the counter identifier.

[0234] (Step S1504) The basis information acquisition unit 134 acquires the basis level corresponding to the original information acquired in step S1503.

[0235] Furthermore, the basis information acquisition unit 134 acquires, for example, the basis level corresponding to the class identifier acquired in step S1503. In this case, two or more basis levels paired with each class identifier are stored in the storage unit 11.

[0236] Furthermore, the evidence information acquisition unit 134 acquires, for example, the evidence level corresponding to the score acquired in step S1503. In this case, the evidence level paired with each of the two or more conditions of the score is stored in the storage unit 11. The conditions of the score are usually information indicating the range of the score. The score is the score acquired in the machine learning prediction process.

[0237] (Step S1505) The basis information acquisition unit 134 acquires one or more training data that served as the basis for acquiring the i-th response identifier.

[0238] (Step S1506) The evidence information acquisition unit 134 uses one or more training data acquired in step S1505 to acquire effect information corresponding to each training data. Next, the evidence information acquisition unit 134 uses one or more effect information to acquire effectiveness information, associates it with the i-th counter identifier, and stores it in a buffer (not shown).

[0239] (Step S1507) The evidence information acquisition unit 134 acquires satisfaction levels from each of the one or more training data acquired in step S1505.

[0240] (Step S1508) The basis information acquisition unit 134 acquires satisfaction information using one or more satisfaction levels obtained in step S1507, associates it with the i-th response identifier, and stores it in a buffer (not shown).

[0241] (Step S1509) The basis information acquisition unit 134 increments counter i by 1. Return to step S1502.

[0242] In the flowchart in Figure 15, evidence level, effectiveness information, and satisfaction information were obtained, but it is also acceptable to obtain only one or two types of information from these.

[0243] Next, an example of reward information processing in step S1003 will be explained using the flowchart in Figure 16.

[0244] (Step S1601) The reward information acquisition unit 135 assigns 1 to counter i.

[0245] (Step S1602) The reward information acquisition unit 135 determines whether or not the i-th action identifier exists. If the i-th action identifier exists, the process proceeds to step S1603; otherwise, it returns to the higher-level process.

[0246] (Step S1603) The reward information acquisition unit 135 determines whether the i-th handling identifier matches the reward conditions. If it matches the reward conditions, the process proceeds to step S1604; otherwise, the process proceeds to step S1606.

[0247] The reward information acquisition unit 135 determines whether the i-th action identifier matches the reward conditions, for example, by using one or more of the following: the i-th action identifier, the rationale level associated with the i-th action identifier, the effectiveness information, and the satisfaction information.

[0248] The reward conditions are the conditions for obtaining reward information. For example, the reward conditions are: "the reward information associated with the i-th response identifier is stored in the storage unit 11," "the rationale level associated with the i-th response identifier is information indicating that there is no rationale," "the rationale level associated with the i-th response identifier is below or less than the threshold," "the effect indicated by the effect information associated with the i-th response identifier is low enough to satisfy the conditions," and "the degree of satisfaction indicated by the satisfaction information associated with the i-th response identifier is low enough to satisfy the conditions."

[0249] (Step S1604) The reward information acquisition unit 135 acquires reward information from the storage unit 11. The reward information acquisition unit 135 acquires, for example, reward information corresponding to the i-th action identifier from the storage unit 11. The reward information acquisition unit 135 acquires, for example, reward information corresponding to the reward condition from the storage unit 11.

[0250] (Step S1605) The reward information acquisition unit 135 temporarily stores the reward information acquired in step S1604 in a buffer (not shown) as output reward information, associating it with the i-th response identifier.

[0251] (Step S1606) The reward information acquisition unit 135 increments counter i by 1. Return to step S1602.

[0252] Next, an example of the operation of terminal device 2 will be described. The terminal receiving unit 22 of terminal device 2 receives user information. Next, the terminal processing unit 23 uses the user information to configure user information to be transmitted. The terminal transmitting unit 24 transmits the user information to the information processing device 1. Next, in response to the transmission of user information, the terminal receiving unit 25 receives output information from the information processing device 1. Next, the terminal processing unit 23 uses the output information to configure information to be output. The terminal output unit 26 outputs the configured information. The output information includes, for example, a response identifier, justification information, and reward information.

