Estimation device, estimation model generation device, estimation system, estimation method, estimation model generation method and program

Abstract

To estimate characteristics, etc. of an object person by using an objective index.SOLUTION: An estimation system includes a model generation unit, a detection unit and an estimation unit. The model generation unit generates a correlation between a feature value and characteristics as an estimation model by using information on a feature value obtained based on a tapping action and information on characteristics of an operator who has performed the tapping action. The detection unit detects a tapping action of an estimation target person, and obtains a feature value based on the detected tapping action. The estimation unit estimates characteristics or a psychological state corresponding to the feature value detected by the detection unit using the estimation model generated by the model generation unit.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to an estimation device, an estimation model generation device, an estimation system, an estimation method, an estimation model generation method, and a program.

Background Art

[0002] Conventionally, there has been a technique for performing a behavioral characteristic diagnosis test that can diagnose and quantify even unconscious behavioral characteristics of a human (see, for example, Patent Documents 1 and 2).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the techniques of Patent Documents 1 and 2 described above, behavioral characteristic analysis is performed using subjective answers given by a person to be diagnosed to evaluation questions. However, the person to be diagnosed may not always answer honestly, being aware that their answers will be seen by others. Analysis based on such answers may have reduced accuracy.

[0005] In view of the above circumstances, an object of the present invention is to provide an estimation device, an estimation model generation device, an estimation system, an estimation method, an estimation model generation method, and a program that can estimate characteristics of a person to be diagnosed using an objective index.

Means for Solving the Problems

[0006] One aspect of the present invention is an estimation device including an estimation unit that estimates a characteristic or a mental state corresponding to a feature amount obtained based on a tapping operation of a person to be estimated, using an estimation model showing a correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation.

[0007] One aspect of the present invention is the above-described estimation device, wherein the feature amount is the frequency or period of tapping in one or a plurality of measurements, the fluctuation of the frequency or period of tapping in one or a plurality of measurements, the tapping interval difference which is the difference between consecutive tapping intervals in one measurement, the fluctuation of the tapping interval difference in one measurement, the fluctuation of the frequency, period or tapping interval difference of tapping in each of a plurality of measurements, or the fluctuation of the fluctuation of the frequency, period or tapping interval difference of tapping in each of a plurality of measurements.

[0008] One aspect of the present invention is the above-described estimation device, wherein the characteristic is a psychological characteristic, motor ability, sociality, or a life rhythm.

[0009] One aspect of the present invention is an estimation model generation device including a model generation unit that generates an estimation model representing a correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristic of the operator who performed the tapping operation.

[0010] One aspect of the present invention is an estimation system including a detection unit that detects a tapping operation of a person to be estimated and obtains a feature amount based on the detected tapping operation, and an estimation unit that estimates a characteristic or a mental state corresponding to the feature amount detected by the detection unit, using an estimation model showing a correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation.

[0011] One aspect of the present invention is the above-described estimation system, further comprising a model generation unit that generates the estimation model representing the correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristics of the operator who performed the tapping operation.

[0012] One aspect of the present invention is the above-described estimation system, further comprising a processing unit that outputs a stimulus recognizable by vision, hearing, or touch for inducing the movement rhythm of the tapping operation detected by the detection unit to approach a constant state.

[0013] One aspect of the present invention is an estimation method having an estimation step of estimating a characteristic or a mental state corresponding to a feature amount obtained based on a tapping operation of a person to be estimated, using an estimation model showing the correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation.

[0014] One aspect of the present invention is an estimation model generation method having an estimation model generation step of generating an estimation model representing the correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristics of the operator who performed the tapping operation.

[0015] One aspect of the present invention is a program for causing a computer to function as the above-described estimation device.

[0016] One aspect of the present invention is a program for causing a computer to function as the above-described estimation model generation device.

Advantages of the Invention

[0017] According to the present invention, it becomes possible to estimate the characteristics of a target person using an objective index.

Brief Description of the Drawings

[0018]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Mode for Carrying Out the Invention

[0019] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In this embodiment, based on feature amounts such as the center frequency and fluctuation obtained from the movement rhythm of an individual, behavioral characteristics such as cognitive characteristics, sociality, and life rhythm, or daily mental states are objectively and simply specified.

[0020] [Overview] "Motor rhythm," which involves moving the body in sync with rhythms such as music, has been shown in previous research to be important not only for sports and music skills, but also for promoting physical and mental health and enhancing social activities such as communication. There are individual differences in the motor rhythm that each person has. Past research has also shown that when required to press a keyboard at a certain rhythm, the central frequency (average value) and fluctuation (dispersion) vary from person to person. The magnitude of this fluctuation is related to hyperactivity-impulsivity, which is one of the individual characteristics, and is also related to other cognitive characteristics, sociality, daily life rhythm, and mental state.

[0021] Therefore, an experiment as described below was conducted, in which the experimental participants performed tapping movements using the touch panel of a smartphone for one week, and furthermore, a questionnaire survey was conducted to measure psychological state (such as the mood of the day), life rhythm (diet and sleep), and psychological characteristics. As a result, it was revealed that the fluctuation of the motor rhythm obtained by the tapping movement is related to the attentional function, which is a cognitive characteristic related to communication with others, and the executive function. Also, it was suggested that the fluctuation of the daily motor rhythm may be related to the daily life rhythm and psychological state. These are the results suggesting that by simply continuously recording tapping, it is possible to estimate the cognitive characteristics related to a person's communication ability and the daily-varying psychological state, and the present embodiment relates to such measurement and estimation.

[0022] Furthermore, the experiment also suggested the possibility that the psychological state may transform by continuing tapping daily to stabilize the motor rhythm. Therefore, in the present embodiment, a training program is used to train the motor rhythm daily to support the improvement of the psychological state.

[0023] Also, by observing the central frequency and fluctuations of the tapping that an individual has, when the keyboard is pressed alternately with others, it has been shown by previous research that these inherent rhythms attract each other, and this degree of attraction is related to each person's inherent rhythm. Therefore, if the factors causing the inherent movement rhythm can be identified, it is assumed that by evaluating the inherent movement rhythm, it can be applied to the evaluation of personal characteristics and the compatibility with others in communication. Thus, in this embodiment, the compatibility between two or more people is determined and a proposal for personnel placement is made using the inherent rhythm.

[0024] [Comparison with the prior art] Regarding the cognitive characteristics and sociality of humans related to communication, it is necessary to conduct objective evaluations, that is, evaluations that cannot be consciously manipulated by oneself, rather than subjective evaluations such as questionnaires that have been conventionally used, that is, evaluations that can be consciously manipulated by oneself. In addition, the psychological experiments that have been conventionally used to objectively evaluate cognitive characteristics and sociality have a problem in that they consider little the daily fluctuations in order to evaluate the results through a single measurement or the like. Furthermore, although biometric experiments are used as an evaluation method for objectifying cognitive characteristics and sociality, there are problems such as the time required for measurement and evaluation, and these also consider little the daily fluctuations.

