Activity support method, program, and activity support device

Abstract

To provide an activity support method, a program, and an activity support device that are capable of allowing a user to learn activities having more effective contents for achieving a goal.SOLUTION: An activity support method includes the following steps of: obtaining a first achievement goal related to an activity performed by a first user; obtaining a plurality of index values representing a performance according to an execution result of the activity by the first user; specifying a related index that changes in conjunction with a goal parameter according to the first achievement goal from the plurality of index values; and outputting information related to an improvement of the related index according to the performance of the activity by the first user.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to an activity support method, a program, and an activity support device. [Background technology]

[0002] Conventionally, many runners have attempted to achieve their goals on their own without the assistance of a trainer. Meanwhile, there is technology that measures the exercises and other aspects of a user's activities and evaluates the activity content based on the measurement results. Patent Document 1 discloses a technology that works in cooperation with fitness clubs, golf courses, etc. to propose exercise menus and other items for improving desired sports abilities based on the user's exercise performance to date. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2014-183867 Summary of the Invention [Problem to be solved by the invention]

[0004] However, even if individual users perform activities to achieve the same goal, their abilities do not improve uniformly, and each user may have weaknesses or stagnation in their abilities. Therefore, in the past, overcoming problems and achieving goals was time-consuming, and even more so, users were forced to grope around without knowing how to achieve the goal.

[0005] An object of the present invention is to provide an activity support method, a program, and an activity support device that enable a user to learn activities that are more effective in achieving a goal. [Means for solving the problem]

[0006] In order to achieve the above object, the present invention provides obtaining a first goal related to an activity performed by a first user; acquiring a plurality of index values ​​representing performance according to the results of the activity performed by the first user; Identifying a related index that changes in conjunction with a target parameter corresponding to the first achievement goal from the plurality of index values; outputting information relating to an improvement in the related indicator in response to the first user performing the activity; It is a method of supporting activities. [Effects of the Invention]

[0007] According to the present invention, it is possible to obtain an effect that a user can learn activities with more effective content for achieving a goal. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a configuration diagram showing an activity support system. [Figure 2] FIG. 2 is a block diagram showing a functional configuration of a server device. [Figure 3] 10 is a table illustrating the contents of registration information and history information. [Figure 4] 10A and 10B are diagrams illustrating the flow of the activity support operation according to the present embodiment. [Figure 5] 10 is a flowchart showing a control procedure of an improvement prediction process executed by the server device. [Figure 6] 10 is a flowchart showing a control procedure of an improvement support process executed by the server device. [Figure 7] FIG. 10 is another example of a configuration diagram of an activity support system. [Figure 8] FIG. 10 is a block diagram showing a functional configuration of a server device according to another example. [Figure 9] 10 is a table illustrating the contents of another example of user registration information. [Figure 10] FIG. 10 is a diagram illustrating a flow of activity support in another example. [Figure 11] 10A and 10B are diagrams illustrating examples of a screen for setting a target to be achieved and a screen for displaying the discrepancy between prediction and actual results. [Figure 12] 10A and 10B are diagrams showing other examples of the screen for setting a target to be achieved and the screen for displaying the discrepancy between prediction and actual results. [Figure 13] FIG. 10 is a diagram showing another example of a display screen showing the discrepancy between prediction and actual results. DETAILED DESCRIPTION OF THE INVENTION

[0009] 1, an activity support system 100 according to an embodiment of the present invention includes an electronic device 1, measuring devices 3 and 5, and a server device 7.

[0010] The measuring devices 3 and 5 measure the user's movements, biological information, and the like, and transmit the measurement data to the electronic device 1. For example, the measuring device 3 is worn on the user's wrist, and the measuring device 5 is worn on the user's waist. The measuring devices 3 and 5 may each include some or all of an acceleration sensor, a direction sensor, a gyro sensor, a pulse sensor, and the like. Communication between the measuring devices 3 and 5 and the electronic device 1 may be possible via short-range wireless communication, for example, Bluetooth (registered trademark).

[0011] The electronic device 1 issues measurement commands to the measuring devices 3 and 5, integrates the measured data, and transmits the integrated data to the server device 7. The electronic device 1 may be a mobile terminal such as a smartphone. The electronic device 1 can receive and display analysis results obtained by the server device 7 based on the measurement results, as well as a training plan according to the goals set in the electronic device 1. Data communication between the electronic device 1 and the server device 7 is possible via an internet line. The electronic device 1 may also perform partial analysis processing when integrating the data.

[0012] The server device 7 is an activity support device of this embodiment, and acquires and stores measurement data of registered users from the electronic device 1. The server device 7 performs an analysis operation based on the acquired measurement data and transmits the analysis results and the like to the electronic device 1. The server device 7 may be a general-purpose electronic computer.

[0013] As shown in the block diagram of FIG. 2, the server device 7 includes a control unit 71, a RAM (Random Access Memory) 72, a storage unit 73, a communication unit 74, a display unit 75, an operation reception unit 76, and the like.

[0014] The control unit 71 has a processor that performs arithmetic processing and controls the overall operation of the server device 7. The processor may be a general-purpose or optimized CPU (Central Processing Unit). The CPU may be a single CPU or multiple CPUs that perform parallel processing or operate independently depending on the processing content, etc. The RAM 72 provides working memory space for the control unit 71 and stores temporary data.