[0253] Furthermore, the terminal receiving unit 22 of the terminal device 2 may also receive training data. In this case, the terminal processing unit 23 uses the training data to construct the training data to be transmitted. Next, The terminal transmission unit 24 transmits the training data to the information processing device 1.

[0254] The following describes a specific example of the operation of information system A in this embodiment.

[0255] Let's assume that the training data storage unit 111 of the information processing device 1 stores a training data management table containing two or more training data transmitted from the terminal device 2. The training data is information about a user who has taken some action, and has first result information and second result information. This training data management table is shown in Figure 17. The training data management table manages two or more records, each containing "ID", "Action Identifier", "First Result Information", "Second Result Information", "Gender", "Age", "Height", "Weight", and "Response Information". The "Response Information" is the user's response to a questionnaire, and in this case, it includes "Initiative Information", "Lifestyle Information", "Goal", and "Satisfaction Level". The "First Result Information" is the test result before taking action, identified by the Action Identifier. The "Second Result Information" is the test result after taking action, identified by the Action Identifier. In this case, the test result is the systolic blood pressure. The "Goal" is the systolic blood pressure the user aims for. The "Satisfaction Level" is the degree of satisfaction of the user who took action, identified by the Action Identifier.

[0256] Furthermore, the storage unit 11 of the information processing device 1 stores the corresponding identifiers "Product A (reduced-sodium soy sauce)", "Product B (reduced-sodium miso)", and "Product C (yogurt)". In this case, the corresponding action for "Product A" is a program (Program A) that uses Product A continuously for one month, the corresponding action for "Product B" is a program (Program B) that uses Product B continuously for one month, and the corresponding action for "Product C" is a program (Program C) that uses Product C continuously for two months. A program refers to the continuous consumption of a product or service, and is also called a challenge program.

[0257] Furthermore, the storage unit 11 stores a basis level determination table having a basis level "3" associated with the class identifier "Class A", a basis level "2" associated with the class identifiers "Class B" and "Class D", a basis level "1" associated with the class identifier "Class C", and a basis level "0" associated with the class identifier "None".

[0258] The evidence level determination table may also be a table for determining the evidence level using one or more pieces of information from effectiveness information and satisfaction information, as shown in Figure 18. Such an evidence level determination table has an "evidence level" and a "determination condition". The "determination condition" is the condition for determining the evidence level. In other words, the evidence information acquisition unit 134 may acquire the evidence level using one or more pieces of information from effectiveness information and satisfaction information. Furthermore, the determination condition for "evidence level = 1 to 3" is applied, for example, when the training data is equal to or greater than a threshold (e.g., 100 items). Furthermore, in "evidence level = 1 to 3" in Figure 18, it is assumed that the conditions are applied in the order of "evidence level = 3", "evidence level = 2", and "evidence level = 1".

[0259] Furthermore, the storage unit 11 stores the target achievement condition "Target achievement rate >= 80%". The rationale information acquisition unit 134 then acquires recommendation information to strongly recommend action for action identifiers that satisfy the target achievement condition. Here, the recommendation information is "necessary" (for example, "1"). Note that "necessary" means that it is necessary to take the action identified by the action identifier in order to achieve the goal. The rationale information acquisition unit 134 also acquires recommendation information "unnecessary" (for example, "0") for action identifiers that do not satisfy the target achievement condition.

[0260] Furthermore, the storage unit 11 stores the reward condition "basis level = 0" and the corresponding reward information "30% OFF". In other words, for products corresponding to "basis level = 0", the reward information of 30% OFF (information indicating that the product can be purchased at a 30% discount) is presented.

[0261] In this situation, the classification unit 131 of the information processing device 1 acquires the first result information, second result information, target, and satisfaction level for each training data in Figure 17 for each response identifier. Next, the classification unit 131 acquires effectiveness information for each response identifier and for each training data using the first result information, second result information, and target. The effectiveness information here is the target achievement rate. The target achievement rate is calculated, for example, by "(first result information - second result information) / (first result information - target) * 100 (%)".