[0025] On the one hand, in this embodiment, the movement rhythm such as the central frequency and fluctuation of the tapping performed by the person to be estimated is measured, and based on the measured movement rhythm, cognitive characteristics and sociality are estimated. Further, in this embodiment, the movement rhythm of the person to be estimated is continuously measured, the movement rhythm such as the central frequency and fluctuation is calculated, and based on the feature quantities such as the central frequency and fluctuation of the continuous movement rhythm, the daily mental state and life rhythm are estimated. In this way, in this embodiment, there are advantages compared with the prior art in terms of realizing objective evaluation, using the fluctuation considering daily variations, not taking much time for measurement and evaluation, and being easily measurable by an application such as a smartphone. And this embodiment calculates characteristics such as personal cognitive characteristics and sociality, or mental state, by an algorithm having a scientific basis in neuroscience and psychology, based on the movement rhythm that can be objectively and easily measured.

[0026] In addition, as a similar study on a method of changing the daily mental state so far, there is tapping therapy. This is a therapy aimed at solving the disorder of the negative mental state by tapping a part of the face. On the other hand, in this embodiment, for the purpose of changing the mental state, training to stabilize the movement rhythm of tapping continuously is supported. The effect of the prior art is qualitative, while the effect of this embodiment can be evaluated quantitatively. In addition, brain training and the like are also similar studies that continuously realize the change of mental state and cognitive characteristics. However, the problem is that it takes time to feel the effect, and in contrast, in this embodiment, the effect can be felt simply in a short period.

[0027] [Experimental Results] Here, the experimental results will be explained. In the experiment, more than 200 subjects were instructed to perform a tapping movement of tapping the touch panel of the smartphone 30 times at regular intervals every day for one week. Further, the subjects were given a questionnaire survey to measure the mental state such as the mood of the day, the life rhythm such as diet and sleep, and the psychological characteristics. FIGS. 1 to 5 are diagrams showing a part of the experimental results.

[0028] Figure 1 shows the relationship between the standard deviation (fluctuation) of the average value of the daily tapping cycle for one week and the total score of the ASRS (Adult ADHD Self-Report Scale). Here, for each participant, the average value ATi of the daily tapping cycle for one week (i = 1 to 7) was obtained, and further, the standard deviation of the average values AT1 to AT7 was obtained as the fluctuation. The ASRS is a self-administered symptom checklist for ADHD (Attention-Deficit / Hyperactivity Disorder) in adulthood. The answers to each of the 18 questions included in the ASRS are quantified by values from 0 to 4. In Figure 1, the relationship between the standard deviation of the average value of the daily tapping cycle for one week for each participant and the total score of the ASRS was plotted.

[0029] As shown in Figure 1, it can be seen that the greater the fluctuation (variation) in the average value of the daily tapping cycle for one week, the higher the total score of the ASRS, indicating a higher tendency of inattention and hyperactivity. The straight line L1 is a regression line representing the correlation between the standard deviation of the average value of the daily tapping cycle for one week and the total score of the ASRS. The correlation coefficient R of the straight line L1 is 0.146 and the P-value is 0.021. From Figure 1, it can be seen that based on the fluctuation (standard deviation) of the average value of the daily tapping cycle during a given period, it is possible to predict the tendency of attention-deficit hyperactivity. The straight line L1 can be used as an estimation model for estimating the total score of the ASRS, that is, the tendency of inattention and hyperactivity.

[0030] In Figure 1, the total score of the ASRS is used. However, it was also found that for each of the scores of the motor hyperactivity items and the language hyperactivity items included in the ASRS, there is a similar correlation with the fluctuation (standard deviation) of the average value of the daily tapping cycle for one week. Specifically, there is a positive correlation between the standard deviation of the average value of the daily tapping cycle for one week and the score of the motor hyperactivity item, with the correlation coefficient R of the regression line being 0.154 and the P-value being 0.015. Also, there is a positive correlation between the standard deviation of the average value of the daily tapping cycle for one week and the score of the language hyperactivity item, with the correlation coefficient R of the regression line being 0.122 and the P-value being 0.053.

[0031] Furthermore, it was found that for each of the scores of the inattentive items included in the ASRS and the scores of the "attention control" item in the executive function, there is a correlation with the fluctuation of the daily tapping cycle variance. The fluctuation of the daily tapping cycle variance is obtained by calculating the variance DTi of the daily tapping cycles for each day (i = 1 to 7) in a week and then calculating the standard deviation of the variances DT1 to DT7. Specifically, there is a positive correlation between the fluctuation (standard deviation) of the daily tapping cycle variance and the scores of the inattentive items, with a correlation coefficient R = 0.145 and a P-value = 0.022 for the regression line. There is also a positive correlation between the fluctuation (standard deviation) of the daily tapping cycle variance and the "attention control" in the executive function.

[0032] Figure 2 is a diagram showing the relationship between the standard deviation of the average value of the daily tapping cycles in a week and the tendency of honesty, and Figure 3 is a diagram showing the relationship between the standard deviation of the average value of the daily tapping cycles in a week and the tendency of neurosis. Values representing the strength of tendencies regarding social aspects such as the tendency of honesty and the tendency of neurosis are obtained using a questionnaire that identifies five personality traits called the Ten Item Personality Inventory (TIPI-J). In Figure 2, the relationship between the standard deviation of the average value of the daily tapping cycles in a week for each participant and the score representing the tendency of honesty obtained from the TIPI-J is plotted, and in Figure 3, the relationship between the standard deviation of the average value of the daily tapping cycles in a week for each participant and the score representing the tendency of neurosis obtained from the TIPI-J is plotted.

[0033] As shown in FIGS. 2 and 3, it can be seen that the smaller the variation in the average value of the daily tapping cycle within a predetermined period, the higher the tendency towards honesty, and the larger the daily variation, the higher the tendency towards neurosis. The straight line L2 in FIG. 2 is a regression line representing the correlation between the standard deviation of the average value of the daily tapping cycle in a week and the score representing the tendency towards honesty. The correlation coefficient R of the straight line L2 is 0.109 and the P value is 0.086. The straight line L2 can be used as an estimation model for the tendency towards honesty. The straight line L3 in FIG. 3 is a regression line representing the correlation between the standard deviation of the average value of the daily tapping cycle in a week and the score representing the tendency towards neurosis. The correlation coefficient R of the straight line L3 is 0.121 and the P value is 0.055. The straight line L3 can be used as an estimation model for the tendency towards neurosis.

[0034] FIG. 4 is a diagram showing the relationship between the average cycle of tapping in a day and the total score of negative emotions according to the PANAS (The Positive and Negative Affect Schedule). FIG. 5 is a diagram showing the relationship between the variance of tapping in a day and the total score of positive emotions according to the PANAS. The PANAS is a simple mood rating scale consisting of 8 items related to negative emotions (nervous, frightened, etc.) and 8 items related to positive emotions (lively, proud, etc.), and the answers to each item are quantified by values from 0 to 6.