[0015] The storage unit 73 has a nonvolatile memory that stores and holds data. The nonvolatile memory may be, for example, a flash memory or a hard disk drive (HDD), but is not limited to these. The storage unit 73 stores and holds a program 731, a machine learning model 7311, registration information 732, history information 733, etc.

[0016] The program 731 includes a control program related to activity support. The contents of the program will be described later. The machine learning model 7311 estimates the user's ability improvement. The machine learning model 731 is generated and stored according to the user's ability and goals. The registration information 732 includes information about registered users regarding the analysis operations by the server device 7 and the provision of advice based on the analysis results. The history information 733 stores a history of measurement results and analysis results measured when the user performs an activity. The history information 733 also stores information about goals set by the user. The registration information 732 and the history information 733 may be stored together in a database device connected to the server device 7 directly or via a network. The registration information 732 and the history information 733 need to store data for a large number of users together and be quickly searchable. Therefore, the database device may be designed and configured specifically to enable efficient storage and retrieval.

[0017] As described above, the communication unit 74 is capable of communicating with the electronic device 1 via the Internet. The communication may be performed according to protocols such as TCP / IP, HTTP, etc. The communication unit 74 includes a network card for performing communication according to these protocols.

[0018] The display unit 75 has a digital display screen and can display characters, symbols, charts, images, etc. under the control of the control unit 71. The digital display screen may be a liquid crystal display screen, an organic EL (Electro-Luminescent) screen, etc., but is not limited to these.

[0019] The operation reception unit 76 receives an input operation from outside and outputs the content of the received operation as an electrical signal to the control unit 71. The operation reception unit 76 may include, for example, a keyboard and a pointing device such as a mouse.

[0020] The server device 7 has a control unit 71, a RAM 72, a storage unit 73, and a communication unit 74 as a computer main body, and may be equipped with peripheral devices such as a display unit 75 and an operation reception unit 76 as needed. In other words, the server device 7 normally operates according to the content received from the external electronic device 1, and does not need to directly receive command operations via the operation reception unit 76 or display on a display screen.

[0021] 3(a), the registration information 732 includes information about the registered user. The user information includes the user's identification ID and personal information such as age and / or date of birth, gender, height, etc. The personal information may also include information about activities other than running that the user is engaged in.

[0022] As shown in FIG. 3(b), the history information 733 stores, for each user and activity, indicators based on the measurement results of the activity along with information on the date and time of the activity. Movements in an activity can be evaluated using multiple indicator values ​​(parameters) as performance indicators corresponding to the results of the activity. Examples of running indicators include, but are not limited to, stride, pitch, push-off time, impact and direction upon landing, vertical and horizontal fluctuations of the center of gravity, body tilt and swing, and arm swing angle and angle range. For example, in FIG. 3(b), indicator 1 may be stride and indicator 2 may be pitch. sn_Tq represents the indicator value for indicator n on date Tq. Examples of indirect indicators include weight, weight change rate, body fat percentage, pulse rate, and oxygen saturation. Activity result indicators include distance traveled, running time, lap time, split time, and running speed. Here, N indicators are arranged in groups of Q. The number of indicators depends on the type of activity, and the amount of data increases with the number of times the activity is performed. Furthermore, a measuring device having a life log function that measures daily energy consumption in addition to activity measurement may also be used. In this case, information such as daily calorie consumption, activity calorie consumption, and number of steps taken may also be stored in the history information 733.

[0023] As shown in FIG. 3(c), the history information 733 stores a list of goals set by the user, along with the date of setting and the achievement status. A user can set one goal at a time, but multiple goals may be set simultaneously. The achievement goals are preferably defined in a common format that is easy to search for for all users. In the history information 733, the index data shown in FIG. 3(b) and the goal data shown in FIG. 3(c) may be stored in any format as long as they are associated with each other.

[0024] Next, we will explain user activities and activity support. In activities such as running, each user has various motivations and goals for engaging in the activity. For example, running goals can include recovering leg strength and cardiopulmonary function, extending the distance one can run, shortening the time it takes to run a specific distance such as 10 km or a marathon, losing weight, and reducing body fat percentage. Among the above-mentioned indices, the result indicators can also be indicator values ​​(target parameters) related to training goals. Other target parameters include weight, weight loss, and body fat percentage. That is, the goal is achieved when the target parameters reach their target values. For example, if the goal is to run 10 km in one hour, the condition for achieving the goal is to achieve the target parameter, which is the time it takes to run 10 km or the split time for a run of 10 km or more, under one hour. The degree of improvement in the target parameters can also be the goal achievement rate.

[0025] There may be a correlation between changes in specific indicators and improvements in target parameters. In other words, when the index value of such an indicator improves, the target parameter also changes and improves accordingly. Since there are often multiple such indicators (related indicators) for one target parameter, an improvement in some related indicators may not necessarily result in a sufficient improvement in the target parameter. In other words, the target parameter can be improved by improving running style in accordance with these related indicators. When performing an activity to achieve a goal, a training direction can be determined and output so as to improve indicators that are directly related to the goal. Some of these correlations and training directions may already have logically established methods. On the other hand, in this embodiment, indicators that are directly correlated with changes (improvements) in target parameters are numerically detected from the training data of multiple users. In each training session, it is determined whether any of these indicators have not improved appropriately. If any indicator has not shown appropriate improvement, advice on improving that indicator is output. In this case, even if data of other users with characteristics significantly different from those of the user (first user) to receive advice is referenced, meaningless results that are not useful to the target user may be obtained. In this embodiment, the correlation (relationship) is calculated based on the practice data of another user (second user) who has corresponding user characteristics and a goal (second achievement goal) in the same range. Corresponding user characteristics include similarities in physical features and behavioral tendencies. Goals in the same range may be defined as, for example, a difference in running distance within ±10% and a difference in target running time within ±10%.

