Information processing apparatus, feature quantity extraction method, teacher data generation method, estimation model generation method, stress degree estimation method, and feature quantity extraction program

Abstract

In order to appropriately extract feature quantities used for machine learning or stress level estimation, these information processing devices (1, 4) comprise: specification means (11, 404) that specify, as focal time periods, time periods for which a chronic stress tendency is clearly indicated in biological signals obtained from a subject across a prescribed time period; and extraction means (12, 405) that extract, from biological signals obtained in a specified focal time period, at least one feature quantity used in machine learning for a stress level estimation model or used in estimating stress levels using an estimation model.

Description

Technical Field

[0001] The present invention relates to a technique for extracting features used for machine learning of a stress level estimation model or estimating a stress level using the estimation model.

Background Art

[0002] In recent years, cases in which employees develop mental disorders such as depression due to occupational stress and leave their jobs or take time off have been increasing. Along with this, the increased burden on companies to maintain and secure employees has also become a problem. Against this background, research on stress monitoring has been underway. For example, research has also been underway on a technique for generating a stress level estimation model using measurement data such as body movement data and biological data of a subject, and estimating the stress level of the subject using the generated estimation model.

[0003] For example, Patent Document 1 describes a stress estimation device and a stress estimation method using a biological signal.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In a stress estimation device, it is required to further improve the accuracy of stress estimation. Appropriately setting the machine learning and the features used for estimating the stress level leads to an improvement in the accuracy of stress estimation.

[0006] One aspect of the present invention has been made in view of the above-described problems, and an example of the object is to provide a technique for extracting appropriate feature amounts used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model.

Means for Solving the Problems

[0007] An information processing apparatus according to one aspect of the present invention includes a specifying means for specifying, as a target time zone, a time zone in which a chronic stress tendency is significantly represented in a biological signal acquired from a subject over a predetermined period, and one or more feature amounts used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model are extracted from the biological signal acquired in the specified target time zone.

[0008] A feature amount extraction method according to one aspect of the present invention includes at least one processor specifying, as a target time zone, a time zone in which a chronic stress tendency is significantly represented in a biological signal acquired from a subject over a predetermined period, and extracting one or more feature amounts used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model from the biological signal acquired in the specified target time zone.

[0009] A feature amount extraction program according to one aspect of the present invention causes a computer to execute a specifying process for specifying, as a target time zone, a time zone in which a chronic stress tendency is significantly represented in a biological signal acquired from a subject over a predetermined period, and an extraction process for extracting one or more feature amounts used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model from the biological signal acquired in the specified target time zone.

Effects of the Invention

[0010] According to one aspect of the present invention, it is possible to extract appropriate feature amounts used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Mode for Carrying Out the Invention

[0012] 〔Exemplary Embodiment 1〕 The inventors of the present invention conceived of improving the accuracy of stress estimation by selectively narrowing down the biological signals used for creating an estimation model or stress estimation among the biological signals acquired from a subject over a predetermined period, and thus completed the present invention. Specifically, the inventors of the present invention conceived of paying attention to the biological signals in the time zone when the chronic stress tendency is significantly manifested in the biological signals due to the subject being in a specific state, and thus completed the present invention. Hereinafter, some exemplary embodiments of the present invention will be described in detail with reference to the drawings.

[0013] First, a first exemplary embodiment of the present invention will be described. Exemplary embodiment 1 is a form that forms the basis of each of the exemplary embodiments described below.

[0014] <Configuration of the information processing apparatus> FIG. 1 is a block diagram showing the configuration of the information processing apparatus 1. As shown in the figure, the information processing apparatus 1 has a configuration including a specifying unit 11 and an extracting unit 12. In this exemplary embodiment, the specifying unit 11 is a configuration that realizes specifying means. In this exemplary embodiment, the extracting unit 12 is a configuration that realizes extracting means.

[0015] The specifying unit 11 specifies, as a target time zone, the time zone in which the chronic stress tendency is significantly manifested in the biological signals among the biological signals acquired from the subject over a predetermined period.

[0016] The extracting unit 12 extracts one or more feature quantities used for machine learning of a stress degree estimation model or for estimating the stress degree using the estimation model from the biological signals acquired in the above-mentioned target time zone specified by the specifying unit 11.

[0017] As described above, in the information processing apparatus 1 according to this exemplary embodiment, a configuration including the above-mentioned specifying unit 11 and extracting unit 12 is adopted.

[0018] According to the above configuration, first, a time period of interest in which a specific tendency caused by chronic stress becomes prominent in the biological signal is identified. Then, a feature amount is extracted from the biological signal of the subject acquired during the time period of interest. As a result, an effect that it is possible to extract an appropriate feature amount for use in machine learning of a stress level estimation model or estimation of the stress level using the estimation model is obtained.

[0019] And the feature amount having a high correlation with chronic stress obtained by the above configuration can be used, for example, together with correct answer data indicating the chronic stress level corresponding to the feature amount, to generate teacher data for use in machine learning of an estimation model for estimating the chronic stress level. As a result, it becomes possible to efficiently construct an estimation model with high estimation accuracy for chronic stress.

[0020] In another example, the feature amount having a high correlation with chronic stress obtained by the above configuration can be used as an input value of the above estimation model, and as a result, it becomes possible to more accurately perform the estimation of chronic stress.

[0021] The information processing apparatus 1 may be realized by a computer and a program of the computer. The above program is a feature amount extraction program that causes the above computer to function as the above specific unit 11 and extraction unit 12. According to this feature amount extraction program, the same effect as that of the above information processing apparatus 1 can be obtained.

[0022] <Flow of Feature Amount Extraction Method> FIG. 2 is a flowchart showing the flow of a feature amount extraction method executed by the information processing apparatus 1.

[0023] In step S11, the specific unit 11 identifies, as the time period of interest, a time period in which a chronic stress tendency appears prominently in the biological signal acquired from the subject over a predetermined period.

[0024] In step S12, the extraction unit 12 extracts one or more feature quantities used for machine learning of a stress level estimation model or estimation of a stress level using the estimation model from the biological signal acquired in the above-mentioned target time period specified by the specifying unit 11.

[0025] As described above, in the feature quantity extraction method according to this exemplary embodiment, a configuration including step S11 and step S12 is adopted. Therefore, according to the feature quantity extraction method according to this exemplary embodiment, similar to the information processing apparatus 1 described above, an effect that appropriate feature quantities used for machine learning of a stress level estimation model or estimation of a stress level using the estimation model can be extracted is obtained.

[0026] <Modification> The extraction unit 12 according to this exemplary embodiment may extract feature quantities using the biological signal acquired in the above-mentioned target time period and the biological signal acquired in another predetermined time period based on the target time period.

[0027] The specifying unit 11 specifies, as the target time period, a time period in which the biological signal shows a remarkable behavior during a day, and the extraction unit 12 may extract feature quantities based on the change in the above-mentioned biological signal at the start or end of the specified target time period.

[0028] Also, in the above-mentioned step S11, the specifying unit 11 may specify, as the target time period, a time period in which the biological signal shows a remarkable behavior during a day. Also, in the above-mentioned step S12, the extraction unit 12 may further extract feature quantities based on the change in the biological signal at the start or end of the target time period.

[0029] According to the above-mentioned configuration and method, first, as a time period in which the chronic stress tendency becomes remarkable, feature quantities are extracted from the biological signal acquired at least in a time period in which the biological signal tends to show a remarkable behavior during a day. Next, in addition to the above-mentioned feature quantities, the extraction unit 12 further extracts feature quantities based on the change in the above-mentioned biological signal observed before and after the target time period.

[0030] During the time period of interest, since specific trends caused by chronic stress are prominently manifested in the biological signals, specific trends caused by chronic stress are also prominently manifested in the changes in the biological signals at the start or end of the time period of interest. Therefore, extracting feature quantities based on the changes in the biological signals at the start or end of the time period of interest leads to an improvement in the estimation accuracy of chronic stress.

[0031] As an example, the extraction unit 12 may extract, as a feature quantity, the amount of change in a predetermined index value of the biological signal observed at the start or end of the time period of interest. Here, the predetermined index value of the biological signal may be the biological signal itself (raw data output from the sensor, so-called raw data) or a calculated value calculated based on the biological signal.

[0032] When it is known in advance that there is a significant correlation between the amount of change in the predetermined index value of the biological signal before and after the time period of interest and chronic stress, the extraction of the above-described amount of change as a feature quantity leads to an improvement in the estimation accuracy of chronic stress.

[0033] Note that the above-mentioned "time period during which the biological signal exhibits prominent behavior" may be a specific time period in which a predetermined index value of the biological signal shows a relatively high value during a day. The predetermined index value indicates, for example, an index obtained from the time-series data of the biological signal, an index obtained from the frequency data of the biological signal, or other values that are important for predicting chronic stress. As an example, the index value may be a value representing the heart rate, sweating amount, breathing rate, pulse wave, and body temperature, etc., that can be derived from the biological signal.

[0034] For example, the specific part 11 may specify a time period during the day when the heart rate tends to be high as the attention time period. According to the above configuration, as the time period when the chronic stress tendency becomes prominent, a feature amount is extracted from the biological signal acquired in a specific time period during the day when the heart rate tends to be high. As a result, an effect that a proper feature amount can be extracted for machine learning of the stress level estimation model or for estimating the stress level using the estimation model is obtained. And it becomes possible to efficiently construct an estimation model with high estimation accuracy of chronic stress or to perform the estimation of chronic stress with higher accuracy.

[0035] 〔Exemplary Embodiment 2〕 A second exemplary embodiment of the present invention will be described in detail with reference to the drawings. Note that components having the same functions as those described in the first exemplary embodiment are denoted by the same reference numerals, and the description thereof will not be repeated.

[0036] The inventors have conceived that the chronic stress tendency is prominently manifested in the biological signal in a time period when a predetermined index value (for example, heart rate, sweating amount, etc.) of the biological signal is high. And based on the circadian rhythm, it is known that a predetermined index value of the biological signal reaches a peak around morning time (for example, 10:00 am), and has a tendency to be high in a predetermined time period before and after that.