[0035] In FIG. 4, the relationship between the average cycle of daily tapping of each participant and the total score of the items related to negative emotions of the PANAS is plotted. As shown in FIG. 4, the higher the average cycle of tapping in a day, the higher the tendency towards negative emotions. The straight line L4 in FIG. 4 is a regression line representing the correlation between the average cycle of tapping in a day and the total score of negative emotions of the PANAS. The correlation coefficient R of the straight line L4 is 0.079 and the P value is 0.005. The straight line L4 can be used as an estimation model for the negative psychological state on that day.

[0036] In addition, in Fig. 5, the relationship between the average cycle of daily tapping of each participant and the total score of the items related to positive affect in PANAS was plotted. As shown in Fig. 5, the smaller the variance of the average cycle of daily tapping, the higher the tendency of positive affect. The straight line L5 in Fig. 5 is a regression line representing the correlation between the variance of the average cycle of daily tapping and the total score of positive affect in PANAS. The correlation coefficient R of the straight line L5 is 0.051, and the P value is 0.051. The straight line L5 can be used as an estimation model of the positive mental state on that day.

[0037] As described above, by performing tapping about 30 times a day and taking data for several days to observe the variation in the exercise rhythm, it is possible to estimate characteristics such as the attention function and sociality of that person. Also, by performing tapping about 30 times a day and observing the frequency and variation of the exercise rhythm, it can be seen that the mental state such as the negative or positive mood state on that day can be estimated. Similarly, it is assumed that the motor function and the deviation of the daily rhythm can be estimated.

[0038] [Embodiment of the Present Invention] Subsequently, the estimation system of this embodiment will be described. The estimation system of this embodiment objectively and simply identifies characteristics such as cognitive characteristics, motor ability, sociality, daily rhythm, etc., or mental states such as mood states, based on the central frequency and fluctuation of the exercise rhythm possessed by an individual.

[0039] Fig. 6 is a diagram showing a configuration example of an estimation system 1 according to an embodiment of the present invention. The estimation system 1 includes an estimation model generation device 2, an estimation device 3, and a terminal 4. The estimation device 3 and one or more terminals 4 are connected via a network 5 such as the Internet. Note that the estimation model generation device 2 and the estimation device 3 may be connected via the network 5, or may be connected by a network different from the network 5.

[0040] The estimation model generation device 2 includes a storage unit 21, a learning data acquisition unit 22, a model generation unit 23, and an output unit 24. The estimation model generation device 2 is realized by a computer device such as a server computer or a personal computer, for example.

[0041] The storage unit 21 stores learning data and various data including an estimation model for each characteristic type. The learning data is data used to generate an estimation model. The characteristic type indicates the type of human characteristics or mental states, and includes cognitive characteristics, motor ability, sociality, daily life rhythm, mood state, etc. More specifically, the cognitive characteristics include characteristic types such as attention deficit and hyperactivity, motor hyperactivity, and speech hyperactivity. The sociality includes characteristic types such as honesty and neurosis. The mood state includes characteristic types such as negative emotions and positive emotions, but is not limited thereto. The estimation model is data representing the correlation between the feature amount obtained from the movement rhythm and the value representing the strength of the tendency of the characteristics or mental states of the type indicated by the characteristic type.

[0042] The movement rhythm indicates one or both of the center frequency and the fluctuation in one measurement. The center frequency of the movement rhythm is calculated by the average of the tapping frequencies. Instead of the center frequency, the average of the tapping periods may be used. The fluctuation of the movement rhythm is calculated by the variance or standard deviation of the frequency or period of each tapping. Also, one or both of the average of the tapping interval differences, which is the difference between the k-th (k is an integer of 2 or more) tapping interval and the (k - 1)-th tapping interval in one measurement, and the fluctuation can be used as the movement rhythm. The fluctuation of the tapping interval differences is calculated by the variance or standard deviation of the tapping interval differences obtained in one measurement.

[0043] The feature quantities obtained from the movement rhythm are: (1) the central frequency, average period, or average tapping interval difference in one measurement; (2) the variance or standard deviation of the frequency, period, or tapping interval difference in one measurement; (3) the central frequency or average period of tapping in the entire measurement of N times (N is an integer of 2 or more) or N days; (4) the variance or standard deviation of the central frequency, average period, or average tapping interval difference in each measurement of N times or N days; (5) the variance or standard deviation of the frequency, period, or tapping interval difference of tapping in the entire measurement of N times or N days; (6) the variance or standard deviation of the variance or standard deviation of the frequency, period, or tapping interval difference in each measurement of N times or N days, etc.

[0044] An example of calculating specific feature quantities is shown. Suppose that the measurement is performed once a day for N days, and 30 taps are made in one measurement. Let i be the date (i is an integer from 1 to N), k be the number of taps (k is an integer from 1 to 30), and the frequency of the k-th tap in the measurement on the i-th day be f i,k Let the period of the k-th tap in the measurement on the i-th day be T i,k Let the difference between the k-th tapping interval and the (k - 1)-th tapping interval in the measurement on the i-th day be the tapping interval difference B i,k In the following, the variance is obtained as the fluctuation, but it can be calculated in the same way when using the standard deviation.

[0045] (1) In the measurement on the i-th day (one time), the central frequency of tapping Af i is calculated by the average of the frequencies f i,1 ~f i,30 The average period of tapping AT i is calculated by the average of the periods T i,1 ~T i,30 The average tapping interval difference AB i is calculated by the average of the tapping interval differences B i,2 ~B i,30 .

[0046] (2) In the measurement on the i-th day (one time), the variance of the frequency of tapping Df i is the frequency fi,1 ~f i,30 Calculated by the variance of ~f, the variance DT of the tapping period i is the period T i,1 ~T i,30 Calculated by the variance of ~T, the variance DB of the tapping interval difference i is the tapping interval difference B i,2 ~B i,30 Calculated by the variance of ~B.

[0047] (3) For each of the center frequency and average period of tapping in the entire measurement for m days (m times, where m is an integer from 2 to i) on the i-th day, they are the averages of the frequencies f (i-m+1),1 ~f i,30 and the periods T (i-m+1),1 ~T i,30 from (i - m + 1) to i days, respectively.

[0048] (4) For the variance of the center frequency, variance of the average period, and variance of the average tapping interval difference in each measurement for m days (m times) on the i-th day, they are the variances of the center frequencies Af (i-m+1) ~Af i from (i - m + 1) to i days, the variances of the average periods AT (i-m+1) ~AT i and the variances of the average tapping interval differences AB (i-m+1) ~AB i from (i - m + 1) to i days, respectively.