[0026] As shown in FIG. 4 , in this embodiment, the history information 733 of all users stored in the server device 7 is referenced (P1), and practice data of a second user who has similar physical characteristics and behavioral tendencies as the support-receiving user and has a goal in the same range is extracted as similarly associated data (P2). The user's own data is extracted separately (P3). The similarly associated data is further classified into data of a third user who is a user who achieved the goal (P5) and data of a fourth user who is a user who did not achieve the goal (P6). Here, all indicators stored in the all-user history information, N indicators in this case, are acquired. Note that, among the fourth users who did not achieve their goals, data of a user with an even lower goal achievement rate may be extracted as goal non-achievement data. The goal achievement rate may be calculated, for example, by the ratio of the difference between the value of the target parameter (initial value) at the time of goal setting and the target setting value of the target parameter to the difference between the value of the target parameter finally achieved and the initial value.

[0027] The similarity of physical characteristics can be determined by, for example, height, weight, sex, age, etc. Height, weight, age, etc. may be determined to be similar if they match within a range (first reference range) of, for example, ±5 cm, ±5 kg, or ±5 years (first condition). Alternatively, a relative range, such as ±10% of each value of the target user, may be set. The ranges may also be set according to different standards depending on age, height, and weight. This information may be acquired from the above-mentioned registered information 732. The latest data of the weight and body fat percentage set by the user's input operation in history information 733 may be acquired.

[0028] The similarity of behavioral tendencies is determined by considering similarities in basic items such as the frequency of training (running), the duration of each session, and the duration of previous training and blank periods. In addition, the status of activities other than running and the amount of exercise in daily life may be included as additional items in determining similarity. Information on basic items may be identified in advance from the user's training history or may be identified each time from the data acquired in process P3. Activities other than running may include, for example, walking, cycling, trail running, mountain climbing, and swimming. The amount of exercise in daily life may include the amount of walking during commuting, the amount of activity during work, and the amount of housework. Some or all of these may be measurable by the measuring devices 3, 5, etc. For exercises that cannot be acquired by the measuring devices 3, 5, etc., the date and type of activity may be separately acquired by the user's input operation. For strength training that is difficult to measure, such as abdominal exercises, back exercises, push-ups, and pull-ups, the activity status may be separately acquired by the user's input operation, or may simply be included in daily life exercise. For example, a quantitative evaluation value may be obtained by adding points according to the number of matches in additional items to a value indicating the degree of similarity in the basic items. The quantitative evaluation value may be compared with a reference value (second reference range) to determine the similarity in behavioral tendencies (second condition). In addition, the similarity between physical characteristics and behavioral tendencies in this embodiment includes cases where height, training frequency, etc. are the same.

[0029] Each user can set a goal (first achievement goal) that they want to achieve for the activity they are performing by inputting the goal as appropriate. The set goal is additionally registered in the setting data shown in FIG. 3(c) above. If the goal is changed to another goal before it is achieved, the goal before the change may be considered not achieved (failed to achieve). A deadline for achieving the goal may also be set. In this case, if the goal is not achieved by the deadline but is maintained as is and achieved later than the deadline, this may be treated separately from a normal achievement or failure to achieve. If the goal is canceled without being achieved before the deadline, this may also be treated separately from a normal failure to achieve.

[0030] A target parameter corresponding to the goal is determined (P4). The correspondence between the goal and the target parameter may be stored in advance in the storage unit 73. For the goal achievement data of the third user, correlations between N-1 indices other than the goal achievement parameter and the target parameter are calculated, and an index having a higher correlation than a certain first criterion is identified as a highly correlated index (first index) (P7). The correlation may be determined, for example, by a correlation coefficient (Pearson's product-moment correlation coefficient) or by another type of correlation index. The type of correlation index calculated may differ depending on the set goal. In this case, the type of correlation index, a formula or correspondence table for calculating the correlation, and a first criterion for determining a highly correlated index may be stored in advance in the storage unit 73 for each goal. For example, the target parameter for a goal of running 10 km in one hour may be the time it takes to run 10 km. The target parameter for a goal of losing 3 kg may be weight transition. The correlation index calculated between the time to complete 10 km and another parameter may be different from the correlation index calculated between the weight transition and another parameter. Here, the other parameters for the time to complete 10 km as a target parameter may include weight transition. The other parameters for the weight transition as a target parameter may include the time to complete 10 km. Therefore, when a parameter A is the target parameter, the type and value of the correlation index calculated between another parameter B may be different from the type and value of the correlation index calculated between another parameter A when parameter B is the target parameter. The number of highly correlated indices varies depending on the goal to be achieved, and is assumed to be M here. Next, of the M highly correlated indices in the goal non-achievement data of the fourth user, indices whose improvement rate compared to the goal achievement rate is not greater than a second criterion are identified as related indices (P8). The second criterion may be determined according to the degree of variation of each indices. For example, the range of variation that occurs in daily training even when ability is not improving may be identified in advance and stored in the storage unit 73. The number of such related indicators also varies depending on the goal to be achieved, but here it is assumed to be K. The relationship N≧M≧K is always satisfied.