[0037] Therefore, in this exemplary embodiment, as an example, as the attention time period when the chronic stress tendency is prominently manifested in the biological signal, a predetermined time period before and after the morning time when the predetermined index value of the biological signal reaches a peak based on the circadian rhythm is specified.

[0038] <Configuration of Information Processing Apparatus> FIG. 3 is a block diagram showing the configuration of the information processing apparatus 4. Also, FIG. 3 also shows a wearable terminal 7 as an example of a device for measuring a biological signal.

[0039] The wearable terminal 7 measures the state of the wearer and outputs a biological signal as an output value. As an example, the wearable terminal 7 has a function of detecting the heart rate of the wearer and a function of detecting the sweating of the wearer. By the subject wearing the wearable terminal 7, heart rate data indicating the heart rate of the subject and sweating data indicating the amount of sweating of the subject are generated as biological signals. These biological signals are transmitted to the information processing device 4.

[0040] The wearable terminal 7 may further include a three-axis acceleration sensor. The output value of this acceleration sensor may be transmitted from the wearable terminal 7 to the information processing device 4 as a biological signal. By the subject wearing the wearable terminal 7, the body movement of the subject is detected by the acceleration sensor. Since it is known that body movement is correlated with the stress level of the subject, the output value of the acceleration sensor can be used as a biological signal to estimate the stress level. Note that the acceleration sensor is not limited to a three-axis one, and may be a one-axis or two-axis one. Also, since the body activity of the subject can be grasped by the output value of the wearable terminal 7 as an acceleration sensor, it is also possible to determine whether the measured biological signal is derived from body activity.

[0041] Hereinafter, for the sake of simplicity of explanation, an example is given in which all the biological signals obtained for the subject are measured by one wearable terminal 7 and transmitted to the information processing device 4. However, the information processing device 4 may obtain several types of biological signals from different devices respectively.

[0042] The information processing device 4 includes a control unit 40 that comprehensively controls each part of the information processing device 4, and a storage unit 41 that stores various data used by the information processing device 4. Further, the information processing device 4 includes an input unit 42 that receives the input of data to the information processing device 4, an output unit 43 for the information processing device 4 to output data, and a communication unit 44 for the information processing device 4 to communicate with other devices (for example, the wearable terminal 7).

[0043] The control unit 40 includes a biological signal acquisition unit 401, a questionnaire data acquisition unit 402, a stress level calculation unit 403, a specification unit 404, an extraction unit 405, a determination unit 406, a teacher data generation unit 407, a learning processing unit 408, and an estimation unit 409. Further, the storage unit 41 stores a biological signal 411, questionnaire data 412, stress level data 413, feature quantity data 414, teacher data 415, an estimation model 416, and estimation result data 417.

[0044] In this exemplary embodiment, the specification unit 404 is configured to implement a specification means. In this exemplary embodiment, the extraction unit 405 is configured to implement an extraction means. In this exemplary embodiment, the determination unit 406 is configured to implement a determination means. The estimation unit 409 is configured to implement an estimation means. Note that in this exemplary embodiment, the determination unit 406 may be omitted. The determination unit 406 will be described in the following exemplary embodiment.

[0045] The biological signal acquisition unit 401 acquires a biological signal of a subject and stores the acquired biological signal in the storage unit 41. The biological signal stored in the storage unit 41 is the biological signal 411. The biological signal 411 may include those used for generating the teacher data 415 and those used for estimating the stress level.

[0046] The questionnaire data acquisition unit 402 acquires the result of a questionnaire related to the stress level of the subject during the period when the biological signal 411 for generating the teacher data 415 is measured, and stores the questionnaire data 412 indicating the acquired result in the storage unit 41. This questionnaire is a questionnaire conducted on the subject in order to calculate the stress level of the subject. This questionnaire may be of any content as long as it reflects the stress level of the subject. For example, it may be a stress questionnaire of PSS (Perceived Stress Scale). The stress questionnaire of PSS is a questionnaire in which, for each of a plurality of questions about how the subject felt and behaved during the target period, the subject is allowed to select the corresponding one from a plurality of options.

[0047] The stress level calculation unit 403 calculates the stress level of the subject using the questionnaire data 412, and stores stress level data 413 indicating the calculated stress level in the storage unit 41. Any arbitrary method can be applied as the method for calculating the stress level. For example, when the questionnaire data 412 is data indicating the result of the PSS stress questionnaire, the stress level calculation unit 403 calculates the PSS score.

[0048] The specifying unit 404 specifies, as the attention time zone, the time zone in which the chronic stress tendency is significantly manifested in the biological signal acquired from the subject over a predetermined period.

[0049] In the present exemplary embodiment, the specifying unit 404 specifies, as the attention time zone, a predetermined time zone before and after the morning time when a predetermined index value of the biological signal peaks based on the circadian rhythm.

[0050] As an example, the predetermined index value may be the heart rate. The inventors have conceived that the chronic stress tendency is significantly manifested in the biological signal in the time zone where the heart rate is high. And based on the circadian rhythm, it is known that the heart rate peaks around the morning time (for example, 10:00 am), and has a tendency to be high in the predetermined time zone before and after that. Therefore, as an example, the specifying unit 404 may be configured to specify, based on the circadian rhythm, a predetermined time zone around 10:00 when the heart rate indicated by the heart rate data peaks (for example, the time zone between 7:00 am and 1:00 pm) as the attention time zone.

[0051] When the peak time in the circadian rhythm of a biological signal different from the heart rate such as the sweating amount is known in advance, the specifying unit 404 may specify, as the attention time zone, a predetermined time zone with a large sweating amount before and after the time when the sweating amount indicated by the sweating data peaks.

[0052] In another example, the specific part 404 may analyze the biological signal acquired over a predetermined period, and identify the time point S when an index contributing to the estimation of a chronic stress tendency begins to prominently appear in the biological signal, and the time point E when the index contributing to the estimation of the prominent chronic stress tendency ends. Then, the specific part 404 may identify the time period from the identified time point S to the time point E as the attention time period. The index contributing to the estimation of the chronic stress tendency in the biological signal may be, for example, the heart rate. For example, the specific part 404 may identify the time period from the time point S when a value such as the numerical value of the heart rate itself, which is an example of the above index, or the index value related to the chronic stress tendency calculated from the heart rate reaches a predetermined threshold to the time point E when the value falls below the predetermined threshold as the attention time period.

[0053] The extraction part 405 extracts one or more feature quantities used for machine learning of the stress degree estimation model or the estimation of the stress degree using the estimation model from the biological signal acquired in the identified attention time period. For example, the extraction part 405 may calculate a feature quantity from the biological signal 411 and store the calculated feature quantity in the storage part 41. The feature quantity data 414 is data indicating the feature quantity extracted by the extraction part 405 and stored in the storage part 41. The feature quantity data 414 may include the feature quantity used for the generation of the teacher data 415. Hereinafter, the feature quantity used for the generation of the teacher data 415 is referred to as a learning feature quantity. That is, the learning feature quantity is the feature quantity used for the machine learning of the stress degree estimation model.

[0054] In addition, the feature quantity data 414 may include at least one or more feature quantities extracted from the biological signal in the attention time period. That is, the feature quantity data 414 may include multiple types of feature quantities. The feature quantity data 414 may include the feature quantity extracted in consideration of the biological signal outside the attention time period. For example, the feature quantity data 414 may include the feature quantity extracted based on the change in the biological signal at the start or end of the attention time period.

[0055] Furthermore, the feature amount data 414 may also include feature amounts used for estimating the degree of stress. Hereinafter, the feature amounts used for estimating the degree of stress are referred to as estimation feature amounts. The estimation feature amount is a feature amount generated from a biological signal of a subject to be estimated for the degree of stress during a predetermined period for which the degree of stress is to be measured.

[0056] The teacher data generation unit 407 generates teacher data by associating the degree of stress shown in the stress degree data 413 with the combination of one or more learning feature amounts extracted by the extraction unit 405 as correct answer data. Then, the teacher data generation unit 407 stores the generated teacher data in the storage unit 41 as teacher data 415.

[0057] The learning processing unit 408 generates an estimation model using the teacher data 415 for learning, with one or more learning feature amounts extracted by the extraction unit 405 as explanatory variables and the degree of stress as the objective variable. Then, the learning processing unit 408 stores the generated estimation model in the storage unit 41 as the estimation model 416.

[0058] The estimation unit 409 estimates the degree of stress of the subject using the estimation feature amount generated from the biological signal of the subject. More specifically, the estimation unit 409 calculates an estimated value of the degree of stress by inputting the estimation feature amount included in the feature amount data 414 into the estimation model 416. Then, the estimation unit 409 stores the estimation result data 417 indicating the estimation result of the degree of stress in the storage unit 41.

[0059] <Modification Example> The influence of chronic stress on biological signals may differ between men and women. For example, there is a paper reporting that in the case of female subjects, in the heart rate data based on the circadian rhythm, the heart rate tends to decrease under chronic stress, while in the case of male subjects, the heart rate tends to increase under chronic stress.

[0060] Therefore, the estimation model 416 is preferably generated separately for men and women. Specifically, the extraction unit 405 extracts male feature quantities from the biological signals of men measured during the target time period (for example, the time period between around 7:00 am and 13:00). Further, the extraction unit 405 extracts female feature quantities from the biological signals of women measured during the target time period in order to generate an estimation model for women. The teacher data generation unit 407 generates male teacher data for generating an estimation model for men using the male feature quantities, and generates female teacher data for generating an estimation model for women using the female feature quantities. In this way, the learning processing unit 408 can generate the estimation model 416 separately for men and women using the teacher data for each gender. Then, the estimation unit 409 can estimate the stress level using the estimation model for men for male subjects, and can estimate the stress level using the estimation model for women for female subjects.