[0049] (5) For the variance of the frequency, variance of the period, and variance of the tapping interval difference of each tapping in the entire measurement for m days (m times) on the i-th day, they are the variances of the frequencies f (i-m+1),1 ~f i,30 from (i - m + 1) to i days, the variances of the periods T (i-m+1),1 ~T i,30 and the variances of the periods B (i-m+1),2 ~B (i-m+1),30 ,..., B i,2 ~B i,30 from (i - m + 1) to i days, respectively.

[0050] (6) For the variance of the variance of the frequency in each measurement for m days (m times) on the i-th day, it is the variance Df of the frequency (i-m+1) ~Dfi It is calculated by the variance of. The variance of the variance of the average period in each measurement for m days (m times) on the i-th day is the variance DT of the period from (i - m + 1) to the i-th day respectively (i-m+1) ~DT i It is calculated by the variance of. The variance of the variance of the tapping interval difference in each measurement for m days (m times) on the i-th day is the variance DB of the tapping interval difference from (i - m + 1) to the i-th day respectively (i-m+1) ~DB i It is calculated by the variance of.

[0051] The estimation model is a regression model obtained based on learning data. By inputting the feature quantity obtained from the movement rhythm of the tapping movement of the person to be estimated into the regression model represented by the estimation model, estimation result data representing the prediction characteristics or psychological state of the person to be estimated is obtained. The estimation result data is obtained as a numerical value that quantitatively represents the strength of the tendency of the characteristics or psychological state indicated by the characteristic type corresponding to the estimation model. The estimation model is added with characteristic type information indicating the type of the characteristics or psychological state of the person to be estimated and feature quantity type information indicating the type of the feature quantity used for the estimation of the characteristics or psychological state.

[0052] The learning data includes the movement rhythm or feature quantity obtained from the tapping movement in each measurement of one or more times, the measurement time of the movement rhythm, and the characteristic information that quantitatively represents the characteristics or psychological state of the subject numerically. Note that the characteristic information also includes the characteristic type information. As an example, the characteristic information indicates the characteristic type information of the tendency of attention deficit and hyperactivity and the total score of ASRS. Also, as another example, the characteristic information indicates the characteristic type information of honesty and the score of the tendency of honesty based on TIPI-J. Further, as still another example, the characteristic information indicates the characteristic type information of negative emotion and the score of negative emotion based on PANAS.

[0053] The learning data acquisition unit 22 reads out the learning data used for generating the estimation model from the storage unit 21. The learning data acquisition unit 22 may receive the learning data from another computer device connected via a network, or may read out the learning data from a computer-readable recording medium. The model generation unit 23 selects, for each characteristic type, which type of feature amount among the feature amounts obtained from the movement rhythm is most related to the characteristic information, and generates an estimation model using the feature amounts of the selected feature amount type. The output unit 24 outputs the estimation model generated by the model generation unit 23. The output estimation model may be transmitted to the estimation device 3 or another computer device connected to the estimation device 3, or may be written to a recording medium.

[0054] The estimation device 3 obtains an estimation result of the user's characteristics or mental state by inputting the feature amount obtained from the movement rhythm of the user detected by the terminal 4 into the estimation model generated by the estimation model generation device 2. The estimation device 3 includes a storage unit 31, an acquisition unit 32, an estimation unit 33, an output unit 34, and a communication unit 35. The estimation device 3 can be realized by a computer device such as a server computer, a personal computer, or a dedicated device.

[0055] The storage unit 31 stores various data including the estimation model and the measurement application. The measurement application is program data distributed to the terminal 4. The measurement application causes the terminal 4 to detect the movement rhythm from the result of the user's tapping, transmit the detection data including the detected movement rhythm to the estimation device 3, and execute the function of outputting the estimation result data returned from the estimation device 3. Further, the storage unit 31 stores, for each user, movement rhythm history information in which the user identification information, the movement rhythm notified from the user's terminal 4, and the measurement date and time of the movement rhythm are associated.

[0056] The acquisition unit 32 acquires the detection data including the movement rhythm of the user from the terminal 4. The acquisition unit 32 adds the movement rhythm set in the detection data to the movement rhythm history information stored in the storage unit 31.

[0057] The estimation unit 33 obtains the type of feature amount indicated by the feature amount type information set in the estimation model from the exercise rhythm set in the detection data, or from the exercise rhythm set in the detection data and the exercise rhythm stored in the storage unit 31. The estimation unit 33 inputs the obtained feature amount into the estimation model, thereby obtaining estimation result data indicating the estimation characteristics or mental state of the user.

[0058] The output unit 34 outputs the estimation result data obtained by the estimation unit 33 to the terminal 4. The communication unit 35 transmits and receives data with other devices. The communication unit 35 communicates with the terminal 4 via the network 5.

[0059] The terminal 4 is, for example, a smartphone, a tablet terminal, a game terminal, a dedicated terminal, a notebook computer, or the like. For example, a measurement application is downloaded and installed from the estimation device 3 to the terminal 4. Note that the application may be provided by a device other than the estimation device 3. The terminal 4 measures the tapping motion using the application, sets the exercise rhythm detected by the measurement as detection data, and transmits it to the estimation device 3. The terminal 4 includes a storage unit 41, an operation unit 42, a display unit 43, an audio output unit 44, a detection unit 45, a processing unit 46, and a communication unit 47.

[0060] The storage unit 41 stores various data. The storage unit 41 stores the measurement application distributed from the estimation device 3. The storage unit 41 may store data on past exercise rhythms and information on the measurement date and time of those exercise rhythms.

[0061] The operation unit 42 is, for example, a keyboard, a game pad, a touch pad, a mouse, or the like. Note that the operation unit 42 may be a touch panel provided on the display unit 43. In this case, the operation unit 42 may also serve as the detection unit 45.

[0062] The display unit 43 is a display device such as a liquid crystal display device or an organic EL (Electro-Luminescence) display device. The display unit 43 may be an interface for connecting the display device to the terminal 4. In this case, the display unit 43 generates a video signal for displaying data and outputs the video signal to the display device connected to its own terminal.

[0063] The audio output unit 44 is an audio output device such as a speaker, headphones, or earphones. The audio output unit 44 may be an interface for connecting the audio output device to the terminal 4. In this case, the audio output unit 44 generates an audio signal for outputting the audio streaming data as audio and outputs the audio signal to the audio output device connected to its own terminal.

[0064] The detection unit 45 is a sensor that detects the operation by the user's operation unit 42. The detection unit 45 may be, for example, a microphone that detects the sound when a keyboard or the like is pressed. The detection unit 45 measures the motion rhythm of the tapping motion performed by the user using the operation unit 42. Conventional techniques can be applied to measure the motion rhythm of the tapping motion.