[0031] The index values ​​are standardized for the K related indexes identified from the similar corresponding data (P9). The numerical values ​​and magnitude of change of each index vary depending on the unit system of the original measured physical quantity and output value. Therefore, in order to equalize the influence of each index, each index is standardized based on its mean value and variance value. In addition, the related indexes for the measurement history up to the point after the target was set are also standardized in the same way (P10). The mean value and variance values ​​used for standardization are the same as those used in process P9.

[0032] A machine learning model 7311 is trained using time-series data of a pair of goal parameters and standardized related indicators as input data, and outputs predicted values ​​of the related indicators after the next activity as estimated values ​​(P11). The input time-series data may be data for the most recent Q0 discrete activities. In this case, a common machine learning model 7311 can be used if the number of times the activity has been performed since the goal was set exceeds Q0. On the other hand, the input time-series data may be all data after the goal was set. In this case, a separate machine learning model 7311 may be trained for each number of times the activity has been performed since the goal was set. For example, deep learning may be used as the machine learning model. Various machine learning models using decision trees may also be used. For training, backpropagation or the like may be used. Note that the values ​​of the related indicators estimated here may be limited to values ​​when the goal is achieved. In this case, the related indicators standardized in process P9 and used to train the machine learning model 7311 may only be goal achievement data.

[0033] Each indicator stored as the target user's practice history is input to the trained machine learning model 7311, and an estimated value of each related indicator after the next activity is obtained (P12). At least the processes of P10 and P12 are performed after the user's activity is performed and before the next activity is started. The processes of P1 to P9 and P11 may be performed once when the user performs the minimum number of activities required for the next prediction after setting the activity, and then stored and retained. Alternatively, the processes of P1 to P9 and P11 may be performed each time the set number of activities is performed, and the stored data may be updated. Alternatively, whether or not to perform the processes P1 to P9 and P11 may be determined depending on the improvement status of the indicator to be improved, which will be described later.

[0034] When the next activity is performed and each indicator is obtained, the obtained indicator values ​​(actual values) are compared with the estimated values, and any actual values ​​that are inferior to the estimated values ​​are identified (P13). If there is an indicator whose actual value is inferior to the estimated value by more than a standard, it is set as an indicator to be improved in the next activity. The set indicator is output to the electronic device 1 or the like and notified to the user immediately or before the start of the next activity (P14). At this time, in addition to simply displaying the indicator, effective training content for improving the indicator may also be displayed. Information indicating the correspondence between the indicator to be improved and the training content may be stored in advance in the storage unit 73.

[0035] In this way, the trained machine learning model 7311 does not need to accurately estimate the index values, but only needs to be able to output values ​​indicating the changes in each index along the way when the user is able to achieve their goal. As a result, if a problem occurs in improving each related index in response to the achievement of the goal, it becomes possible to encourage improvement of the related index, leading to more reliable improvement of the target parameters.

[0036] 5 corresponds to the above-mentioned processes P1 to P12. In the improvement prediction process, the program 731 is read and executed when the activity measurement results are acquired from the electronic device 1 of a registered user and an index of the number of times required for prediction is obtained.

[0037] The control unit 71 sets a goal for the target user (S1; goal acquisition means). The goal may be, for example, information determined based on a user's input operation accepted by the electronic device 1, acquired by the communication unit 74. The control unit 71 acquires physical characteristics and behavioral tendencies as the user's characteristics, and acquires data after the goal for the user has been set (S2; index acquisition means). The control unit 71 acquires values ​​of the physical characteristics other than those related to weight from the registered information 732. The control unit 71 acquires values ​​and index data related to the most recent weight from the history information 733. The control unit 71 acquires information on other activities among the behavioral tendencies from the registered information 732, and acquires activity continuation information, execution frequency, execution time, blank information, etc. from the history information 733.

[0038] The control unit 71 refers to the history information 733 and extracts similar corresponding data of other second users whose physical characteristics, behavioral tendencies, and goals are similar to those of the user within a reference range (S3). The control unit 71 determines goal parameters corresponding to the goals by referring to the storage unit 73, etc. (S4).

[0039] The control unit 71 extracts goal achievement data of a third user who has achieved the goal from the similarly associated data of the second user (S5). The control unit 71 calculates the correlation between each index of the goal achievement data and the goal parameter, and identifies a highly correlated index whose magnitude of correlation is equal to or greater than a first criterion (S6). As described above, the control unit 71 may refer to the storage unit 73 to acquire the type and calculation method of the correlation index according to the goal parameter, as well as the first criterion.

[0040] The control unit 71 extracts goal non-achievement data of a fourth user who did not achieve the goal from the similar corresponding data of the second user (S7). The control unit 71 identifies related indices that have not improved more than the goal parameters from the highly correlated indices included in the goal non-achievement data (S8; identifying means). As described above, the control unit 71 may extract data in which the goal achievement rate is lower than a second standard, rather than simply indicating that the goal was not achieved. The control unit 71 standardizes the related indices of the similar corresponding data and the user data (S9). The control unit 71 trains the machine learning model 7311 using the related indices and the goal parameters in the similar corresponding data (S10). The control unit 71 inputs the user's practice history data into the trained machine learning model 7311 and obtains an estimated value of the related indices after the next activity (S11). Then, the control unit 71 ends the improvement prediction process.