[0262] Next, the classification unit 131 classifies the training data for each treatment identifier using the effectiveness information and satisfaction level. Here, the classification unit 131 classifies the training data for each treatment identifier as follows: for example, training data with an effectiveness information of 60% or more and a satisfaction level of "4 or 5" is classified as "Class A"; training data with an effectiveness information of less than 60% and a satisfaction level of "4 or 5" is classified as "Class B"; training data with an effectiveness information of less than 60% and a satisfaction level of "1, 2, or 3" is classified as "Class C"; and training data with an effectiveness information of 60% or more and a satisfaction level of "1, 2, or 3" is classified as "Class D". Figure 19 illustrates this concept of training data classification.

[0263] Next, the learning information acquisition unit 132 acquires "first result information," "gender," "age," "height," "weight," "lifestyle information," and "goal" for each training data from the training data management table in Figure 17 for each response identifier, and constructs a vector with the "first result information," "gender," "age," "height," "weight," "lifestyle information," and "goal" as explanatory variables and the class identifier as the objective variable.

[0264] Next, the learning information acquisition unit 132 performs machine learning learning processing using two or more vectors for each corresponding identifier, constructs a learner for each corresponding identifier, and stores the learner for each corresponding identifier in the learning information storage unit 112, associating it with the corresponding identifier. This learner is a learner for predicting a class identifier using a vector containing "first result information," "gender," "age," "height," "weight," "lifestyle information," and "goal."

[0265] In the above situation, suppose user A inputs user information into terminal device 2, which has the following characteristics: "<First result information>183 <Gender>Male <Age>68 <Height>175 <Weight>88 <Lifestyle information>Smoking <Target>140". Next, terminal device 2 receives the user information and transmits it to information processing device 1.

[0266] Next, the user information receiving unit 121 of the information processing device 1 receives the user information from the terminal device 2 of user A.

[0267] Next, the processing unit 13 performs output information acquisition processing as follows. That is, first, the countermeasure determination unit 133 determines a class identifier for each of the countermeasure identifiers "Product A (low-salt soy sauce)", "Product B (low-salt miso)", and "Product C (yogurt)". That is, first, the countermeasure determination unit 133 acquires a learning device paired with the countermeasure identifier "Product A (low-salt soy sauce)" from the learning information storage unit 112. Next, the countermeasure determination unit 133 provides the user attribute value "<First result information> 183 <Gender> male <Age> 68 <Height> 175 <Weight> 88 <Lifestyle information> Smoking <Goal> 140" included in the user information and the acquired learning device to a prediction module that performs prediction processing of machine learning, executes the prediction module, and assumes that the class identifier "Class A" is acquired. Similarly, the countermeasure determination unit 133 acquires a learning device paired with the countermeasure identifier "Product B (low-salt miso)" from the learning information storage unit 112. Next, the countermeasure determination unit 133 provides two or more user attribute values included in the user information and the acquired learning device to a prediction module that performs prediction processing of machine learning, executes the prediction module, and assumes that the class identifier "Class D" is acquired. Note that it is assumed that the countermeasure determination unit 133 could not acquire a learning device paired with the countermeasure identifier "Product C (yogurt)" from the learning information storage unit 112.

[0268] Next, the basis information acquisition unit 134 acquires the basis level "3" paired with the class identifier "Class A" corresponding to the countermeasure identifier "Product A (low-salt soy sauce)" from the storage unit 11. The basis information acquisition unit 134 also acquires the basis level "2" paired with the class identifier "Class D" corresponding to the countermeasure identifier "Product B (low-salt miso)" from the storage unit 11. Further, the basis information acquisition unit 134 acquires the basis level "0" paired with the countermeasure identifier "Product C (yogurt)" for which the class could not be determined from the storage unit 11.

[0269] Next, for each countermeasure identifier for which the class identifier has been acquired, the basis information acquisition unit 134 acquires one or more teacher data corresponding to the countermeasure identifier and the class identifier. Then, the basis information acquisition unit 134 acquires the effect information of each teacher data using the acquired one or more teacher data. Next, the basis information acquisition unit 134 acquires "effective information", which is the average of the effect information of each teacher data. It is assumed that the basis information acquisition unit 134 has acquired "80%" of effective information from one or more teacher data corresponding to the countermeasure identifier "Product A (low-salt soy sauce)" and the class identifier "Class A". Also, the basis information acquisition unit 134 acquires the satisfaction level possessed by each of the acquired one or more teacher data. Next, it is assumed that the basis information acquisition unit 134 has acquired "60%" of satisfaction information, which is the ratio of teacher data having a satisfaction level of "4" or "5" in the same class.

