JP7297236B2 - Estimation processing device and estimation model construction device - Google Patents

Description

The present invention relates to an estimation processing device and an estimation model building device for estimating user's attention function.

A decline in attentional function has become an international problem because it leads to a decline in labor productivity. Therefore, it is considered to quantitatively measure human attention function. For example, it is considered to conduct cognitive function tests such as PVT and Go-NoGo, but this test requires several minutes, and in order to capture changes over time, it is necessary to perform multiple tests a day. It was necessary to conduct a test. Therefore, the burden on the user is heavy. Therefore, a system that automatically judges whether the attentional function is high or low would be useful.

Non-Patent Document 1 describes estimating human attention function (response speed of PVT test) using a machine learning model based on feature values such as the number of times the smartphone screen is lit, lighting time, and sleep time. .

Cognitive Rhythms: Unobtrusive and Continuous Sensing of AlertnessUsing a Mobile PhoneSaeedAbdullah, Elizabeth Murnane, Mark Matthews, Matthew Kay, Julie Kientz, GeriGay, Tanzeem Choudhury, Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing. Retrieved February 6, 2016] <http://pac.cs.cornell.edu/pubs/Ubicomp2016_CognitiveRhythms.pdf>

However, feature values related to how to use a smartphone, such as the lighting of the screen, cannot be obtained without using a smartphone. It is assumed that it will not be obtained.

Therefore, the present invention provides an estimation processing device and an estimation model construction device that can collect and judge information that can passively measure the attention function even when the user is not using a terminal such as a smartphone. intended to provide

In order to solve the above-described problems, an estimation processing device of the present invention estimates a user's attention function using feature values generated based on sensor values of a motion sensor of an information processing device owned by the user. an attentional function estimator.

According to this invention, it is possible to estimate the attention function of the user from the movement of the communication terminal held by the user without detecting the operation of the user. Since the user's operation is not required, the attention function can be estimated without making the user aware of it. Note that "a user possesses a communication terminal" is not limited to a form in which the user holds the communication terminal by holding it in his/her hand or on his/her body. It includes a state in which it is operable, for example, a state in which it is placed in a predetermined place within reach of the user.

According to the present invention, the attention function of the user can be estimated from the movement of the communication terminal held by the user without detecting the user's operation.

2 is a block diagram showing functional configurations of an estimation model execution device 100 and a communication terminal 200 of this embodiment; FIG. FIG. 4 is a schematic diagram showing generation processing when generating an attention function score estimation model; It is a table that associates dates, attention function scores, and feature quantities such as sensor value logs of each sensor value for each user ID. 4 is a diagram showing a sensor value table of each sensor value collected by the communication terminal log collection unit 101; FIG. It is a figure which shows the sensor value table in another example. Specific examples of a screen on/off history table and an earphone connection history table are shown. Specific examples of a positioning history table and a stay point table are shown. Specific examples of an application activation history table and an application information table 101b are shown. FIG. 11 is a flow chart showing a process of constructing an attention function score estimation model; FIG. FIG. 10 is a flow chart in which the estimation model execution device 100 performs a process of estimating a user's attention function using an estimation model; FIG. 9 is a flowchart showing processing for calculating a feature amount from an application activation log; FIG. 10 is a diagram showing changes in screen history for showing an example of a terminal log showing operations from turning on the screen of the communication terminal 200 to starting an application; It is a specific example of an operation log table that stores operation logs. 9 is a flow chart showing processing when calculating a feature amount from an operation log; FIG. 10 is a diagram showing an example of a management table for generating an exclusion list and its exclusion list; 2 is a diagram illustrating an example of hardware configurations of estimation model execution device 100 and communication terminal 200 according to an embodiment of the present disclosure; FIG.

Embodiments of the present invention will be described with reference to the accompanying drawings. Where possible, the same parts are denoted by the same reference numerals, and duplicate descriptions are omitted.

FIG. 1 is a block diagram showing functional configurations of an estimation model execution device 100 and a communication terminal 200 of this embodiment. As shown in FIG. 1, the estimation model execution device 100 can acquire the operation log of each motion sensor provided in the communication terminal. The estimation model execution device 100 includes a communication terminal log collection unit 101 (sensor value acquisition unit, program management unit), a log table 101a (program management unit), an application information table 101b, and a usage feature amount calculation unit 102 (feature amount conversion unit). , attention function estimation unit 103 (attention function estimation unit), estimation result display unit 104 (attention function presentation unit), estimation model construction unit 105 (estimation model construction unit), estimation model storage unit 106, exclusion list 107, and exclusion list It is configured including a generation unit 108 . Further, the communication terminal 200 includes an acceleration sensor 201, an angular velocity sensor 202, an inclination sensor 203, a screen on/off sensor 204, an application use sensor 205, a GPS 206, a communication section 207, a control section 208, and a program 209. . Each configuration will be described below. These configurations are examples, and not all of them are essential configurations. For example, the exclusion list 107 and the exclusion list generator 108 may be omitted.