[0041] The improvement support process shown in Fig. 6 starts after the improvement prediction process has been executed, when the measurement data for each activity is acquired after the activity has finished. The control unit 71 acquires the index value of the activity (S21; index acquisition means). The control unit 71 may perform a process to calculate the index value from the measurement value, or may simply acquire an index value that has already been calculated externally. The control unit 71 standardizes the related index (S22). The mean value and variance used for standardization are the values ​​used when standardizing the similar corresponding data.

[0042] The control unit 71 obtains a difference value by subtracting the standardized estimated value from the standardized actual value (S23). The control unit 71 sets the related indicators whose difference values ​​are equal to or greater than a reference value as targets for improvement (S24). The control unit 71 outputs information on the related indicators to be improved to the electronic device 1 (S25; output means). As described above, the control unit 71 may also output information on advice on practice content for improving the related indicators to the electronic device 1. Then, the control unit 71 ends the improvement support process.

[0043] The activity content, i.e., the activity, is not limited to the running-related content described above. The activity is not limited to physical movement, but may also include the study of various subjects. As shown in the system configuration diagram of FIG. 7, an activity support system 100a according to the second embodiment, which aims to support academic ability improvement, may include a server device 7a and an electronic device 1a. The electronic device 1a may be a tablet terminal or a notebook PC. The electronic device 1a may also be a desktop PC. The server device 7a may include an external storage unit 73 as a database.

[0044] As shown in Fig. 8, the server device 7a of this embodiment stores learning information 734 in addition to registration information 732a and history information 733a in the storage unit 73. The learning information 734 includes explanatory texts to be provided to the user, i.e., texts and practice questions. The texts and practice questions are associated with subjects and units, respectively. A plurality of units may be associated. The level of understanding and proficiency of each unit required to understand each text and to correctly answer each practice question may be quantified and stored.

[0045] 9, information such as the user's physique is not required as user registration information 732a, and instead, information such as the school year is stored. Furthermore, the subject of the support target may be stored as activity information.

[0046] The history information 733a stores information on each user's viewing of the text and their answering of the practice questions. The level of understanding and proficiency may be calculated for each unit based on each user's viewing of the text and their correct answers to the practice questions. These levels of understanding and proficiency may be included in the indices. The history information 733a also stores performance information for exams such as mock exams. The performance information may include not only scores for each subject but also scores for each question (section). The performance information may be entered by the user. Alternatively, the operator of the server device 7a may obtain performance information directly from the person conducting the mock exam through a contract or the like. This performance information may also be used to calculate the level of understanding and proficiency. For example, item response theory (IRT) or knowledge training may be used to calculate the level of understanding and proficiency.

[0047] In addition to the direct level of understanding and proficiency in each unit, indirect parameters that affect grades across units may also be set as indicators of the user's academic ability. Examples of indirect parameters include calculation ability, such as the degree of deduction of points due to calculation errors, ability to read and understand questions, such as the relationship between the length of the question or the presence or absence of an explanatory diagram and the percentage of correct answers, and spatial recognition ability, such as the difference in the percentage of correct answers between questions about two-dimensional figures or vectors and questions about three-dimensional or higher-dimensional figures or vectors.

[0048] In this embodiment, the goal (first achievement goal) may be, for example, passing a school entrance exam, a certification exam, or a qualification exam, or a score, ranking, or deviation value on a specific mock exam. Furthermore, if the achievement of the goal involves comprehensive evaluation of multiple subjects, the goal may include whether to focus on a strong subject to earn points, or on a weak subject to achieve a score that can be covered by other strong subjects. The question trends in entrance exams and mock exams, the difficulty level of each subject, the frequency of questions in each unit, the number of test takers, and the academic level of the test takers can be obtained from statistical information on each index (multiple index values) according to the test taker's past test and mock exam scores.

[0049] As shown in FIG. 10, the flow of providing advice to improve a user's indices based on changes in each indices for those with similar goals is basically the same as that shown in FIGS. 4 to 6 of the first embodiment. Specifically, in FIGS. 4 and 5, the goal parameters (P4, S4) set according to the first achievement goal, i.e., the expected passing level (score), change. Accordingly, the range of similarly related data (P2, S3), which is data for similar users, i.e., those who set a similar expected passing level, changes. As a result, the data ranges (P5, S5) for those who achieved their goals and the data ranges (P6, S7) for those who did not achieve their goals for those users also change. Depending on the results of processes P4 to P6 and steps S4, S5, and S7, the units of the identified highly correlated indices (P7, S6) and related indices (P8, S8) may change. That is, the proficiency, comprehension, and indirect parameters (referred to as unit proficiency) of a unit that increase in a goal achiever through learning training are highly correlated indicators, while the unit proficiency that does not increase or increases insufficiently in a goal non-achiever is a related indicator. Changes in unit proficiency, which is a related indicator, may include an increase or decrease in the number of units. Changes in unit proficiency, which is a related indicator, at each timing of the learning period are learned from similar corresponding data corresponding to the first achievement goal (P11, S10). In FIGS. 4 and 6, the difference between the user's most recent changes (actual results) in each related indicator (S21) and the changes (predictions) in the related indicators estimated by each machine learning model 7311 is compared (S23), and units that have not kept up with the increase observed in goal achievers, particularly those with a large difference (deviation), are detected (P13, S25). When the goal is to improve academic ability, the control unit 71 of the server device 7a can not only generate and output advice, but also search for and provide study content corresponding to units where the unit proficiency level is lower than predicted. In Figure 10, steps P35 to P38 related to this content search and provision have been added.