[0061] Each of the above-described units included in the information processing apparatus 4 may be realized by a plurality of computers. For example, a feature quantity extraction apparatus including a control unit 40 including the biological signal acquisition unit 401, the specification unit 404, the extraction unit 405, and the determination unit 406, and a storage unit 41 storing the biological signal 411 and the feature quantity data 414 may be realized. A teacher data generation apparatus including a control unit 40 including the questionnaire data acquisition unit 402, the stress level calculation unit 403, and the teacher data generation unit 407, and a storage unit 41 storing the questionnaire data 412, the stress level data 413, and the teacher data 415 may be realized. An estimation model generation apparatus including a control unit 40 including the learning processing unit 408 and a storage unit 41 storing the estimation model 416 may be realized. An estimation apparatus including a control unit 40 including the estimation unit 409 and a storage unit 41 storing the estimation result data 417 may be realized. The estimation apparatus may be configured to include components of the feature quantity extraction apparatus, specifically, the specification unit 404 and the extraction unit 405. And an information processing system configured such that the wearable terminal 7, the feature quantity extraction apparatus, the teacher data generation apparatus, the estimation model generation apparatus, and the estimation apparatus are communicably connected to each other via a communication network also falls within the scope of the present invention.

[0062] <Flow of information processing method in learning phase> FIG. 4 is a flowchart showing the flow of an information processing method in a learning phase executed by an information processing apparatus 4 according to Exemplary Embodiment 2 of the present invention. The information processing method shown in FIG. 4 includes, as an example, a feature extraction method, a teacher data generation method, and an estimation model generation method of the present invention. In this exemplary embodiment, steps S32 and S33 are processes for realizing the feature extraction method, step S35 is a process for realizing the teacher data generation method, and step S35 is a process for realizing the estimation model generation method.

[0063] Each of the above processes can also be realized by a program. That is, a feature extraction program for causing a computer to execute the processes of steps S32 to S33 is also included in the scope of this exemplary embodiment. Similarly, a teacher data generation program for causing a computer to execute the process of step S35 of generating teacher data using the feature amount extracted in step S33 is also included in the scope of this exemplary embodiment. And an estimation model generation program for causing a computer to execute the process of step S36 of generating an estimation model using the teacher data generated in step S35 is also included in the scope of this exemplary embodiment.

[0064] Note that the series of information processing methods shown in FIG. 4 may be executed by the above-described information processing system instead of the information processing apparatus 4. In this case, the execution subject of steps S31 to S33 is the above-described feature extraction apparatus, the execution subject of steps S34 to S35 is the above-described teacher data generation apparatus, and the execution subject of step S36 is the above-described estimation model generation apparatus.

[0065] The following describes an example of generating an estimation model using, as biological signals, the heartbeat data and sweating data of a subject measured by the wearable terminal 7. The biological signals to be used may be the biological signals of a single subject or the biological signals of a plurality of subjects, but are preferably the biological signals of a subject whose responsiveness to stress is close to that of the subject for whom the stress level is to be estimated. Also, for each subject, a questionnaire for calculating the stress level during the period when the biological signal was measured has been implemented, and it is assumed that the results are stored in the storage unit 41 as questionnaire data 412. Further, since all the feature amounts in FIG. 4 are the above-described feature amounts for learning, they are simply referred to as feature amounts in the description of FIG. 4.

[0066] In step S31, the biological signal acquisition unit 401 acquires a biological signal to be used for generating the estimation model. As described above, the biological signal acquired here is the heartbeat data and sweating data of the subject measured by the wearable terminal 7. Then, the biological signal acquisition unit 401 stores the acquired biological signal in the storage unit 41 as the biological signal 411.

[0067] In step S32, the specifying unit 404 specifies, as the attention time zone, the time zone in which the tendency of chronic stress is significantly manifested in the biological signal 411 recorded in step S31. In this exemplary embodiment, as an example, the time zone in which the biological signal shows significant behavior within a day is specified as the attention time zone. In this exemplary embodiment, more specifically, the predetermined time zone around the morning time when a predetermined index value of the biological signal peaks based on the circadian rhythm is specified as the attention time zone. As an example, the specifying unit 404 may specify, as the attention time zone, the predetermined time zone around 10 o'clock (for example, from 7 o'clock to 13 o'clock) when the heart rate obtained from the biological signal peaks.

[0068] In step S33, the extraction unit 405 extracts feature quantities from the biological signals measured in the target time period specified in step S32, out of the biological signals 411 stored in step S31. Specifically, the extraction unit 405 may extract multiple types of feature quantities from each of the heartbeat data and the sweating data. The extracted feature quantities are stored in the storage unit 41 as feature quantity data 414.

[0069] In step S34, the stress level calculation unit 403 calculates the stress level of the subject using the questionnaire data 412. Then, the stress level calculation unit 403 stores the calculated stress level in the storage unit 41 as stress level data 413. Note that the process of step S34 may be performed prior to step S35, may be performed prior to step S31, or may be performed in parallel with steps S31 to S33.

[0070] In step S35, the teacher data generation unit 407 generates teacher data by associating, with respect to the combination of one or more feature quantities extracted in step S33, the stress level calculated in step S34, indicated in the stress level data 413, as correct answer data. Then, the teacher data generation unit 407 stores the generated teacher data in the storage unit 41 as teacher data 415.

[0071] In step S36, the learning processing unit 408 generates an estimation model for the stress level by machine learning using the teacher data generated in step S35. Note that step S36 may include a series of processes of generating a plurality of estimation models, evaluating the estimation accuracy of each generated estimation model, and selecting a final estimation model based on the evaluation result. Then, the learning processing unit 408 stores the generated estimation model in the storage unit 41 as an estimation model 416. Thereby, the estimation model generation method ends.

[0072] <Flow of information processing method in the inference phase> FIG. 5 is a flowchart showing the flow of an information processing method in an inference phase executed by the information processing apparatus 4 according to Exemplary Embodiment 2 of the present invention. The information processing method shown in FIG. 5 includes, as an example, the feature extraction method of the present invention and the stress level estimation method. In this exemplary embodiment, steps S42 and S43 are processes for realizing the feature extraction method, and step S44 is a process for realizing the stress level estimation method.

[0073] Each of the above-described processes can also be realized by a program. That is, a feature extraction program for causing a computer to execute the processes of steps S42 to S43 is also included in the scope of this exemplary embodiment. And a stress level estimation program for causing a computer to execute the process of step S44 of inputting the feature amount extracted in step S43 into the estimation model generated in step S36 and estimating the stress level is also included in the scope of this exemplary embodiment.

[0074] Note that when the series of information processing methods shown in FIG. 5 are executed by the above-described information processing system instead of the information processing apparatus 4, the execution subject of steps S41 to S43 is the above-described feature extraction apparatus, and the execution subject of step S43 is the above-described estimation apparatus. Of course, the configuration may be such that the above-described estimation apparatus executes the processes of steps S41 to S44.

[0075] Note that hereinafter, an example of estimating the stress level of a subject in a month using the heartbeat data and sweating data for one month measured by the wearable terminal 7 as biological signals will be described, but the measurement period may be less than one month or longer than one month. Also, since all of the "feature amounts" described in FIG. 5 are the above-described feature amounts for estimation, they are simply referred to as feature amounts in the description of FIG. 5.

[0076] In step S41, the biological signal acquisition unit 401 acquires a biological signal. As described above, the biological signal acquired here is the heartbeat data and sweating data of the subject measured by the wearable terminal 7 for one month. Then, the biological signal acquisition unit 401 stores the acquired biological signal in the storage unit 41 as the biological signal 411.

[0077] In step S42, the specifying unit 404 specifies the target time period. The process of specifying the target time period executed in step S42 is the same as the process of specifying in step S32 in the above-described learning phase. That is, in the present exemplary embodiment, a predetermined time period (for example, from 7:00 to 13:00) around the morning time when a predetermined index value of the biological signal peaks based on the circadian rhythm is specified as the target time period.

[0078] In step S43, the extraction unit 405 extracts a feature amount from the biological signal measured in the target time period specified in step S42 among the biological signals 411 stored in step S41. The extracted feature amount is stored in the storage unit 41 as feature amount data 414.

[0079] In step S44, the estimation unit 409 estimates the stress level of the subject. Specifically, the estimation unit 409 inputs the feature amount extracted in step S43 to the estimation model 416. This estimation model 416 is the one generated in step S36 of FIG. 4. Then, the estimation unit 409 stores the output value of the estimation model 416 in the storage unit 41 as estimation result data 417. Note that the estimation unit 409 may output the estimated stress level to the output unit 43. Thereby, the method for estimating the stress level is completed.

[0080] As described above, in the information processing apparatus 4 according to the present exemplary embodiment, a configuration including the above-described specifying unit 404 and extraction unit 405 is adopted. In particular, the specifying unit 404 is configured to specify a predetermined time period around the morning time when a predetermined index value of the biological signal peaks based on the circadian rhythm as the target time period.

[0081] According to the above-mentioned configuration, a feature is extracted from a biosignal acquired during a predetermined time period (e.g., in the morning) when a predetermined index value of the biosignal (e.g., heart rate) is relatively high, including a time (e.g., around 10:00) when the predetermined index value of the biosignal is at its peak. The inventors have found that a chronic stress tendency is prominently shown in the biosignal during a time period when the predetermined index value of the biosignal (e.g., heart rate, sweat rate, etc.) is high. Based on the circadian rhythm, it is known that the predetermined index value of the biosignal peaks around morning time (e.g., 10:00 a.m.) and tends to be high during a predetermined time period around that time.

[0082] Therefore, features are extracted from the biosignal in a predetermined time period before and after the morning time when a predetermined index value of the biosignal peaks based on the circadian rhythm. This has the effect of making it possible to extract appropriate features to be used in machine learning of a stress level estimation model or in estimating a stress level using the estimation model. This makes it possible to efficiently build an estimation model with high estimation accuracy of chronic stress and to estimate chronic stress with even higher accuracy.