[0270] Similarly, it is assumed that the basis information acquisition unit 134 has acquired one or more teacher data corresponding to the class identifier "Class D" corresponding to the countermeasure identifier "Product B (low-salt miso)", and using the teacher data, has acquired "80%" of effective information and "30%" of satisfaction information.

[0271] Also, the basis information acquisition unit 134 acquires the target achievement condition "target achievement rate >= 80%" from the storage unit 11. Next, for each countermeasure identifier, the basis information acquisition unit 134 acquires the countermeasure identifier "Product A (low-salt soy sauce)" for which the average of the target achievement rates of one or more teacher data for the countermeasure identifier and the corresponding class identifier satisfies the target achievement condition. Next, the basis information acquisition unit 134 acquires "necessary" as recommended information in association with the countermeasure identifier "Product A (low-salt soy sauce)". Also, the basis information acquisition unit 134 acquires "not necessary" as recommended information in association with the countermeasure identifiers "Product B (low-salt miso)" and "Product C (yogurt)" that do not satisfy the target achievement condition.

[0272] Also, the basis information acquisition unit 134 was unable to acquire the effective information and satisfaction information corresponding to the countermeasure identifier "Product C (yogurt)". This is because the basis level is "0".

[0273] Next, the reward information acquisition unit 135 refers to the reward condition "basis level = 0" in the storage unit 11 and acquires 30% OFF reward information for the corresponding action identifier "product C (yogurt)" for "basis level = 0".

[0274] Next, the processing unit 13 constructs output information using the acquired action information, rationale information, recommendation information, and reward information. Then, the information output unit 141 transmits the acquired output information to the terminal device 2.

[0275] Next, terminal device 2 receives and outputs output information. An example of such output is shown in Figure 20. In Figure 20, 2001 is the evidence level, which is referred to here as the "evidence level." Also, 2002 is evidence information containing effectiveness information and satisfaction information, and is labeled "Reason." 2003 is recommendation information, and is labeled "To achieve ToBe." Furthermore, 2004 is reward information, and is labeled "Incentive."

[0276] As described above, according to this embodiment, it is possible to determine appropriate actions to take in response to the inspection results.

[0277] Furthermore, according to this embodiment, when proposing a course of action in response to the inspection results, the basis for that proposal can be presented.

[0278] Furthermore, according to this embodiment, when proposing a course of action regarding the inspection results, it is also possible to clearly indicate that there is no basis for that proposal.

[0279] Furthermore, according to this embodiment, when a solution to the test results is proposed, it can be shown that an incentive is given to the user to take that solution.

[0280] Furthermore, the processing in this embodiment may be implemented in software. This software may be distributed via software download or the like. Alternatively, this software may be recorded on a recording medium such as a CD-ROM and distributed. This also applies to other embodiments described herein. The software that implements the information processing device in this embodiment is the following program. In other words, this program is a program that causes a computer to function as a user information receiving unit that receives user information having one or more user attribute values ​​including result information that identifies the results of a biological examination of a single user; a learning information storage unit that stores learning information based on two or more training data sets, each containing two or more user attribute values, which are associated with a response identifier that identifies the action taken by the user, and each training data set includes a first result information that identifies the examination results before the action was taken and a second result information that identifies the examination results after the action was taken; a response determination unit that uses the learning information and the user information received by the user information receiving unit to obtain a response identifier that identifies the action corresponding to the result information held by the user information; a basis information acquisition unit that uses one or more training data sets corresponding to the response identifier obtained by the response determination unit and the user information received by the user information receiving unit to obtain basis information regarding the basis for recommending the action identified by the response identifier obtained by the response determination unit; and an information output unit that outputs the response identifier obtained by the response determination unit and the basis information obtained by the basis information acquisition unit.