[0065] The processing unit 46 executes the measurement application installed in the terminal 4. By executing the measurement application, the processing unit 46 performs processes such as displaying various data on the display unit 43, outputting various data as audio from the audio output unit 44, and detecting the motion rhythm of the tapping performed by the user according to the measurement application using the detection unit 45. Further, the processing unit 46 transmits detection data including the detection result of the motion rhythm, the user identification information for identifying the user, and the measurement date and time of the motion rhythm to the estimation device 3 via the communication unit 47. The communication unit 47 transmits and receives data to and from other devices via the network 5.

[0066] Note that the estimation model generation device 2 may be realized by a plurality of computer devices connected to a network. In this case, it can be arbitrary which of these plurality of computer devices realizes each functional unit included in the estimation model generation device 2. For example, the storage unit 21, or the storage unit 21 and the learning data acquisition unit 22, and other functional units may be realized by different computers connected by a network. Also, the same functional unit of the estimation model generation device 2 may be realized by a plurality of computer devices.

[0067] Similarly, the estimation device 3 may be realized by a plurality of computer devices connected to a network. In this case, it can be arbitrary which of these plurality of computer devices realizes each functional unit included in the estimation device 3. For example, the storage unit 31 of the estimation device 3 and other functional units of the estimation device 3 may be realized by different computers connected by a network.

[0068] Also, the estimation device 3 and the terminal 4 may be an integrated device. Also, the estimation model generation device 2 and the estimation device 3 may be an integrated device. In this case too, as described above, the functions of the estimation device 3 may be realized by a plurality of computer devices connected to a network. Also, the estimation model generation device 2, the estimation device 3, and the terminal 4 may be an integrated device. In this case too, as described above, the functions of the estimation model generation device 2 and the functions of the estimation device 3 may be realized by a plurality of computer devices connected to a network.

[0069] FIG. 7 is a diagram for explaining an example of a tapping operation using the terminal 4. The measurement application instructs the user, for example, to press an operation unit 42 such as a keyboard or a button at a constant time interval at a preferred pace, or to touch the operation unit 42 such as a touch panel. The instructed user presses the keyboard or the button as shown in FIG. 7(a) or taps the touch panel as shown in FIG. 7(b) at a preferred pace and at a constant time interval. The detection unit 45 of the terminal 4 detects that the keyboard or the button has been pressed, or that the touch panel has been tapped.

[0070] FIG. 8 is a diagram showing an example of the detection result of a tapping motion. In FIG. 8, the horizontal axis represents the tapping interval (milliseconds), and the vertical axis represents the number of taps (times). The detection unit 45 of the terminal 4 obtains the tapping frequency based on the reciprocal of each tapping interval (tapping period), and sets the average of the obtained tapping frequencies as the average frequency of the user's motion rhythm. Note that the average of the tapping intervals may be used as the motion rhythm. Further, the detection unit 45 calculates the standard deviation or variance in the distribution of the tapping interval or tapping frequency and the number of taps, and sets the calculated standard deviation or variance as the fluctuation of the user's motion rhythm. The detection unit 45 calculates the tapping interval difference, which is the difference between the k-th tapping interval and the (k - 1)-th tapping interval, and the average of the tapping intervals may be used as the motion rhythm. Further, the detection unit 45 may use the standard deviation or variance of the tapping interval difference as the fluctuation of the motion rhythm.

[0071] Subsequently, the operation of the estimation system 1 will be described. FIG. 9 is a flowchart showing the processing of the estimation model generation device 2. The estimation model generation device 2 performs the processing of FIG. 9 for each characteristic type information.

[0072] First, the learning data acquisition unit 22 of the estimation model generation device 2 reads a plurality of learning data in which the characteristic type information of the estimation model generation target is set from the storage unit 21 (step S11). The characteristic type information of the estimation model generation target is set in advance by the user of the estimation model generation device 2.

[0073] The model generation unit 23 selects, as a processing target, one unselected candidate feature quantity type to be used in the estimation model (step S12). The candidate feature quantity types are, for example, the center frequency of tapping, the average period, or the average tapping interval difference in one measurement (one day's measurement), the variance or standard deviation of the frequency, period, or tapping interval difference of tapping in one measurement (one day's measurement), the center frequency or average period of tapping in the entire measurement for a predetermined number of times or a predetermined number of days, the variance or standard deviation of the center frequency, average period, or average tapping interval difference of tapping in the entire measurement for a predetermined number of times or a predetermined number of days, the variance or standard deviation of the variance or standard deviation of the frequency, period, or tapping interval difference in each measurement for a predetermined number of times or a predetermined number of days, and so on. The selected feature quantity type to be processed is referred to as the processing target feature quantity type. Note that there is one or more candidate feature quantity types to be used in the motion model, which are set in advance by the user of the estimation model generation device 2.

[0074] Based on the motion model of each learning data read in step S11, the model generation unit 23 acquires the feature quantity indicated by the processing target feature quantity type. The model generation unit 23 generates a regression model showing the relationship between the feature quantity indicated by the processing target feature quantity type and the characteristic information using the set of the feature quantity and the characteristic information obtained from each learning data, and further calculates the correlation coefficient R and the P value (step S13). The correlation coefficient R represents the strength of the correlation between the feature quantity and the characteristic information, and whether there is a positive correlation or a negative correlation, and the P value represents statistical superiority.

[0075] The model generation unit 23 determines whether all candidate feature quantity types to be used in the estimation model have been selected (step S14). When the model generation unit 23 determines that there is an unselected feature quantity type (step S14: NO), it repeats the processing from step S12.

[0076] When the model generation unit 23 determines that all candidate feature type have been selected (step S14: YES), among the regression models calculated for each feature type, the regression model with a P-value of a predetermined value or less and the highest correlation coefficient R is selected (step S15). When there is only one candidate feature type, the model generation unit 23 selects the regression model generated in step S13.

[0077] The model generation unit 23 adds the feature type used in the selected regression model and the characteristic type information of the estimation target to the selected regression model to generate an estimation model (step S16). The model generation unit 23 stores the generated estimation model in the storage unit 21. The output unit 24 outputs the estimation model stored in the storage unit 21 (step S17). The estimation device 3 acquires the estimation model output from the estimation model generation device 2 and stores it in the storage unit 31.

[0078] FIG. 10 is a flowchart showing an example of the processing of the estimation device 3 and the terminal 4. The terminal 4 acquires a measurement application from the estimation device 3 in advance and stores it in the storage unit 31. When the user inputs the execution of the measurement application by operating the operation unit 42, the processing unit 46 reads out the measurement application from the storage unit 31 and starts the processing. The measurement application executed by the processing unit 46 instructs the user to tap at a constant time interval at a preferred pace, such as by displaying a message on the display unit 43 and outputting a sound from the voice output unit 44 (step S21).

[0079] The user taps about 30 times by pressing the keys or buttons of the keyboard of the terminal 4, tapping with a finger on a game pad or touch pad, or the like. The detection unit 45 detects that the tapping operation has been performed (step S22). When the detection unit 45 calculates the motion rhythm based on the detection result of the tapping (step S23), the detection data set with the calculated motion rhythm, the user identification information, and the measurement date and time is transmitted from the communication unit 47 to the estimation device 3 (step S24).