[0050] When a user achieves a goal set by inputting information into the electronic device 1a (P31; corresponding to S1), the control unit 71 of the server device 7a sets conditions corresponding to the user's goal and current situation, and searches the histories of other users who meet the conditions (P32; corresponding to S2 to S4). Regarding the user's situation, physical characteristics are not required as search conditions for improving academic ability. Behavioral tendencies may include study frequency, study time, time zone, and subject assignment. Behavioral tendencies may also include whether or not the user attends a cram school or prep school, whether or not they use online tutoring services, and past study methods and methods. Indirect parameters that are different from parameters that directly indicate the study situation may also include, for example, whether or not the user is using background music, whether or not they use social media while studying, and how they use their mobile device.

[0051] When the control unit 71 acquires the search results from the storage unit 73 (database) (P33; corresponding to S5 and S7), it predicts the academic ability of the user as he / she studies. After that, when the actual results are obtained, the control unit 71 calculates the deviation of the actual results from the prediction (P34; corresponding to S6, S8, S9 to S11, S21 to S23).

[0052] The control unit 71 searches the storage unit 73 for improved content corresponding to parameters with high deviations (P35). The control unit 71 generates and acquires an improved content list (P36) and outputs it to the electronic device 1a as display data together with deviation information (P37). The electronic device 1a displays the input information (P38). When the user selects the displayed improved content, the electronic device 1a requests the improved content from the server device 7a, acquires it, and displays it (P38).

[0053] As shown in Fig. 11(a), the goal setting (S1) may be arbitrarily set by the user through an input operation to the operation reception unit 76. The setting content may be identified through natural language analysis or the like. Alternatively, the goal may be selectable from predetermined options.

[0054] As shown in FIG. 11(b), the display of the comparison results between predictions and actual results (P13) and advice on improving indicators (P14) may be displayed on the same screen. Advice on improving instruction may be obtained, for example, by inserting the name of the corresponding unit or the like into a pre-prepared template sentence. Alternatively, the advice sentence may be obtained by writing a composition incorporating the name of the unit or the like based on AI technology such as a transformer. Furthermore, learning data corresponding to the advice content may be searched from the learning information 734, and selection buttons for viewing the text included in the searched learning data (content) or answering practice questions may also be displayed.

[0055] In entrance exams and mock exams, pass / fail and overall rankings are comprehensive assessments. Therefore, even if a student lacks understanding or proficiency in a specific unit of a particular subject, or even if they abandon their academic improvement in that unit midway, they may still be able to achieve their goal. Furthermore, entrance exams and mock exams do not necessarily cover all units. Units related to indicators with a high percentage of users who have abandoned their academic improvement may not appear in activity goals. Furthermore, overcoming a weak unit and improving grades do not necessarily correspond in real time depending on the timing of overcoming the weak unit and the timing of questions related to that unit. Furthermore, when there are multiple weak units, the order in which they are overcome may vary from user to user. Therefore, the learning and use of machine learning model 7311 may not simply use the overall indicators of the target achiever, but may also use a reference number of times based on the timing when the target achiever focused on studying the unit of interest.

[0056] Alternatively, even if the passing goal is the same, the passing level may be selectable depending on the user's level or inclination. As shown in FIG. 12(a), when setting the passing goal, the user may also be able to select the level at which the user aims to pass. For example, passing at the highest score level, passing near the middle of the passers, and passing near the borderline may be selectable as options that more precisely define the user's first achievement goal.

[0057] When one of the options is selected, the deviation between the unit proficiency required for the selected level and the user's current unit proficiency for each unit may be displayed, as shown in FIG. 12(b). When initially setting the level or considering moving the goal up or down, the deviations for all levels may be displayed together by selecting button B1. In this case, the above processes / steps may be performed sequentially or in parallel for multiple first achievement goals, and units with large deviations, i.e., the difference between predicted and actual results, may be detected. As shown in FIG. 13(a), the all-at-once display may display the deviations between the unit proficiency for each goal level and the current unit proficiency side by side. Depending on the goal level, the deviations for multiple units may be large. In this case, the user may be able to select which of the multiple units to study first. Alternatively, since the deviations increase as the goal becomes higher, the deviations may be displayed overlapping each other in different colors or patterns within the same graph, as shown in FIG. 13(b). If the goal is too high, the number of other users who have achieved the goal with the same level of ability in the past is expected to be small. For example, in cases where there is no ceiling, such as the highest ranking at a top-level school, or when a student is far below the borderline and hopes to pass at a mid-level or higher level, a lack of data may make it difficult to obtain accurate changes in related indicators. In such cases, at the deviation display stage, a note may be displayed to the user indicating that it is difficult to accurately display the deviation or provide support for future activities.