[0083] As described above, the teacher data generation method according to this exemplary embodiment employs a configuration including step S35 of generating teacher data to be used in machine learning by associating the subject's stress level as correct answer data with one or more features extracted by the feature extraction method including steps S32 to S33. Therefore, the teacher data generation method according to this exemplary embodiment has the effect of generating teacher data that can efficiently build an estimation model with high estimation accuracy of chronic stress.

[0084] As described above, in the estimation model generation method according to the present exemplary embodiment, a configuration including step S36 of generating an estimation model by machine learning using the teacher data generated by the teacher data generation method including step S35 is adopted. Therefore, according to the estimation model generation method according to the present exemplary embodiment, an effect that an estimation model with high estimation accuracy of chronic stress can be generated is obtained.

[0085] As described above, in the stress level estimation method according to the present exemplary embodiment, a configuration including step S44 of estimating the stress level of a subject using the estimation model generated by the estimation model generation method including step S36 is adopted. Therefore, according to the estimation method according to the present exemplary embodiment, an effect that the stress level related to chronic stress can be accurately estimated is obtained.

[0086] 〔Exemplary Embodiment 3〕 The third exemplary embodiment of the present invention will be described in detail with reference to the drawings. Note that components having the same functions as those described in the above exemplary embodiments are denoted by the same reference numerals, and the description thereof will not be repeated.

[0087] In the present exemplary embodiment, feature amounts are extracted for each attribute of the subject, teacher data is generated for each attribute, and an estimation model is generated for each attribute. Then, when estimating the stress level of the subject, the stress level of the subject is estimated using the estimation model corresponding to the attribute of the subject. The attribute of the subject may be, for example, gender.

[0088] In the present exemplary embodiment, as an example, a time zone of interest in which a chronic stress tendency is significantly manifested in a biological signal is specified based on a meal time zone.

[0089] Generally, a predetermined index value of a biological signal (e.g., heart rate, sweating amount, etc.) tends to increase with meals. Furthermore, since men tend to increase their food intake when they have a tendency towards chronic stress, the inventors have focused on the fact that a tendency towards chronic stress is prominently manifested in a predetermined index of the biological signal during meal time zones. Therefore, in this exemplary embodiment, as an example, for male subjects, a standard lunch time zone is specified as the attention time zone in which a tendency towards chronic stress is prominently manifested in the biological signal.

[0090] In addition, the inventors have speculated that a predetermined index value of a female's biological signal tends to decrease during meal time zones, and the manifestation of a tendency towards chronic stress is dulled. Therefore, in this exemplary embodiment, as an example, for female subjects, time zones other than the standard lunch time zone are specified as the attention time zones. For example, the final attention time zone may be specified by excluding the standard lunch time zone from the attention time zones specified based on Exemplary Embodiment 1 or 2.

[0091] Note that specifying the attention time zone based on the lunch time zone is particularly due to the fact that the lunch time zone has less variation in time zone among people. Note that the lunch time zone has even less variation if it is limited to working days, so it is more suitable as the attention time zone.

[0092] <Configuration of the information processing device> In this exemplary embodiment, the specifying unit 404 is configured to specify the attention time zone based on a predetermined lunch time zone. As an example, a questionnaire may be taken in advance from the subject about the lunch time zone, and the most standard time zone (e.g., from 12:00 to 13:00) may be defined as the lunch time zone. Specifically, the specifying unit 404 specifies the above-mentioned standard lunch time zone of the day as the attention time zone for male subjects. Also, the specifying unit 404 specifies a time zone other than the above-mentioned standard lunch time zone of the day as the attention time zone for female subjects. For example, after the specifying unit 404 specifies the attention time zone by the method shown in Exemplary Embodiment 1 or Exemplary Embodiment 2, if the above-mentioned standard lunch time zone is included in the attention time zone, the standard lunch time zone may be excluded from the above-mentioned attention time zone.

[0093] In the present exemplary embodiment, the extraction unit 405 extracts feature amounts for each gender based on the time zone of interest specified by the specifying unit 404 for each gender.

[0094] Specifically, the extraction unit 405 extracts feature amounts from the biological signals acquired during the specified lunch time zone for male subjects. Hereinafter, the feature amounts for males extracted from the biological signals acquired during the lunch time zone are referred to as first feature amounts. Among the first feature amounts, the feature amounts used for generating the teacher data 415 are referred to as first learning feature amounts. Among the first feature amounts, the feature amounts used for estimating the stress level are referred to as first estimation feature amounts.

[0095] Specifically, the extraction unit 405 extracts feature amounts from the biological signals acquired during time zones other than the lunch time zone for female subjects. Hereinafter, the feature amounts for females extracted from the biological signals acquired during time zones other than the lunch time zone are referred to as second feature amounts. Among the second feature amounts, the feature amounts used for generating the teacher data 415 are referred to as second learning feature amounts. Among the second feature amounts, the feature amounts used for estimating the stress level are referred to as second estimation feature amounts.

[0096] That is, in the present exemplary embodiment, the feature amount data 414 includes first learning feature amounts, first estimation feature amounts, second learning feature amounts, and second estimation feature amounts.

[0097] In the present exemplary embodiment, the teacher data generation unit 407 generates teacher data by associating the stress levels indicated in the stress level data 413 of male subjects with the combinations of one or more first learning feature amounts as correct data. The above-described first learning feature amounts are those extracted from the biological signals of male subjects by the extraction unit 405, and the teacher data generated as described above becomes teacher data for constructing an estimation model for males. Hereinafter, the teacher data for constructing an estimation model for males is referred to as first teacher data.

[0098] Further, the teacher data generation unit 407 generates teacher data by associating the stress level indicated by the stress level data 413 of female subjects with the combination of one or more second learning feature amounts as correct answer data. The above-described second learning feature amounts are extracted from the biological signals of female subjects by the extraction unit 405, and the teacher data generated as described above becomes teacher data for constructing an estimation model for females. Hereinafter, the teacher data for constructing an estimation model for females is referred to as second teacher data.

[0099] That is, in this exemplary embodiment, the teacher data 415 includes the first teacher data and the second teacher data.

[0100] In this exemplary embodiment, the learning processing unit 408 generates an estimation model using the first teacher data, with one or more first learning feature amounts extracted by the extraction unit 405 as explanatory variables and the stress level of male subjects as the objective variable. Hereinafter, the above-described estimation model for estimating the stress level of male subjects is referred to as the first estimation model.

[0101] Further, the learning processing unit 408 generates an estimation model using the second teacher data, with one or more second learning feature amounts extracted by the extraction unit 405 as explanatory variables and the stress level of female subjects as the objective variable. Hereinafter, the above-described estimation model for estimating the stress level of female subjects is referred to as the second estimation model.

[0102] That is, in this exemplary embodiment, the estimation model 416 includes the first estimation model and the second estimation model.

[0103] When the subject to be estimated is male, the estimation unit 409 estimates the stress level of the subject using the first estimation model, and when the subject to be estimated is female, the estimation unit 409 estimates the stress level of the subject using the second estimation model.

[0104] <Flow of the information processing method in the learning phase> The flow of the information processing method in the learning phase, which is executed by the information processing apparatus 4 according to the exemplary embodiment 3 of the present invention, will be described based on FIG. 4. In the information processing method of the exemplary embodiment 3, the differences from the information processing method of the exemplary embodiment 2 are as follows.

[0105] In step S32, when the biological signal acquired in step S31 is a biological signal of a male subject, the specifying unit 404 specifies the standard lunch time zone of the day as the attention time zone. When the biological signal acquired in step S31 is a biological signal of a female subject, the specifying unit 404 specifies a time zone other than the standard lunch time zone of the day as the attention time zone.

[0106] In step S33, when the biological signal acquired in step S31 is a biological signal of a male subject, the extraction unit 405 extracts the first learning feature amount from the biological signal acquired in the specified lunch time zone. When the biological signal acquired in step S31 is a biological signal of a female subject, the extraction unit 405 extracts the second learning feature amount from the biological signal acquired in the time zone other than the lunch time zone.

[0107] In step S35, when the biological signal acquired in step S31 is a biological signal of a male subject, the teacher data generation unit 407 generates the first teacher data. The first teacher data is generated by associating the combination of the first learning feature amounts extracted in step S33 with the stress level calculated in step S34 as correct answer data. When the biological signal acquired in step S31 is a biological signal of a female subject, the teacher data generation unit 407 generates the second teacher data. The second teacher data is generated by associating the combination of the second learning feature amounts extracted in step S33 with the stress level calculated in step S34 as correct answer data.

[0108] In step S36, when the biological signal acquired in step S31 is a biological signal of a male subject, the learning processing unit 408 generates a first estimation model for estimating the stress level of the male subject by machine learning using the first teacher data generated in step S35. When the biological signal acquired in step S31 is a biological signal of a female subject, the learning processing unit 408 generates a second estimation model for estimating the stress level of the female subject by machine learning using the second teacher data generated in step S35.

[0109] <Flow of information processing method in inference phase> The flow of the information processing method in the inference phase executed by the information processing apparatus 4 according to the exemplary embodiment 3 of the present invention will be described with reference to FIG. 5. In the information processing method of the exemplary embodiment 3, the differences from the information processing method of the exemplary embodiment 2 are as follows.

[0110] In step S42, when the biological signal acquired in step S41 is a biological signal of a male subject, the specifying unit 404 specifies the standard lunch time zone of the day as the attention time zone. When the biological signal acquired in step S41 is a biological signal of a female subject, the specifying unit 404 specifies a time zone other than the standard lunch time zone of the day as the attention time zone.

[0111] In step S43, when the biological signal acquired in step S41 is a biological signal of a male subject, the extraction unit 405 extracts the first feature quantity for estimation from the biological signal acquired in the specified lunch time zone. When the biological signal acquired in step S41 is a biological signal of a female subject, the extraction unit 405 extracts the second feature quantity for estimation from the biological signal acquired in a time zone other than the above-mentioned lunch time zone.