[0080] The acquisition unit 32 of the estimation device 3 receives the detection data transmitted by the terminal 4 via the communication unit 35 (step S31). The acquisition unit 32 identifies the exercise rhythm history information stored in the storage unit 31 based on the user identification information set in the detection data. The acquisition unit 32 adds and updates the pair of the exercise rhythm and the measurement date and time read from the detection data to the identified exercise rhythm history information (step S32).

[0081] The estimation unit 33 reads out the estimation model used for estimation from the storage unit 31 (step S33). The information of the estimation model used for estimation may be set in the storage unit 31 in advance corresponding to the user identification information. Alternatively, the estimation unit 33 may receive the characteristic type information indicating the type of the characteristic or mental state to be estimated from the terminal 4 together with the detection data, and read out the estimation model set with the received characteristic type information from the storage unit 31 and use it as the estimation model for estimation.

[0082] The estimation unit 33 reads out the feature quantity of the feature quantity type set in the read estimation model from the detection data received in step S31 or from the detection data history information updated in step S32 (step S34). For example, when the feature quantity type is the center frequency or fluctuation (variance or standard deviation) for one time, the estimation unit 33 reads out the feature quantity of the feature quantity type from the exercise rhythm of the received detection data. When the feature quantity type is the standard deviation of the center frequency for each day of one week, the estimation unit 33 reads out the exercise rhythm for each day within the period one week back from the current measurement time from the detection data history information, and calculates the variance of the center frequency set in the read exercise rhythm. When the storage unit 41 of the terminal 4 stores the past exercise rhythm, the terminal 4 may include the exercise rhythm of the past predetermined period in the detection data and transmit it to the estimation device 3.

[0083] The estimation unit 33 obtains, as estimation result data, a value representing an estimated characteristic or mental state by inputting the feature amount acquired in step S34 into the estimation model read out in step S33 (step S35). The output unit 34 outputs the estimation result data obtained by the estimation unit 33 to the terminal 4 via the communication unit 35 (step S36).

[0084] The processing unit 46 of the terminal 4 receives the estimation result data from the estimation device 3. The processing unit 46 displays the received estimation result data on the display unit 43 or outputs it as voice from the voice output unit 44 (step S25). Note that the processing unit 46 may output an explanation of the characteristic corresponding to the numerical value indicated by the estimation result data. In this case, the output unit 34 may add the explanation to the estimation result data and transmit it to the terminal 4.

[0085] Note that the estimation model used by the estimation device 3 to estimate the characteristics or mental state of the user of the terminal 4 may be an estimation model generated based on learning data obtained for a subject different from the user, or an estimation model generated based on learning data obtained in the past for the user. Further, the estimation model generated based on the learning data obtained for a subject different from the user may be an estimation model updated based on the learning data obtained for the user of the terminal 4.

[0086] When updating the estimation model based on the learning data obtained for the user of the terminal 4, the user inputs, via the operation unit 42 of the terminal 4, the correct characteristics or mental state corresponding to the detection data. The terminal 4 transmits the detection data and the characteristic information indicating the correct characteristics or mental state that has been input to the estimation device 3. The estimation device 3 transmits the learning data in which the correct characteristic information received from the terminal 4 and the movement rhythm and the measurement time of the movement rhythm read from the detection data history information stored in the storage unit 31 are set, and the model identification information for identifying the estimation model used for the estimation to the estimation model generation device 2. The storage unit 21 of the estimation model generation device 2 stores the learning data received from the estimation device 3. The estimation model generation device 2 performs the processing of FIG. 9 using the learning data based on the movement rhythm detected at the terminal 4 of each user, and updates the estimation model specified by the model identification information. Alternatively, the model generation unit 23 of the estimation model generation device 2 may update the estimation model specified by the model identification information using the learning data received from the estimation device 3. Alternatively, the estimation device 3 may be provided with a model update unit (not shown), and the model update unit of the estimation device 3 may update the estimation model used for the estimation using the detection data and the correct characteristic information received from the terminal 4. Thereby, while providing the user with the estimation result of the characteristics and mental state based on the tapping operation, the accuracy of the estimation by the estimation model can be improved.

[0087] By the above-described processing, it becomes possible to estimate the daily state and the abilities required for various communications from the movement rhythm obtained by several taps. For example, it becomes possible to estimate the cognitive state and mental state that vary daily with just several taps.

[0088] Also, for example, by using the characteristic information estimated by the estimation model as the learning effect, it is possible to visualize the learning effect in the educational field and in the personnel training of employees with just several taps.

[0089] Note that the estimation unit 33 can further propose personnel allocation in team formation in the workplace, educational institutions, sports, etc. using the estimation results. For example, the storage unit 31 stores in advance the characteristics information required for or suitable for each role and position in the team. The estimation unit 33 replaces or adds to the estimation result data the information on the role and position corresponding to the characteristics information obtained as the estimation result, and transmits it to the terminal 4 for display.

[0090] In addition, the estimation unit 33 can also determine the relationship between the two using the estimation results of the characteristics information of each of the two. The relationship between the two refers to the degree of connection of the interpersonal relationship between the two. For example, in the case of lovers, spouses, friends, work colleagues, etc., whether the compatibility between the two is good or bad, whether they get along well or not, whether the work tempo matches or not, etc. In this case, the estimation unit 33 performs the processes of steps S21 to S24 and steps S31 to S35 in FIG. 10 for the two users. The estimation unit 33 determines that the closer the feature amounts of the two are, the better the compatibility. Alternatively, the storage unit 31 stores in association with each other the set of characteristics information and the degree of connection of the interpersonal relationship, such as whether the compatibility is good or bad, whether they get along well or not, and whether the work tempo matches or not. For example, the storage unit 31 stores that the characteristics information representing leadership temperament and the specific information representing follower temperament are compatible, and that the characteristics information representing leadership temperament is incompatible with each other. The estimation unit 33 reads out from the storage unit 31 the information on the degree of connection of the interpersonal relationship corresponding to the set consisting of the characteristics information obtained for each of the two, and uses it as the estimation result data. The estimation unit 33 transmits the estimation result data to the terminal 4 for display. As a result, it becomes possible to match friends and romantic partners with just a few taps. Note that in the case of three or more people, similar estimation can be performed by storing in the storage unit 31 the set of characteristics information and the degree of connection of the interpersonal relationship (relationship).

[0091] Also, the smaller the fluctuation and dispersion of the tapping period, the more stable the mental state and life rhythm are. Moreover, experiments have shown that by continuing to tap, the mental state and life rhythm become more stable. Therefore, the terminal 4 prompts the user to stabilize the exercise rhythm.