[0281] Figure 21 also shows the external appearance of a computer that executes the program described herein to realize the information processing device 1 and other devices of the various embodiments described above. The embodiments described above can be realized with computer hardware and computer programs executed thereon. Figure 21 is an overview of this computer system 300, and Figure 22 is a block diagram of the system 300.

[0282] In Figure 21, the computer system 300 includes a computer 301 with a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.

[0283] In Figure 22, the computer 301 includes, in addition to the CD-ROM drive 3012, an MPU 3013, a bus 3014 connected to the CD-ROM drive 3012, a ROM 3015 for storing programs such as boot-up programs, 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 that provides connectivity to a LAN.

[0284] The program that causes the computer system 300 to execute the functions of the information processing device 1, etc., as described above, may be stored on the 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 during execution. The program may also be loaded directly from the CD-ROM 3101 or the network.

[0285] The program does not necessarily have to include an operating system (OS) or third-party program that causes the computer 301 to execute the functions of the information processing device 1, etc., as described above. The program only needs to include the instruction portion that calls the appropriate function (module) in a controlled manner and obtains the desired result. How the computer system 300 operates is well known, so a detailed explanation is omitted.

[0286] In the above program, in steps such as the step of transmitting information and the step of receiving information, processing performed by hardware, for example, processing performed by a modem or an interface card in the transmission step (processing that can only be performed by hardware) is not included.

[0287] Also, the computer that executes the above program may be singular or plural. That is, centralized processing may be performed, or distributed processing may be performed.

[0288] Also, in each of the above embodiments, it is needless to say that two or more communication means existing in one device may be physically realized by one medium.

[0289] Also, in each of the above embodiments, each process may be realized by being centrally processed by a single device, or may be realized by being distributedly processed by a plurality of devices.

[0290] Needless to say, the present invention is not limited to the above embodiments, and various modifications are possible, and those are also included in the scope of the present invention.

Industrial Applicability

[0291] As described above, the information processing apparatus according to the present invention has an effect that when proposing a countermeasure against the result of inspection, the basis for the proposal can be presented, and it is useful as a server or the like that outputs a countermeasure identifier for identifying the recommended countermeasure and basis information.

Explanation of Signs

[0292] 1 Information processing apparatus 3 Terminal device 11 Storage unit 12 Reception unit 13 Processing unit 14 Output unit 21 Terminal storage unit 22 Terminal reception unit 23 Terminal processing unit 24 Terminal transmission unit 25 Receiving part of the terminal 26 Terminal output section 111 Training data storage unit 112 Learning Information Storage Unit 121 User Information Reception Department 131 Classification Department 132 Learning Information Acquisition Department 133 Decision-making department 134 Base Information Acquisition Department 134 Information Acquisition Department 135 Compensation Information Acquisition Department 141 Information Output Unit

Claims

1. A user information receiving unit that receives user information having one or more user attribute values, including result information that identifies the results of a biological examination of a single user, A learning information storage unit acquires learning information from a learning information storage unit which stores learning information based on two or more training data sets, each training data set having two or more user attribute values, including a first result information that identifies the inspection result before the action was taken and a second result information that identifies the inspection result after the action was taken; and an action determination unit acquires an action identifier that identifies the action according to the result information held by the user information receiving unit, using the learning information and the user information received by the user information receiving unit. A basis information acquisition unit acquires basis information relating to the basis for recommending an action identified by the action identifier acquired by the action decision unit, using one or more training data corresponding to the action identifier acquired by the action decision unit and the user information received by the user information receiving unit. An information processing apparatus comprising: an information output unit that outputs the countermeasure identifier acquired by the countermeasure determination unit and the basis information acquired by the basis information acquisition unit.

2. Each of the two or more aforementioned teacher data sets corresponds to a class. The aforementioned information acquisition unit, The information processing device according to claim 1, which acquires the basis information using one or more training data associated with the class to which the user information belongs.

3. The one or more user attribute values ​​mentioned above have one or more pieces of information, such as a goal or goal achievement rate. The aforementioned information acquisition unit, Determine the action identifier that satisfies the objective achievement conditions stored in the storage unit, and obtain recommendation information to strongly recommend that action identifier be used. The aforementioned information output unit is An information processing device according to claim 1 or claim 2, which outputs the aforementioned recommended information.