[0058] As described above, the activity support method of this embodiment includes the following steps: (1) Acquire a goal for the activity performed by the target user; (2) Acquire multiple index values ​​representing actual results according to the results of the user's performance of the activity; (3) Identify related indexes that change in conjunction with goal parameters corresponding to the goal from the multiple index values; (4) Output information regarding improvement of the related indexes according to the user's performance of the activity. In this way, this activity support method checks the degree of improvement of related indexes that change in conjunction with and have a high correlation with goal parameters representing the goal, and outputs information encouraging improvement, particularly for related indexes with low degrees of improvement. This allows the user to grasp more specific areas in the implementation of the activity where improvement is lagging in order to achieve the goal. Therefore, the user can perform more effective activities that take into account the areas that need improvement in order to achieve the goal.

[0059] Additionally, the related indicators may be identified based on the relationship between multiple index values ​​and goal parameters when another user with characteristics corresponding to the target user set a second goal similar to the target goal. Because the related indicators are identified based on actual data from when other users in the past attempted to achieve goals under similar conditions to the current target user, more realistic relationships can be obtained depending on the current level and goal level. Therefore, realistic improvement trends in the related indicators are obtained, allowing the target user to understand the relative improvement status of each related indicator.

[0060] The second user may also include a third user who achieved the second achievement goal and a fourth user who did not achieve the second achievement goal. The related indicator may be identified by an indicator among the plurality of indicator values ​​for the third user that has a correlation with the target parameter greater than a first standard, and an indicator among the plurality of indicator values ​​for the fourth user that has an improvement rate that is not higher than the goal achievement rate of the target parameter by a second standard or more. In other words, the related indicator may be an indicator that is highly correlated with the target parameter as much as possible, but does not improve due to other factors. This reduces the influence of other factors, and the target parameter can be more effectively improved by improving the activity content related to the related indicator.

[0061] Furthermore, the characteristics corresponding to the target user may include, as a first condition, that the user's physical characteristics fall within a first reference range. In other words, a user similar to the target user may be a user whose physical characteristics are similar to those of the target user. Differences in physical characteristics, such as age, gender, height, and weight, may result in different problems, or different solutions to the same problem. Therefore, by referencing only the data of other users with similar physical characteristics, it becomes easier to understand the path to goal achievement and deviations from that path when the user has similar conditions to the target user.

[0062] Furthermore, the characteristics corresponding to the target user may include, as a second condition, that the target user's behavioral tendencies fall within a second reference range. In addition to or instead of the physical characteristics described above, behavioral tendencies can also be important parameters for identifying other users similar to the target user. That is, when comparing users with completely different activity execution strategies or with other users who are engaged in other activities outside of the target activity, the basic conditions are significantly different, and therefore the paths to goal achievement are likely to differ. Therefore, by comparing the changes in related indicators of other users with similar behavioral tendencies, the target user's progress toward their goals and problems can be more accurately identified.

[0063] Furthermore, an estimated value of the related indicator that changes as the target user performs an activity may be calculated based on the correlation between the goal parameters and the related indicators of other users. Information regarding improvement may be output based on the difference between the estimated value and the actual value of the related indicator obtained by performing the activity. Therefore, the user can easily determine if the rate of improvement, particularly if the rate of improvement, is deviating from the expected rate during the activity toward achieving the goal. This allows the user to quickly determine if the improvement status of the related indicator is deviating from the standard and adjust the content of the activity. Therefore, this activity support method allows the user to more efficiently obtain and implement the information necessary to achieve their goal.

[0064] Furthermore, the estimated value of the related index may be obtained by a machine learning model 7311 that learns based on the correlation between the target parameter and the related index and outputs an estimated value of the next related index in response to multiple inputs of time-series data of the related index. In other words, a user's ability does not necessarily improve linearly, but often has areas where the user tends to stall or take time to master, and areas where the user improves rapidly once key points are grasped. By using the machine learning model 7311 to more accurately predict the improvement of ability, the user can be prevented from becoming overly impatient or, conversely, complacent, and can be more effectively guided to achieve the user's goal.

[0065] Furthermore, it may be possible to obtain first achievement goals for multiple stages at once. Related indicators may be set for each of the first achievement goals for multiple stages. Mainly in the early stages or at intermediate levels of goals that can be set based on performance, it becomes possible to know the details of the delay in improvement for each of the multiple goals and the extent of the delay. Therefore, the user can proceed with training by more effectively knowing not only the degree of goal achievement but also the progress status during activities including learning.

[0066] Furthermore, activities may include learning behaviors related to improving academic ability. In other words, the contents of the above disclosure may be applicable to brain training applications, rather than physical training. If ability and its improvement can be numerically expressed using multiple parameters, past results by similar users can be obtained using the same algorithm. Therefore, the activity support method of this embodiment can provide the user with effective information for improving their ability.

[0067] Furthermore, by installing the program 731 of this embodiment on a computer, the processing related to the activity support method can be executed in a software manner. Therefore, the program 731 can easily output information that is more suitable for the target user to achieve their goal, without using special hardware, etc.

[0068] The server device 7, which is an activity support device according to this embodiment, also includes a control unit 71. The control unit 71 acquires a goal for an activity performed by the target user. The control unit 71 acquires multiple index values ​​representing performance results according to the results of the activity performed by the target user. The control unit 71 identifies related indexes, which change in conjunction with goal parameters corresponding to the goal, from the multiple index values. The control unit 71 outputs information related to improvement of the related indexes according to the performance of the activity by the target user. The server device 7 checks the degree of improvement of the related indexes, which change in conjunction with and have a high correlation with the goal parameters representing the goal, and can output information encouraging improvement, particularly for related indexes with low degrees of improvement. This allows the user to identify specific areas in the implementation of the activity where improvement is lagging in order to achieve the goal. Therefore, the user can perform more effective activities that take into account areas that need improvement in order to achieve the goal.