[0092] Specifically, the storage unit 41 of the terminal 4 stores the exercise rhythm detected up to a period retroactively from the present for a predetermined period. The processing unit 46 of the terminal 4 activates the measurement application and gives instructions such as displaying a message on the display unit 43 and outputting a voice from the voice output unit 44 so that the user performs tapping for training. When the user performs tapping, the processing unit 46 calculates the exercise rhythm by the detection unit 45. When the fluctuation of the calculated exercise rhythm is larger than a predetermined value, or when there is a deviation of a predetermined value or more from the center frequency detected in the past or when the past mental state was good, etc., the processing unit 46 outputs a sound from the voice output unit 44 at a constant rhythm, vibrates the terminal 4, switches the display on the display unit 43, etc., and outputs stimuli through hearing, touch, vision, etc. to assist in stabilizing the exercise rhythm. For example, the processing unit 46 outputs stimuli at a constant period according to the calculated center frequency and the center frequency detected in the past or when the past mental state was good.

[0093] In this way, by stabilizing the exercise rhythm, it is possible to provide an application of a training program for stabilizing the mental state and life rhythm.

[0094] According to the above-described embodiment, by estimating characteristic information and training to stabilize the exercise rhythm, it is possible to realize a program for self-monitoring and self-care using tapping. Also, it is possible to provide a cognitive training program using tapping.

[0095] The above-described estimation system 1 can be used as follows. (1) Measure the exercise rhythm and calculate the center frequency and fluctuation. The user performs tapping about several tens of times using a touch panel of a smartphone or the like, or a keyboard of a computer device or the like. The movement rhythm such as the central frequency (average) and fluctuation (dispersion) of the user individual is calculated.

[0096] (2) Calculate cognitive characteristics, motor ability, and sociality from the movement rhythm. The estimation device 3 uses the feature amount obtained from the movement rhythm calculated in (1) above and the learning data in which the values of characteristic information such as the execution function (working memory), attention function, motor function, and sociality related to the user's communication when the movement rhythm was obtained are set, and constructs a regression model for predicting the values of each characteristic information by the process shown in FIG. 9 as the algorithm of the prediction model. After the prediction model is generated, the estimation device 3 uses the movement rhythm of the user obtained as in (1) above using the terminal 4, and estimates the value of the characteristic information by the prediction model.

[0097] (3) Continuously measure the movement rhythm and calculate the central frequency and fluctuation. The user continuously performs the measurement of the movement rhythm measured as in (1) above for several days (for example, one week), and calculates the central frequency (average) and fluctuation (dispersion) of the movement rhythm that varies within the user individual.

[0098] (4) Calculate daily life rhythm and mood state from continuous movement rhythm data. The estimation device 3 uses the feature amount such as the fluctuation of the central frequency (average) and fluctuation (dispersion) of the predetermined period (for example, one week) calculated in (3) above and the learning data in which the values of characteristic information such as the mental state and the deviation of the life rhythm related to the user's communication when the feature amount was obtained are set, and constructs a regression model for predicting the values of each characteristic information by the process shown in FIG. 9 as the algorithm. After the prediction model is generated, the estimation device 3 uses the movement rhythm of the user obtained by (3) above, and estimates the value of the characteristic information by the prediction model.

[0099] (5) Construct a large-scale database of movement rhythms. Using the above measurement techniques (1) to (4) and analysis techniques, obtain movement rhythms, and construct a large-scale database in combination with an individual's psychological characteristics, motor ability, sociality, daily life rhythm, and mood state. For example, taking the actual measurement data widely collected from a large number of people as learning data, according to (2) and (4), for the attention function, y1 (value of the attention function) = a1·x1 (fluctuation of the daily tapping variance) + b1; for the motor function, y2 (value of the motor function) = a2·x2 (fluctuation of the daily tapping cycle) + b2; for the deviation of the life rhythm, y3 = a3·x3 (fluctuation of the tapping cycle / variance over several days) + b3;... Suppose regression lines are obtained. The large-scale database stores all these regression models as estimation models, and when new learning data based on new measurements is obtained, it uses them to update the regression lines of the estimation models.

[0100] (6) Quantify communication ability from movement rhythms. Using the large-scale database in (5) above, establish an algorithm for quantifying communication ability. For example, set the algorithm for quantifying communication ability as αy1 + βy2 + γy3 +... obtained by weighted addition of the values y1, y2, y3,... obtained by each estimation model, and determine the coefficients α, β, γ,... representing the weights. Furthermore, based on the synchronization rate of this movement rhythm and the combination of characteristic information, judge the compatibility and relationship between the two.

[0101] (7) Develop an application that proposes communication types from movement rhythms. Applicationize the above (1) to (6) and implement them on a computer device, smartphone, etc.

[0102] (8) Develop an application for training movement rhythms. Implement a function in the application installed on the terminal 4 to notify the deviation by message or voice when the user's tapping is faster / slower than a predetermined amount compared to a stable movement rhythm. The application induces a certain movement rhythm by vibrating the device, changing the display, etc. according to the stable movement rhythm. This application supports the user's training to perform the task of stabilizing the movement rhythm every day.

[0103] Describe the hardware configuration examples of the estimation model generation device 2 and the estimation device 3. FIG. 11 is a device configuration diagram showing the hardware configuration examples of the estimation model generation device 2 and the estimation device 3. The estimation model generation device 2 and the estimation device 3 each include a processor 71, a storage unit 72, a communication interface 73, and a user interface 74.

[0104] The processor 71 is a central processing unit that performs operations and controls. The processor 71 is, for example, a CPU. The processor 71 reads and executes a program from the storage unit 72. The storage unit 72 further has a work area, etc. when the processor 71 executes various programs. The communication interface 73 is connected to be communicable with other devices. The user interface 74 is an input device such as a keyboard, a pointing device (mouse, tablet, etc.), a button, a touch panel, etc., a display device such as a display, and an audio output device such as a speaker or headphones.

[0105] At least a part of the functions of the learning data acquisition unit 22, the model generation unit 23, and the model generation unit 23 of the estimation model generation device 2, and the functions of the acquisition unit 32, the estimation unit 33, and the output unit 34 of the estimation device 3 are realized by the processor 71 reading and executing a program from the storage unit 72. The above program may be transmitted via a telecommunication line or may be recorded on a portable recording medium. Also, the above program may be for realizing a part of the functions described above, and furthermore, it may be possible to realize the functions described above in combination with a program already recorded in the computer system. Note that all or part of the above functions may be realized using hardware such as an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). Also, the storage unit 21 of the estimation model generation device 2 and the storage unit 31 of the estimation device 3 are realized by the storage unit 72.

[0106] Also, the hardware configuration of the terminal 4 is the same as that in FIG. 11. At least a part of the functions of the detection unit 45 and the processing unit 46 of the terminal 4 are realized by the processor 71 reading and executing a program from the storage unit 72.

[0107] This embodiment has the following effects.