[0112] In step S44, when the biological signal acquired in step S41 is a biological signal of a male subject, the estimation unit 409 inputs the first feature quantity for estimation extracted in step S43 into the first estimation model generated in step S36. The estimation unit 409 stores the output value of the first estimation model in the storage unit 41 as the above-described estimation result data 417 of the male subject. When the biological signal acquired in step S41 is a biological signal of a female subject, the estimation unit 409 inputs the second feature quantity for estimation extracted in step S43 into the second estimation model generated in step S36. The estimation unit 409 stores the output value of the second estimation model in the storage unit 41 as the above-described estimation result data 417 of the female subject.

[0113] As described above, in the information processing apparatus 4 according to the present exemplary embodiment, a configuration including the above-described specifying unit 404 and extraction unit 405 is adopted. In particular, the specifying unit 404 is configured to specify, for a male subject, the standard lunch time zone of the subject during a day as the attention time zone. Further, in particular, the extraction unit 405 is configured to extract a feature quantity from the biological signal acquired in the specified lunch time zone for the male subject.

[0114] According to the above configuration, during a day, for a male, a feature quantity is extracted from the biological signal of the male subject, which is acquired during a meal time when a predetermined index value (for example, heart rate, sweating amount, etc.) of the biological signal tends to be relatively high, and particularly during the lunch time zone with little variation in time zone among individuals. The meal time zone is considered to be a time zone in which the tendency of chronic stress becomes prominent in the biological signal of a male. Therefore, by narrowing down the biological signal for which the feature quantity is to be extracted to the biological signal in the lunch time zone, it becomes possible to efficiently construct an estimation model with high estimation accuracy for the chronic stress of a male or to more accurately perform the estimation of the chronic stress of a male.

[0115] As described above, in the information processing apparatus 4 according to the present exemplary embodiment, a configuration including the above-described specifying unit 404 and extracting unit 405 is adopted. In particular, the specifying unit 404 is configured to specify, for a female subject, a time period other than the subject's standard lunch time period within a day as a target time period. Further, in particular, the extracting unit 405 is configured to extract a feature amount from a biological signal acquired in a time period other than the lunch time period for a female subject.

[0116] According to the above configuration, during a meal time when a predetermined index value (for example, heart rate, sweating amount, etc.) of a biological signal tends to be relatively low for a woman within a day, and in particular, a feature amount is extracted excluding the biological signal of the lunch time period with little variation in time period among individuals. When a woman has a tendency of chronic stress, the amount of food intake decreases, and thus it can be estimated that the biological signal associated with the meal becomes dull during the meal time period. Therefore, by narrowing down the biological signal for which the feature amount is to be extracted by excluding at least the lunch time period from the target time period, it becomes possible to efficiently construct an estimation model with improved estimation accuracy of a woman's chronic stress or to improve the estimation accuracy of a woman's chronic stress.

[0117] 〔Exemplary Embodiment 4〕 A fourth exemplary embodiment of the present invention will be described in detail with reference to the drawings. Note that components having the same functions as those described in the above exemplary embodiments are denoted by the same reference numerals, and the description thereof will not be repeated.

[0118] In the present exemplary embodiment, a feature amount is extracted for each attribute of the subject, teacher data is generated for each attribute, and an estimation model is generated for each attribute. Then, when estimating the stress level of the subject, the stress level of the subject is estimated using the estimation model corresponding to the attribute of the subject. The attribute of the subject may be, for example, gender.

[0119] In the present exemplary embodiment, as an example, a time period of interest in which a chronic stress tendency is significantly manifested in a biological signal is specified based on a time period during which a subject is exposed to an acute stress stimulus. Hereinafter, the time period during which a subject is exposed to an acute stress stimulus is referred to as a stress occurrence time period.

[0120] The inventors of the present invention presumed that a predetermined index value (for example, heart rate, sweating amount, etc.) of a biological signal of a woman tends to increase during an acute stress occurrence time period, and a chronic stress tendency may be significantly manifested. Therefore, in the present exemplary embodiment, as an example, for female subjects, the above-described stress occurrence time period is specified as the time period of interest in which a chronic stress tendency is significantly manifested in a biological signal.

[0121] In addition, the inventors of the present invention focused on the fact that a predetermined index value of a biological signal of a man with a chronic stress tendency tends to be blunted during an acute stress occurrence time period, and thus the manifestation of the chronic stress tendency that can be estimated from the biological signal is blunted. Therefore, in the present exemplary embodiment, as an example, for male subjects, a time period other than the above-described stress occurrence time period is specified as the time period of interest. For example, the final time period of interest may be specified by excluding the stress occurrence time period from the time period of interest specified based on the configuration of at least any one of Exemplary Embodiments 1 to 3.

[0122] <Configuration of Information Processing Apparatus> In the present exemplary embodiment, as shown in FIG. 3, the control unit 40 includes a determination unit 406. The determination unit 406 determines whether or not a subject is in a state of being exposed to an acute stress stimulus based on a biological signal. For example, the determination unit 406 may analyze the biological signal as follows to detect the stress occurrence time period.

[0123] For example, the determination unit 406 determines whether or not a subject is in a state of being exposed to an acute stress stimulus by using at least either the heart rate indicated by the heart rate data obtained from the subject or the sweating amount indicated by the sweating data.

[0124] For example, when the heartbeat data is data in which the heart rates measured every unit time are arranged in time series, the determination unit 406 may compare the heart rate with a predetermined threshold value every unit time to determine whether acute stress stimulation has been exposed. For example, the determination unit 406 may determine that "acute stress stimulation has been exposed" at that time when the heart rate is equal to or higher than a predetermined threshold value. The determination unit 406 may detect the set of times when it is determined that "acute stress stimulation has been exposed" as the stress occurrence time zone and output it to the specifying unit 404.

[0125] In another example, the determination unit 406 detects a sharp rise time SS of the heart rate at which the increase amount of the heart rate per unit time in the heartbeat data is equal to or more than a predetermined threshold value, and a time point EE at which the heart rate that has risen in this way starts to decline, and may output the time zone from the time point SS to the time point EE as the stress occurrence time zone to the specifying unit 404.

[0126] In addition, the determination unit 406 may determine the presence or absence of acute stress stimulation using a plurality of types of biological signals and detect the stress occurrence time zone. For example, the time point at which an increase in heart rate and an increase in sweating amount are simultaneously observed may be specified as the start time point of the stress occurrence time zone, and the time point at which at least one of the heart rate and the sweating amount has decreased to the normal level may be specified as the end time point of the stress occurrence time zone. Also, for example, the determination unit 406 may detect, as the stress occurrence time zone, a period in which the pattern of fluctuations in the measured biological signal during a predetermined period corresponds to a pattern specific to a state of being exposed to acute stress stimulation.

[0127] The specifying unit 404 specifies, for a female subject, the stress occurrence time zone determined by the determination unit 406 as the attention time zone.

[0128] In addition, the specifying unit 404 specifies, for a male subject, a time zone other than the stress occurrence time zone determined by the determination unit 406 as the attention time zone. For example, the specifying unit 404 may exclude the above-described stress occurrence time zone from the attention time zone specified by the configuration of at least any one of the exemplary embodiments 1 to 3 to specify the final attention time zone.

[0129] Similar to the third exemplary embodiment, the extraction unit 405 extracts feature quantities by attribute, that is, by gender, based on the specified time zone of interest. The feature quantity data 414 may include first learning feature quantities and first estimation feature quantities for men, and second learning feature quantities and second estimation feature quantities for women.

[0130] Similar to the third exemplary embodiment, the teacher data generation unit 407 generates teacher data by gender. The teacher data 415 may include first teacher data for men and second teacher data for women.

[0131] Similar to the third exemplary embodiment, the learning processing unit 408 generates estimation models by gender. The estimation model 416 may include a first estimation model for men and a second estimation model for women.

[0132] Similar to the third exemplary embodiment, the estimation unit 409 estimates the stress level by gender.

[0133] <Flow of the information processing method in the learning phase> FIG. 6 is a flowchart showing the flow of the information processing method in the learning phase executed by the information processing apparatus 4 according to the fourth exemplary embodiment of the present invention. The information processing method shown in FIG. 6 includes the feature quantity extraction method, the teacher data generation method, and the estimation model generation method of the present invention, similar to the second and third exemplary embodiments. In this exemplary embodiment, steps S52 to S54 are processes for realizing the feature quantity extraction method, step S56 is a process for realizing the teacher data generation method, and step S57 is a process for realizing the estimation model generation method.

[0134] Each of the above processes can also be realized by a program. That is, a feature extraction program for causing a computer to execute the processes of steps S52 to S54 is also included in the scope of the exemplary embodiment. Similarly, a teacher data generation program for causing a computer to execute the process of step S56 of generating teacher data using the feature amounts extracted in step S54 is also included in the scope of the exemplary embodiment. And an estimation model generation program for causing a computer to execute the process of step S57 of generating an estimation model using the teacher data generated in step S56 is also included in the scope of the exemplary embodiment.

[0135] Note that the series of information processing methods shown in FIG. 6 may be executed by the above-described information processing system instead of the information processing apparatus 4. In this case, the execution subject of steps S51 to S54 is the above-described feature extraction apparatus, the execution subject of steps S55 to S56 is the above-described teacher data generation apparatus, and the execution subject of step S57 is the above-described estimation model generation apparatus.

[0136] Hereinafter, similar to Exemplary Embodiments 2 and 3, an example of generating an estimation model using the subject's heartbeat data and sweating data measured by the wearable terminal 7 as biological signals will be described. Hereinafter, in the information processing method of Exemplary Embodiment 4, the points common to each of the above information processing methods will be described as "similarly to Exemplary Embodiment ~" or "similarly to step S ~", and the same explanation will not be repeated.

[0137] In step S51, the biological signal acquisition unit 401 acquires a biological signal in the same manner as in step S31.