[0069] The present invention is not limited to the above-described embodiment, and various modifications are possible. For example, in the above, the next related index is estimated by training the machine learning model 7311, but the next related index may be estimated by other methods. For example, the next related index may be estimated based on trend analysis, multiple regression analysis, or the like.

[0070] In the above example, the related indicators are estimated based on the next activity, but this is not limiting. The average change in the related indicators over multiple activities may also be estimated. This allows the effects of weather conditions such as temperature, humidity, and wind to be averaged out and reduced.

[0071] Furthermore, the identification of related indicators does not require a comparison of the goal achievement rate and improvement rate of each indicator based on the goal non-achievement data. That is, highly correlated indicators that have a high correlation with the target parameter based only on the goal achievement data may be determined as related indicators. Alternatively, the presence or absence of correlation with other parameters may be determined using a method different from the above.

[0072] Although the above describes the case where the improvement of running activity ability and academic achievement is supported, the present invention is not limited to this. The activity to be supported may be other sports, or may be specialized knowledge or special skills whose progress can be evaluated from multiple angles using numerical values, such as board games such as Go and Shogi.

[0073] Furthermore, depending on the activity to be supported as described above, the range of characteristics that are the same as those of the user may be appropriately adjusted depending on the characteristics of the activity.

[0074] In the above description, the server device 7 performs various processes related to the support, but this is not limiting. The processes related to the support may be executed by each of the electronic devices 1. Alternatively, the processes related to the support may be distributed and executed by a plurality of information processing devices.

[0075] In the above description, the storage unit 73 including a nonvolatile memory such as an HDD or flash memory has been used as an example of a computer-readable medium for storing the program 731 and machine learning model 7311 related to the activity support control of the present invention, but the present invention is not limited to this. Other computer-readable media may include other nonvolatile memories such as MRAM, and portable recording media such as CD-ROMs and DVD discs. Furthermore, a carrier wave may also be used as a medium for providing data of the program according to the present invention via a communication line.

[0076] In addition, the specific configurations, contents and procedures of the processing operations, etc. shown in the above embodiments can be modified as appropriate without departing from the spirit of the present invention. The scope of the present invention includes the scope of the invention described in the claims and its equivalents. [Explanation of symbols]

[0077] Server device 7, 7a, control unit 71, program 731

Claims

1. obtaining a first goal related to an activity performed by a first user; acquiring a plurality of index values ​​representing performance according to the results of the activity performed by the first user; Identifying a related index that changes in conjunction with a target parameter corresponding to the first achievement goal from the plurality of index values; outputting information relating to an improvement in the related indicator in response to the first user performing the activity; Activity support methods.

2. The activity support method described in claim 1, wherein the related indicators are identified based on the relationship between the multiple indicator values ​​and the goal parameters when another second user having characteristics corresponding to the first user set a second achievement goal similar to the first achievement goal.

3. The second users include a third user who achieved the second achievement goal and a fourth user who did not achieve the second achievement goal, 3. The activity support method of claim 2, wherein the related index is identified by an index among the plurality of index values ​​for the third user that has a correlation with the target parameter greater than a first criterion, and an index among the plurality of index values ​​for the fourth user that has an improvement rate that is not higher than a second criterion or more than the target achievement rate of the target parameter.

4. The activity support method according to claim 2 , wherein the characteristics corresponding to the first user include, as a first condition, that the physical characteristics of the first user are within a first reference range.

5. The activity support method according to claim 2 , wherein the characteristic corresponding to the first user includes, as a second condition, that the behavioral tendency of the first user is within a second reference range from the behavioral tendency of the first user.

6. 3. The activity support method according to claim 2, further comprising: calculating an estimated value of the related indicator that changes when the first user performs the activity based on a correlation between the target parameter and the related indicator for the second user; and outputting information relating to the improvement based on a difference between the estimated value and an actual value of the related indicator obtained by performing the activity.

7. The activity support method according to claim 6, wherein the estimated value of the related index is obtained using a machine learning model that learns based on the correlation and outputs an estimated value of the related index for the next time in response to input of time series data of the related index multiple times.

8. The first achievement goal can be obtained in a plurality of stages, The related indicators are determined in accordance with the first achievement goals of the plurality of stages, respectively. The activity support method according to claim 2.

9. The activity support method according to claim 1 , wherein the activity includes a learning behavior related to improving academic ability.

10. Computer, goal acquisition means for acquiring a first goal related to an activity performed by a first user; index acquisition means for acquiring a plurality of index values ​​representing achievements according to the results of the activity performed by the first user; an identification means for identifying a related index that changes in conjunction with a target parameter corresponding to the first achievement goal from the plurality of index values; an output means for outputting information relating to an improvement in the related index in response to the first user's performance of the activity; A program that functions as a

11. obtaining a first goal related to an activity performed by a first user; acquiring a plurality of index values ​​representing performance according to the results of the activity performed by the first user; Identifying a related index that changes in conjunction with a target parameter corresponding to the first achievement goal from the plurality of index values; outputting information relating to an improvement in the related indicator in response to the first user performing the activity; An activity support device comprising a control unit.

Images