[0108] (Objective, quantitative, unconscious estimation of an individual's ability and state) There has been no research to date on developing a system that identifies an individual's cognitive characteristics, sociality, and psychological state from movement rhythm synchronization. This embodiment is innovative. Most conventional psychological questionnaires, etc., are based on subjective profiling and questionnaires, making it difficult to obtain objective evaluations. As a result, there has been a problem of lack of reliability as individuals can consciously manipulate their answers so that they achieve such results for cognitive characteristics and psychological states. The movement rhythm in this embodiment is difficult for an individual to consciously manipulate the output, and has the merit of enabling visualization of its ability in that it is quantitative, unlike qualitative conventional methods.

[0109] (Estimation of an individual's daily-varying abilities and states) Even when an individual can honestly answer subjective profiling and questionnaires, since an individual's state varies daily, an index that reflects each change is essential. This embodiment has the merit of enabling the estimation of cognitive characteristics, sociality, and psychological state in consideration of daily variations in that it uses the fluctuations of the movement rhythm.

[0110] (Simple estimation of an individual's abilities and states) As similar research, there are many developments and studies on estimating cognitive characteristics, sociality, and psychological state from ecological information such as brain waves and heartbeats. These have merits similar to this embodiment in that they are difficult to consciously manipulate. On the other hand, many of the devices themselves are large-scale, and even if they are small, there is a problem that they are not easy for everyone to use. In this embodiment, with just one smartphone, by downloading an application and performing several taps, there is the merit of being simple in that cognitive characteristics, sociality, and psychological state can be estimated.

[0111] (Estimation of an individual's abilities and states with scientific basis) There is no example of a study that analyzes and databases motion rhythm data through continuous measurement of the same participants on the scale of hundreds of people, and this embodiment is innovative. Most of the previous studies had difficulty in constructing a unified database because of the small amount of sample data in individual studies and the fact that the experimental conditions were too different to conduct a meta-analysis of multiple research results. This embodiment is scientifically novel and highly innovative in that it can establish a motion rhythm database based on the motion rhythm measurement technology and analysis technology established through past experience and achievements.

[0112] (Technology for training to continuously stabilize the motion rhythm for the purpose of transforming the mental state) Similar studies on methods for transforming the daily mental state up to now include tapping therapy. The effect of this method is qualitative, whereas the effect of this embodiment can be quantitatively evaluated. Also, brain training and the like are similar studies that continuously realize the transformation of the mental state and cognitive characteristics, but it takes time to feel the effect. In contrast, this embodiment enables the effect to be felt simply in a short time.

[0113] According to the embodiment described above, the estimation system includes a detection unit and an estimation unit. The detection unit may be provided in the terminal, and the estimation unit may be provided in the estimation device. The detection unit detects the tapping operation of the person to be estimated and obtains a feature amount based on the detected tapping operation. The estimation unit uses an estimation model showing the correlation between the feature amount obtained based on the tapping operation and the characteristics or mental state of the operator who performed the tapping operation to estimate the characteristics corresponding to the feature amount detected by the detection unit.

[0114] The feature quantity is the tapping frequency or period in one or multiple measurements, the fluctuation of the tapping frequency or period in one or multiple measurements, the tapping interval difference which is the difference between consecutive tapping intervals in one measurement, the fluctuation of the tapping interval difference in one measurement, the fluctuation of the tapping frequency, period or tapping interval difference in each of multiple measurements, or the fluctuation of the fluctuation of the tapping frequency, period or tapping interval difference in each of multiple measurements. For example, the fluctuation is the variance or standard deviation. Also, the characteristic is a psychological characteristic, motor ability, sociality, or life rhythm.

[0115] The estimation system may further include a model generation unit. The model generation unit may be provided in the estimation model generation device or in the estimation device. The model generation unit generates an estimation model representing the correlation between the feature quantity and the characteristic using the information on the feature quantity obtained based on the tapping operation and the information on the characteristic of the operator who performed the tapping operation.

[0116] The estimation system may further include a processing unit. The processing unit outputs a stimulus recognizable by vision, hearing or touch for guiding the movement rhythm of the tapping operation detected by the detection unit to approach a certain state.

[0117] As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and designs within the scope not departing from the gist of the present invention are also included.

Explanation of Reference Numerals

[0118] 1 Estimation system 2 Estimation model generation device 3 Estimation device 4 Terminal 5 Network 21 Storage unit 22 Learning data acquisition unit 23 Model generation unit 24 Output unit 31 Storage unit 32 Acquisition unit 33 Estimation Unit 34 Output Unit 35 Communication Unit 41 Memory Unit 42 Operation Unit 43 Display Unit 44 Voice Output Unit 45 Detection Unit 46 Processing Unit 47 Communication Unit 71 Processor 72 Memory Unit 73 Communication Interface 74 User Interface

Claims

1. An estimation unit that estimates a characteristic or mental state corresponding to a feature amount obtained based on a tapping operation of a person to be estimated, using an estimation model showing a correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation, An estimation device comprising the same.

2. The feature amount is the frequency or period of tapping in one or more measurements, the fluctuation of the frequency or period of tapping in one or more measurements, the tapping interval difference that is the difference between consecutive tapping intervals in one measurement, the fluctuation of the tapping interval difference in one measurement, the fluctuation of the frequency, period, or tapping interval difference of tapping in each of a plurality of measurements, or the fluctuation of the fluctuation of the frequency, period, or tapping interval difference of tapping in each of a plurality of measurements. The estimation device according to Claim 1.

3. The characteristic is a psychological characteristic, motor ability, sociality, or lifestyle rhythm. The estimation device according to Claim 1.

4. A model generation unit that generates an estimation model representing the correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristic of the operator who performed the tapping operation. An estimation model generation device comprising the same.

5. A detection unit that detects a tapping operation of a person to be estimated and obtains a feature amount based on the detected tapping operation, and An estimation unit that estimates a characteristic or mental state corresponding to the feature amount detected by the detection unit, using an estimation model showing a correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation. An estimation system comprising the same.

6. Further comprising a model generation unit that generates the estimation model representing the correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristic of the operator who performed the tapping operation. The estimation system according to Claim 5.

7. Further comprising a processing unit that outputs a stimulus recognizable by vision, hearing, or touch for guiding the movement rhythm of the tapping operation detected by the detection unit to approach a certain state. The estimation system according to Claim 5 or 6.

8. An estimation step of estimating a characteristic or mental state corresponding to a feature amount obtained based on a tapping operation of a person to be estimated, using an estimation model showing a correlation between the feature amount obtained based on the tapping operation and the characteristic or mental state of the operator who performed the tapping operation. An estimation method having the above. **Claim 9** An estimation model generation step of generating an estimation model representing a correlation between the feature amount and the characteristic, using information on the feature amount obtained based on the tapping operation and information on the characteristic of the operator who performed the tapping operation. An estimation model generation method having the above. **Claim 10** A program for causing a computer to function as the estimation device according to claim 1. **Claim 11** A program for causing a computer to function as the estimation model generation device according to claim 4.

Images