[0138] In step S52, the determination unit 406 determines whether or not the subject is in a state of being exposed to an acute stress stimulus based on the biological signal acquired in step S51. For example, the determination unit 406 may adopt some of the above-described specific determination methods to detect the stress occurrence time zone.

[0139] In step S53, when the biological signal acquired in step S51 is the biological signal of a male subject, the specifying unit 404 specifies a time zone other than the stress occurrence time zone as the target time zone. When the biological signal acquired in S51 is the biological signal of a female subject, the specifying unit 404 specifies the stress occurrence time zone as the target time zone.

[0140] In step S54, when the biological signal acquired in step S51 is the biological signal of a male subject, the extraction unit 405 extracts the first learning feature amount from the biological signals acquired in the time zone other than the stress occurrence time zone. When the biological signal acquired in step S51 is the biological signal of a female subject, the extraction unit 405 extracts the second learning feature amount from the biological signals acquired in the stress occurrence time zone.

[0141] In step S55, the stress level calculation unit 403 calculates the stress level of the subject in the same manner as in step S34.

[0142] In step S56, when the biological signal acquired in step S51 is the biological signal of a male subject, the teacher data generation unit 407 generates the first teacher data. The first teacher data is generated by associating the stress level calculated in step S55 with the combination of the first learning feature amounts extracted in step S54 as correct answer data. When the biological signal acquired in step S51 is the biological signal of a female subject, the teacher data generation unit 407 generates the second teacher data. The second teacher data is generated by associating the stress level calculated in step S55 with the combination of the second learning feature amounts extracted in step S54 as correct answer data.

[0143] In step S57, when the biological signal acquired in step S51 is the biological signal of a male subject, the learning processing unit 408 generates a first estimation model for estimating the stress level of the male subject by machine learning using the first teacher data generated in step S56. When the biological signal acquired in step S51 is the biological signal of a female subject, the learning processing unit 408 generates a second estimation model for estimating the stress level of the female subject by machine learning using the second teacher data generated in step S56.

[0144] <Flow of information processing method in inference phase> FIG. 7 is a flowchart showing the flow of an information processing method in the inference phase executed by the information processing apparatus 4 according to the exemplary embodiment 4 of the present invention. The information processing method shown in FIG. 7 includes, as an example, the feature extraction method of the present invention and the stress level estimation method. In this exemplary embodiment, steps S62 to S64 are processes for realizing the feature extraction method, and step S65 is a process for realizing the stress level estimation method.

[0145] Each of the above processes can also be realized by a program. That is, a feature extraction program for causing a computer to execute the processes of steps S62 to S64 is also included in the scope of this exemplary embodiment. And a stress level estimation program for causing a computer to execute the process of step S65 of estimating the stress level using the estimation model generated in step S57 is also included in the scope of this exemplary embodiment.

[0146] When the series of information processing methods shown in FIG. 7 are executed by the above-described information processing system instead of the information processing apparatus 4, the execution entity of steps S61 to S64 is the above-described feature extraction apparatus, and the execution entity of step S65 is the above-described estimation apparatus.

[0147] Hereinafter, similar to Exemplary Embodiments 2 and 3, an example of estimating the stress level of a subject in a month using one-month worth of heartbeat data and sweating data measured by the wearable terminal 7 as biological signals will be described. Hereinafter, in the information processing method of Exemplary Embodiment 4, points common to each of the above-described information processing methods will be described as "similarly to Exemplary Embodiment ~" or "similarly to Step S~", and the same explanation will not be repeated.

[0148] In Step S61, the biological signal acquisition unit 401 acquires a biological signal in the same manner as in Step S41.

[0149] In Step S62, the determination unit 406 determines whether or not the subject is in a state of being exposed to acute stress stimuli in the same manner as in Step S52. For example, the determination unit 406 may adopt some of the specific determination methods described above to detect the stress occurrence time zone.

[0150] In Step S63, when the biological signal acquired in Step S61 is a biological signal of a male subject, the specifying unit 404 specifies a time zone other than the stress occurrence time zone as the attention time zone. When the biological signal acquired in Step S61 is a biological signal of a female subject, the specifying unit 404 specifies the stress occurrence time zone as the attention time zone.

[0151] In Step S64, when the biological signal acquired in Step S61 is a biological signal of a male subject, the extraction unit 405 extracts a first estimation feature amount from the biological signal acquired in a time zone other than the stress occurrence time zone. When the biological signal acquired in Step S61 is a biological signal of a female subject, the extraction unit 405 extracts a second estimation feature amount from the biological signal acquired in the stress occurrence time zone.

[0152] In step S65, when the biological signal acquired in step S61 is a biological signal of a male subject, the estimation unit 409 inputs the first feature quantity for estimation extracted in step S64 into the first estimation model generated in step S57. The estimation unit 409 stores the output value of the first estimation model in the storage unit 41 as the above-described estimation result data 417 of the male subject. When the biological signal acquired in step S61 is a biological signal of a female subject, the estimation unit 409 inputs the second feature quantity for estimation extracted in step S64 into the second estimation model generated in step S57. The estimation unit 409 stores the output value of the second estimation model in the storage unit 41 as the above-described estimation result data 417 of the female subject.

[0153] As described above, in the information processing apparatus 4 according to the present exemplary embodiment, a configuration including the above-described determination unit 406, identification unit 404, and extraction unit 405 is adopted. The determination unit 406 is configured to determine whether or not the subject is in a state of being exposed to acute stress stimulation based on the biological signal. The identification unit 404 is configured to identify, for a female subject, the stress occurrence time period determined to be in a state of being exposed to acute stress stimulation as the attention time period. The extraction unit 405 is configured to extract a feature quantity from the biological signal acquired during the identified stress occurrence time period for the female subject.

[0154] It is presumed that the acute stress occurrence time period becomes prominent in the chronic stress tendency in the biological signal in females. According to the above-described configuration, a feature quantity is extracted from the biological signal acquired during the stress occurrence time period among the biological signals of the female subject. Therefore, by narrowing down the biological signal for which the feature quantity is to be extracted to the biological signal during the stress occurrence time period, it becomes possible to efficiently construct an estimation model with high estimation accuracy for chronic stress in females or to perform the estimation of chronic stress in females with higher accuracy.

[0155] As described above, in the information processing apparatus 4 according to the present exemplary embodiment, a configuration including the above-described determination unit 406, identification unit 404, and extraction unit 405 is adopted. The determination unit 406 is configured to determine whether or not the subject is in a state of being exposed to acute stress stimulation based on the biological signal. The identification unit 404 is configured to identify, for a male subject, a time period other than the stress occurrence time period determined to be a state in which the subject is exposed to acute stress stimulation as the target time period. The extraction unit 405 is configured to extract a feature amount from the biological signal acquired in a time period other than the stress occurrence time period for a male subject.

[0156] It is considered that the acute stress occurrence time period is a time when the manifestation of the chronic stress tendency in the biological signal is dulled in men. According to the above-described configuration, a feature amount is extracted from the biological signal acquired in a time period other than the stress occurrence time period among the biological signals of the male subject. As a result, it becomes possible to efficiently construct an estimation model with high estimation accuracy for male chronic stress or to perform the estimation of male chronic stress with higher accuracy.

[0157] 〔Exemplary Embodiment 5〕 The fifth exemplary embodiment of the present invention will be described in detail with reference to the drawings. Note that components having the same functions as those described in the above exemplary embodiments are denoted by the same reference numerals, and the description thereof will not be repeated.

[0158] In the present exemplary embodiment, the attribute of the subject who provided the acquired biological signal is discriminated, and different information processing is executed for each discriminated attribute.

[0159] <Configuration of Information Processing Apparatus> In the present exemplary embodiment, the biological signal acquisition unit 401 shown in FIG. 3 acquires, together with the biological signal, attribute information indicating the attribute of the subject who provided the biological signal. In the present exemplary embodiment, the attribute of the subject is, for example, gender. Therefore, in the present exemplary embodiment, the attribute information is information indicating the gender of the subject.

[0160] <Flow of Information Processing Method in Learning Phase> FIG. 8 is a flowchart showing the flow of an information processing method in a learning phase, which is executed by the information processing apparatus 4 according to Exemplary Embodiment 5 of the present invention. Hereinafter, similar to each of the above-described exemplary embodiments, an example of generating an estimation model using the heartbeat data and sweating data of the subject measured by the wearable terminal 7 as biological signals will be described. Hereinafter, in the information processing method of Exemplary Embodiment 5, the points common to each of the above-described information processing methods will be described as "similarly to Exemplary Embodiment ~" or "similarly to Step S ~", and the same explanation will not be repeated.

[0161] In step S101, the biological signal acquisition unit 401 acquires a biological signal used for generating the estimation model and attribute information of the subject who provides the biological signal. For example, the wearable terminal 7 may transmit the attribute information of the wearer of the wearable terminal 7 registered in advance to the information processing apparatus 4 together with the biological signal.

[0162] In step S102, the specifying unit 404 determines whether the attribute information acquired in step S101 indicates male or female. If the attribute information indicates male, the specifying unit 404 proceeds with the process from A in step S102 to step S103. If the attribute information indicates female, the specifying unit 404 proceeds with the process from B in step S102 to step S108.

[0163] In step S103, the specifying unit 404 specifies, in the same manner as step S32 of Exemplary Embodiment 3, among the biological signals of the male subject, a predetermined standard lunch time zone as the attention time zone.

[0164] In step S104, the extraction unit 405 extracts the first learning feature amount from the biological signal acquired in the lunch time zone, in the same manner as step S33 of Exemplary Embodiment 3.

[0165] In step S105, the stress level calculation unit 403 calculates the stress level, in the same manner as step S34 of each of the above-described exemplary embodiments.

[0166] In step S106, the teacher data generation unit 407 generates first teacher data by associating the stress level with the first learning feature amount in the same manner as in step S35 of the exemplary embodiment 3.

[0167] In step S107, the learning processing unit 408 generates a first estimation model by machine learning using the first teacher data in the same manner as in step S36 of the exemplary embodiment 3.