4. The aforementioned information acquisition unit, An information processing device according to any one of claims 1 to 3, wherein the action determination unit obtains one or more training data corresponding to the action identifier obtained and the user information received by the user information receiving unit, and obtains evidence information which includes one or more types of information, such as evidence level, which identifies the degree of strength of the evidence for recommending the action, and reason information, which indicates the reason for recommending the action and includes effectiveness information regarding the effectiveness of the action if it is performed, or satisfaction information regarding the satisfaction level after the action is performed.

5. The aforementioned information acquisition unit, Using the aforementioned effectiveness information or satisfaction information, obtain the evidence level, The aforementioned information output unit is The information processing device according to claim 4, which outputs the basis information including the basis level.

6. The aforementioned information acquisition unit, An information processing device according to any one of claims 1 to 5, which, if it is not possible to obtain supporting information, obtains supporting information that indicates there is no supporting information.

7. The system further comprises a reward information acquisition unit that acquires reward information which is information that identifies a reward for recommending the user to take the aforementioned action, and which corresponds to the aforementioned justification information. The aforementioned information output unit is The information processing apparatus according to any one of claims 1 to 6, which also outputs the reward information acquired by the reward information acquisition unit.

8. The aforementioned reward information acquisition unit, The information processing device according to claim 7, which acquires reward information that identifies a reward corresponding to the level of justification included in the aforementioned justification information.

9. The system further comprises a classification unit that, for each response identifier, classifies the two or more training data into two or more classes using effect information, which is information obtained using the second result information contained in each of the two or more training data, and which is information concerning the effectiveness of the response, and associates each of the two or more training data with a class identifier that identifies the class. The aforementioned action decision unit, The information processing apparatus according to any one of claims 1 to 7, wherein the user information received by the user information receiving unit and the learning information based on the two or more training data determine the class to which the user information belongs for each of the two or more corresponding identifiers, and distinguishes and obtains the corresponding identifier corresponding to the class with a large difference between the first result information and the second result information and the corresponding identifier corresponding to the class with a small difference.

10. The aforementioned training data also includes the user satisfaction level as a result of the aforementioned measures. The aforementioned classification unit is The information processing apparatus according to claim 9, wherein for each of the two or more training data, the two or more training data are classified into two or more classes using the effect information and satisfaction level for each of the two or more training data, and each of the two or more training data is associated with a class identifier that identifies the class.

11. Each of the two or more classes classified by the aforementioned classification unit is associated with a rationale level based on effectiveness information for the training data corresponding to that class. The aforementioned information acquisition unit, The information processing apparatus according to claim 9, which acquires a basis level corresponding to the class, acquires effectiveness information using effectiveness information which is information obtained using the second result information contained in the training data corresponding to the class determined by the action decision unit and is information relating to the effect of the action, and acquires basis information which is reason information which is information including the basis level and the effectiveness information.

12. The aforementioned training data also includes the user satisfaction level as a result of the aforementioned measures. The aforementioned classification unit is For each of the two or more training data, the effectiveness information and satisfaction level for each of the two or more training data are used to classify the two or more training data into two or more classes, and each of the two or more training data is associated with a class identifier that identifies the class. The aforementioned information acquisition unit, The information processing apparatus according to claim 11, which obtains satisfaction information using the satisfaction level of training data corresponding to the class to which the user information belongs, and obtains justification information including reason information that includes the satisfaction level information.

13. The system further comprises a learning information acquisition unit that acquires learning information using two or more training data sets having one or more user attribute values ​​that satisfy the same conditions as the one or more user attribute values ​​included in the user information received by the user information receiving unit, The information processing apparatus according to any one of claims 1 to 12, wherein the learning information in the learning information storage unit is learning information acquired by the learning information acquisition unit.

14. The information processing device according to any one of claims 1 to 13, wherein the measure is a challenge to consume the product for a certain period of time or longer in order to improve the test results, and the training data includes the degree of participation in the challenge, which is information from responses to a questionnaire about the challenge.

15. An information processing method comprising all the steps performed by the information processing device described in any one of claims 1 to 14.

16. Computers, A program for causing an information processing device to function as described in any one of claims 1 to 14.

Images