[0168] In step S108, the determination unit 406 determines whether the subject is in a state of being exposed to acute stress stimulation based on the biological signal in the same manner as in step S52. For example, the determination unit 406 may detect the stress occurrence time zone.

[0169] In step S109, the specifying unit 404 specifies the stress occurrence time zone as the attention time zone in the same manner as in step S53.

[0170] In step S110, the extraction unit 405 extracts the second learning feature amount from the biological signal acquired in the stress occurrence time zone in the same manner as in step S54.

[0171] In step S111, the stress level calculation unit 403 calculates the stress level in the same manner as in steps S34 and S55 of the above-described exemplary embodiments.

[0172] In step S112, the teacher data generation unit 407 generates second teacher data by associating the stress level with the second learning feature amount in the same manner as in step S56.

[0173] In step S113, the learning processing unit 408 generates a second estimation model by machine learning using the second teacher data in the same manner as in step S57.

[0174] <Flow of information processing method in inference phase> FIG. 9 is a flowchart showing the flow of an information processing method in an inference phase executed by the information processing apparatus 4 according to Exemplary Embodiment 5 of the present invention. Hereinafter, in the same manner as in each of the above-described exemplary embodiments, an example will be described in which the stress level of a subject in a month is estimated using one-month worth of heartbeat data and sweating data measured by the wearable terminal 7 as biological signals. Also, hereinafter, in the information processing method of Exemplary Embodiment 5, points common to each of the above-described information processing methods will be described as "similarly to Exemplary Embodiment ~" or "similarly to Step S ~", and the same explanation will not be repeated.

[0175] In step S201, the biological signal acquisition unit 401 acquires a biological signal used for estimating the stress level of the subject and the attribute information of the above-described subject who provides the biological signal. Similarly to step S101, the attribute information may be transmitted from the wearable terminal 7 to the information processing apparatus 4 together with the biological signal.

[0176] In step S202, the specifying unit 404 discriminates the gender indicated by the attribute information, similarly to step S102. When the attribute information indicates male, the specifying unit 404 proceeds with the process from A in step S202 to step S203. When the attribute information indicates female, the specifying unit 404 proceeds with the process from B in step S202 to step S205.

[0177] In step S203, the extraction unit 405 extracts a first estimation feature amount from the biological signal acquired in the attention time zone specified in step S103, that is, the standard lunch time zone.

[0178] In step S204, the estimation unit 409 estimates the stress level of the male subject who is the provider of the biological signal acquired in step S201 using the first estimation model generated in step S107, similarly to step S65. Specifically, the first estimation feature amount extracted in step S203 is input to the first estimation model generated in step S107. Then, the estimation unit 409 stores the output value of the first estimation model in the storage unit 41 as the above-described estimation result data 417 of the male subject.

[0179] In step S205, the determination unit 406 determines, in the same manner as in step S108, whether or not the subject is in a state of being exposed to acute stress stimuli based on the biological signal. For example, the determination unit 406 may detect the stress occurrence time zone.

[0180] In step S206, the specifying unit 404 specifies, in the same manner as in step S109, the stress occurrence time zone as the attention time zone.

[0181] In step S207, the extraction unit 405 extracts the second estimation feature amount from the biological signal acquired in the stress occurrence time zone, in the same manner as in step S64.

[0182] In step S208, the estimation unit 409 estimates, in the same manner as in step S65, the stress level of the female subject who is the provider of the biological signal acquired in step S201 using the second estimation model generated in step S113. Specifically, the second estimation feature amount extracted in step S207 is input to the second estimation model generated in step S113. Then, the estimation unit 409 stores the output value of the second estimation model in the storage unit 41 as the estimation result data 417 of the above-described female subject.

[0183] According to each information processing method according to this exemplary embodiment, for each of men and women, feature amounts can be extracted by narrowing down to the time zone in which the chronic stress tendency becomes prominent. Specifically, for men, feature amounts are extracted from the biological signal acquired in the lunch time zone. Also, for women, feature amounts are extracted from the biological signal acquired in the stress occurrence time zone.

[0184] In this way, by more appropriately narrowing down the time zone in which the chronic stress tendency becomes prominent according to gender, it becomes possible to efficiently construct an estimation model with high estimation accuracy for chronic stress for each of men and women, or to more accurately perform the estimation of chronic stress.

[0185] 〔Modification example〕 In each of the above exemplary embodiments, an example in which the attribute of the subject is gender has been described. However, the attribute may be any attribute related to the time period during which the tendency of chronic stress is significantly manifested in the biological signal, and is not limited to gender. For example, the age group, occupation, etc. of the subject may be used as the attribute of the subject, and the attention time period corresponding to these attributes may be specified. Further, from the biological signal acquired during the attention time period specified in this way, feature amounts corresponding to attributes such as the age group and occupation of the subject are extracted, and an estimation model for each attribute such as the age group and occupation of the subject is constructed using these feature amounts. Then, by inputting the feature amounts corresponding to the attributes such as the age group and occupation of the subject into the estimation model constructed in this way, it becomes possible to estimate the stress level with high accuracy corresponding to the attributes such as the age group and occupation of the subject.

[0186] 〔Example of Realization by Software〕 Some or all of the functions of the information processing apparatuses (1, 4) may be realized by hardware such as an integrated circuit (IC chip) or by software.

[0187] In the latter case, the above-described information processing apparatus is realized by, for example, a computer that executes instructions of a program that is software for realizing each function. An example of such a computer (hereinafter referred to as computer C) is shown in FIG. 10. Computer C includes at least one processor C1 and at least one memory C2. A program P for operating computer C as the above-described information processing apparatus is recorded in memory C2. In computer C, processor C1 reads and executes program P from memory C2, whereby each function of the above-described information processing apparatus is realized.

[0188] As the processor C1, for example, a CPU (Central Processing Unit), GPU (Graphic Processing Unit), DSP (Digital signal Processor), MPU (Micro Processing Unit), FPU (Floating point number Processing Unit), PPU (Physics Processing Unit), a microcontroller, or a combination thereof can be used. As the memory C2, for example, a flash memory, HDD (Hard Disk Drive), SSD (Solid State Drive), or a combination thereof can be used.

[0189] Note that the computer C may further include a RAM (Random Access Memory) for expanding the program P during execution and temporarily storing various data. Also, the computer C may further include a communication interface for transmitting and receiving data to and from other devices. Further, the computer C may further include an input / output interface for connecting input / output devices such as a keyboard, mouse, display, and printer.

[0190] Also, the program P can be recorded on a non-transitory tangible recording medium M readable by the computer C. As such a recording medium M, for example, a tape, disk, card, semiconductor memory, or programmable logic circuit can be used. The computer C can acquire the program P via such a recording medium M. Also, the program P can be transmitted via a transmission medium. As such a transmission medium, for example, a communication network or broadcast wave can be used. The computer C can also acquire the program P via such a transmission medium.

[0191] 〔Supplementary Note 1〕 The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims. For example, embodiments obtained by appropriately combining the technical means disclosed in the above-described embodiments are also included in the technical scope of the present invention.

[0192] [Supplementary Note 2] Some or all of the above-described embodiments may also be described as follows. However, the present invention is not limited to the aspects described below.

[0193] (Supplementary Note 1) Specific means for specifying, as a time zone of interest, a time zone in which a chronic stress tendency is significantly manifested in a biological signal acquired from a subject over a predetermined period; Extraction means for extracting one or more feature quantities used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model from the biological signal acquired in the specified time zone of interest. An information processing apparatus comprising the same.

[0194] According to the above configuration, an effect can be obtained in that it is possible to extract appropriate feature quantities used for machine learning of a stress level estimation model or for estimating a stress level using the estimation model.

[0195] (Supplementary Note 2) The specific means specifies, as the time zone of interest, a time zone in which the biological signal shows significant behavior within one day; The extraction means extracts the feature quantity based on a change in the biological signal at the start or end of the time zone of interest. The information processing apparatus according to Supplementary Note 1.

[0196] According to the above configuration, when it is known in advance that the mode of change in the biological signal due to, for example, the subject transitioning to a specific state has a significant correlation with chronic stress, the following effects are achieved. That is, by extracting a feature quantity based on the above-described change, an effect can be obtained in that the estimation accuracy of chronic stress can be improved.

[0197] (Appendix 3) The information processing apparatus according to Appendix 1 or 2, wherein the specifying means specifies, as a target time period, a predetermined time period before and after the morning time when a predetermined index value of the biological signal peaks based on the circadian rhythm.

[0198] According to the above configuration, a feature amount is extracted from the biological signal in a predetermined time period before and after the morning time when the above index value peaks based on the circadian rhythm. Therefore, an effect is obtained in that it is possible to extract an appropriate feature amount to be used for machine learning of the stress level estimation model or for estimating the stress level using the estimation model.

[0199] (Appendix 4) The specifying means specifies, for a male subject, the standard lunch time period of the subject in a day as the target time period, The information processing apparatus according to any one of Appendices 1 to 3, wherein the extracting means extracts the feature amount from the biological signal acquired in the specified lunch time period for a male subject.

[0200] According to the above configuration, the biological signal for extracting the feature amount can be narrowed down to the biological signal in the lunch time period. As a result, it becomes possible to efficiently construct an estimation model with high estimation accuracy for chronic stress in men, or to more accurately perform the estimation of chronic stress in men.

[0201] (Appendix 5) The specifying means specifies, for a female subject, a time period other than the standard lunch time period of the subject in a day as the target time period, The information processing apparatus according to any one of Appendices 1 to 4, wherein the extracting means extracts the feature amount from the biological signal acquired in the time period other than the lunch time period for a female subject.

[0202] According to the above configuration, at least the lunch time zone can be excluded from the target time zone, and the biological signals to be used for extracting the feature amounts can be narrowed down. As a result, it becomes possible to efficiently construct an estimation model with improved estimation accuracy of chronic stress in women or to improve the estimation accuracy of chronic stress in women.

[0203] (Appendix 6) The information processing apparatus according to any one of Appendices 1 to 5, further comprising determination means for determining whether or not the subject is in a state of being exposed to an acute stress stimulus based on the biological signal. For a female subject, the specifying means specifies, as the target time zone, a stress occurrence time zone in which it is determined that the subject is in a state of being exposed to an acute stress stimulus. For a female subject, the extraction means extracts the feature amount from the biological signal acquired in the specified stress occurrence time zone.

[0204] According to the above configuration, by narrowing down the biological signals for which the feature amounts are to be extracted to the biological signals in the stress occurrence time zone, it becomes possible to efficiently construct an estimation model with high estimation accuracy of chronic stress in women or to more accurately estimate the chronic stress in women.

[0205] (Appendix 7) The information processing apparatus according to any one of Appendices 1 to 6, further comprising determination means for determining whether or not the subject is in a state of being exposed to an acute stress stimulus based on the biological signal. For a male subject, the specifying means specifies, as the target time zone, a time zone other than the stress occurrence time zone in which it is determined that the subject is in a state of being exposed to an acute stress stimulus. For a male subject, the extraction means extracts the feature amount from the biological signal acquired in the time zone other than the stress occurrence time zone.

[0206] According to the above configuration, the feature amount is extracted by narrowing down to the biological signals acquired in time zones other than the stress occurrence time zone. As a result, it becomes possible to efficiently construct an estimation model with high estimation accuracy for male chronic stress, or to more accurately estimate male chronic stress.

[0207] (Appendix 8) At least one processor In the biological signal acquired from the subject over a predetermined period, identify the time zone in which the chronic stress tendency is significantly manifested in the biological signal as the attention time zone, Extract one or more feature amounts used for machine learning of the stress level estimation model or estimation of the stress level using the estimation model from the biological signal acquired in the identified attention time zone. A feature amount extraction method including this.

[0208] According to the above method, the same effect as that of the information processing apparatus of Appendix 1 is obtained.

[0209] (Appendix 9) The at least one processor In the specifying, specify the time zone in which the biological signal shows significant behavior during the day as the attention time zone, In the extracting, extract the feature amount based on the change of the biological signal at the start or end of the attention time zone. The feature amount extraction method according to Appendix 8.

[0210] According to the above method, when it is known in advance that the mode of change of the biological signal due to the subject shifting to a specific state, etc. has a significant correlation with chronic stress, the following effects are obtained. That is, by extracting the feature amount based on the above change, an effect of improving the estimation accuracy of chronic stress can be obtained.

[0211] (Appendix 10) At least one processor A teacher data generation method, including associating the stress level of a subject with one or more feature quantities extracted by the feature quantity extraction method described in Supplementary Note 8 or 9 as correct answer data, and generating teacher data used for the machine learning.

[0212] According to the above method, an effect can be obtained that teacher data capable of efficiently constructing an estimation model with high estimation accuracy for chronic stress can be generated.

[0213] (Supplementary Note 11) At least one processor An estimation model generation method, including generating the estimation model by machine learning using the teacher data generated by the teacher data generation method described in Supplementary Note 10.

[0214] According to the above method, an effect can be obtained that an estimation model with high estimation accuracy for chronic stress can be generated.

[0215] (Supplementary Note 12) At least one processor A stress level estimation method, including estimating the stress level of a subject using the estimation model generated by the estimation model generation method described in Supplementary Note 11.

[0216] According to the above method, an effect can be obtained that the stress level related to chronic stress can be accurately estimated.

[0217] (Supplementary Note 13) A computer In a biological signal acquired from a subject over a predetermined period, a specifying means for specifying a time zone in which a chronic stress tendency is significantly manifested in the biological signal as a target time zone, and An extraction means for extracting one or more feature quantities used for machine learning of a stress level estimation model or estimation of a stress level using the estimation model from the biological signal acquired in the specified target time zone. According to the above configuration, the same effect as the information processing apparatus of Supplementary Note 1 is achieved.

[0218] (Appendix 14) In a biological signal acquired from a subject over a predetermined period, specific means for specifying, as a target time period, a time period in which a chronic stress tendency is significantly manifested in the biological signal; Extraction means for extracting one or more feature quantities (feature quantities for estimation) used for estimating the stress level using an estimation model of the stress level from the biological signal acquired in the specified target time period; Estimation means for estimating the stress level of the subject based on an output value obtained by inputting the one or more extracted feature quantities into the estimation model. An estimation device comprising the above.

[0219] (Appendix 15) At least one processor In a biological signal acquired from a subject over a predetermined period, specifying, as a target time period, a time period in which a chronic stress tendency is significantly manifested in the biological signal; From the biological signal acquired in the specified target time period, extracting one or more feature quantities (feature quantities for estimation) used for estimating the stress level using an estimation model of the stress level; Based on an output value obtained by inputting the one or more extracted feature quantities into the estimation model, estimating the stress level of the subject. A method for estimating the stress level including the above.

[0220] [Appendix Item 3] Some or all of the above-described embodiments can also be expressed as follows.

[0221] An information processing apparatus including at least one processor, the processor performing a specific process of specifying, as a target time period, a time period in which a chronic stress tendency is significantly manifested in a biological signal acquired from a subject over a predetermined period, and an extraction process of extracting one or more feature quantities used for machine learning of an estimation model of the stress level or for estimating the stress level using the estimation model from the biological signal acquired in the specified target time period.

[0222] Note that this information processing apparatus may further include a memory, and a program for causing the processor to execute the specific processing and the extraction processing may be stored in this memory. Further, this program may be recorded on a non-transitory tangible computer-readable recording medium.

Explanation of Signs

[0223] 1 Information processing apparatus 4 Information processing apparatus 7 Wearable terminal 11 Specific part (specific means) 12 Extraction part (extraction means) 404 Specific part (specific means) 405 Extraction part (extraction means) 406 Judgment part (judgment means) 409 Estimation part (estimation means)

Claims

1. Specific means for specifying, as a time period of interest, a time period in which a chronic stress tendency is prominently manifested in a biological signal acquired from a subject over a predetermined period; Extraction means for extracting, from the biological signal acquired in the specified time period of interest, one or more feature quantities used for machine learning of a stress level estimation model or for estimating the stress level using the estimation model; Determination means for determining whether or not the subject is in a state of being exposed to an acute stress stimulus based on the biological signal, the information processing apparatus comprising: The specific means specifies the time period of interest with reference to a stress occurrence time period determined to be a state in which the subject is exposed to an acute stress stimulus.

2. The specific means specifies, as the time period of interest, a time period in which the biological signal shows prominent behavior within one day; The extraction means extracts the feature quantity based on a change in the biological signal at the start or end of the time period of interest. The information processing apparatus according to claim 1.

3. The specific means specifies, as the time period of interest, a predetermined time period before and after a morning time when a predetermined index value of a biological signal peaks based on a circadian rhythm. The information processing apparatus according to claim 1 or 2.

4. For a male subject, the specific means specifies, as the time period of interest, the standard lunch time period of the subject within one day; For a male subject, the extraction means extracts the feature quantity from the biological signal acquired in the specified lunch time period. The information processing apparatus according to any one of claims 1 to 3.

5. For a female subject, the specific means specifies, as the time period of interest, a time period other than the standard lunch time period of the subject within one day; For a female subject, the extraction means extracts the feature quantity from the biological signal acquired in the time period other than the lunch time period. The information processing apparatus according to any one of claims 1 to 4.

6. For a female subject, the specific means specifies, as the time period of interest, a stress occurrence time period determined to be a state in which the subject is exposed to an acute stress stimulus; For a female subject, the extraction means extracts the feature quantity from the biological signal acquired in the specified stress occurrence time period. The information processing apparatus according to any one of claims 1 to 5. **Claim 7**: For a male subject, a time period other than the stress occurrence time period determined to be a state where the subject is exposed to acute stress stimulation is specified as the target time period, and the extraction means extracts the feature amount from the biological signal acquired in the time period other than the stress occurrence time period for the male subject. The information processing apparatus according to any one of claims 1 to 6. **Claim 8** At least one processor specifies, as a target time period, a time period in which a chronic stress tendency is significantly manifested in the biological signal acquired from a subject over a predetermined period, extracts one or more feature amounts used for machine learning of a stress degree estimation model or estimation of a stress degree using the estimation model from the biological signal acquired in the specified target time period, and determines whether or not the subject is in a state of being exposed to acute stress stimulation based on the biological signal, wherein, in the specifying, the target time period is specified with reference to the stress occurrence time period determined to be a state where the subject is exposed to acute stress stimulation. A feature amount extraction method. **Claim 9** The at least one processor in the specifying, specifies, as the target time period, a time period in which the biological signal shows significant behavior within one day, and in the extracting, extracts the feature amount based on a change in the biological signal at the start or end of the target time period. The feature amount extraction method according to claim 8. **Claim 10** At least one processor includes associating the stress degree of a subject as correct answer data with one or more feature amounts extracted by the feature amount extraction method according to claim 8 or 9, and generating teacher data used for the machine learning. A teacher data generation method. **Claim 11** At least one processor includes generating the estimation model by machine learning using the teacher data generated by the teacher data generation method according to claim 10. An estimation model generation method. **Claim 12** At least one processor includes estimating the stress degree of a subject using the estimation model generated by the estimation model generation method according to claim 11. A stress degree estimation method. **Claim 13** A computer In a biological signal acquired from a subject over a predetermined period, specific means for specifying, as a target time period, a time period in which a chronic stress tendency is significantly manifested in the biological signal, and extraction means for extracting one or more feature quantities used for machine learning of a stress degree estimation model or estimation of a stress degree using the estimation model from the biological signal acquired in the specified target time period, functioning as determination means for determining whether or not the subject is in a state of being exposed to an acute stress stimulus based on the biological signal, wherein the specific means specifies the target time period with reference to a stress occurrence time period determined to be a state in which the subject is exposed to an acute stress stimulus, a feature quantity extraction program.

Images