The communication terminal log collection unit 101 collects sensor values such as the acceleration sensor 201, the angular velocity sensor 202, the tilt sensor 203, and the screen on/off sensor 204 provided in the communication terminal 200, the program activation state of the application use sensor 205, This is a part that collects terminal log information such as screen on/off and position information of the communication terminal 200 . As sensor values, the number of WIFI access points, WIFI access rate, and the like are also collected. In this case, the communication unit of communication terminal 200 corresponds to the motion sensor. The communication terminal log collection unit 101 can communicate with the communication terminal 200 by wire or wirelessly.

The log table 101a includes a sensor value table that stores sensor values such as an acceleration sensor and an angular velocity sensor, a position information table, an application activation history table, and a terminal log table. The communication terminal log collection unit 101 sorts and stores the collected sensor values and terminal log information (positioning information, application activation information, screen on/off) into the respective tables.

The application information table 101b is a table that associates and stores package names, category numbers, and tags of applications. A package name is an ID for specifying an application and is an application ID. The category number is a number indicating the category of the application. The category number is, for example, a number indicating a category such as social/transportation/work optimization. A tag is additional information, and in this embodiment, information indicating that it is a communication application such as CALL (telephone) or SMS is added.

The usage feature amount calculation unit 102 detects each sensor (acceleration sensor 201, angular velocity sensor 202, tilt sensor 203, screen on/off sensor 204, application usage sensor 205, and other communication status (WIFI access status) in the log table 101a. It is a part that calculates feature amounts based on sensor values and terminal log information. For example, the feature values are the average value of sensor values, standard deviation, maximum value, minimum value, maximum/minimum difference, correlation coefficient of axis pair, total moving distance, number of times screen is turned on, number of times application is started, time period WIFI connection turn-on rate. , is the number of observed WIFI access points. In addition, in the case of an acceleration sensor, the Pearson correlation coefficient of the value series of two of the three axes or the square root of the sum of squares of the three axes (signal power) is calculated as a feature amount. In addition, angular velocity, tilt, and acceleration (gravitational component removed) may also be calculated in the same manner.

The usage feature quantity calculation unit 102 can calculate the feature quantity by designating the user ID and the date and acquiring the sensor values and terminal log information for each predetermined time period from each table. For example, the utilization feature amount calculation unit 102 calculates the average sensor value of the acceleration sensor every hour (every hour), the average sensor value of the acceleration sensor every 6 hours, the sensor value of the acceleration sensor all day (every 24 hours) Calculate the average of the values.

The attention function estimation unit 103 is a part that estimates the attention function of the user possessing the communication terminal 200 based on the usage feature amount and the estimation model stored in the estimation model storage unit 106 . The "attention function" in this embodiment means the user's concentration and attention, or a similar state of the user. The estimation model is a model learned in advance based on the usage feature amount and the Go-NoGo test results. The model construction processing will be described later.

In this embodiment, it is important that the attentional function estimation unit 103 estimates the attentional function score by inputting the sensor values into the estimation model. A sensor value in the present embodiment is a sensor value passively obtained in the communication terminal 200, and is a sensor value obtained without necessarily being operated by the user. By using such sensor values, the attention function of the user can be estimated even if the user does not operate the terminal.

The estimation result display section 104 is a section that displays the attention function of the user who owns the communication terminal 200 . It should be noted that the present invention is not limited to the display, and may be any function as long as it presents a user's attention function. In addition, the estimation result display unit 104 may have a function of presenting (notifying) the communication terminal 200 with information indicating the user's attention function instead of the display.

The estimation model construction unit 105 uses the feature amount based on the sensor value of the communication terminal 200 and the terminal log information (application activation history, screen on/off log, position information, etc.) for a predetermined time unit as an explanatory variable, and for the predetermined time unit In the form of correspondence, the communication terminal 200 builds an estimation model by performing machine learning using the result of the Go-NoGo test obtained from the user of the communication terminal 200 as a teacher signal.

For example, the estimation model building unit 105 uses the feature amount calculated based on the information sorted by predetermined time units from among the sensor values and the terminal log information (application activation history, screen on/off, position measurement, etc.) as the usage feature. Acquired from the amount calculation unit 102 . On the other hand, the estimation model construction unit 105 acquires the user's Go-NoGo test result as an index of the user's attention function. This test result may be information obtained by matching the time period with the feature values based on the sensor values and the terminal log information in the predetermined time period, but it is not necessarily limited to this. For example, the feature value from 0:00 to 6:00 on the day may be associated with the attention function score obtained at 12:00 on the day.

For example, the estimation model construction unit 105 obtains the feature amount in units of 6 hours, and the user takes a Go-NoGo test every 6 hours, and the estimation model construction unit 105 acquires the result.

Then, the estimation model construction unit 105 performs learning processing using the feature amount as an explanatory variable and the Go-NoGo test result as a teacher signal. Existing methods are applied as machine learning methods. For example, SVM (Support Vector Machine), Random Forest, K nearest neighbor method, neural network, etc. can be used.

The construction processing will be described with reference to FIG. FIG. 2 is a schematic diagram showing generation processing when generating an attention function score estimation model based on sensor values of each sensor, terminal log information, and Go-NoGo test results. The input S11 indicates the log information of the sensor value of each sensor of the communication terminal 200, the application activation history, the terminal log, and the positioning history (GPS). Based on the log information such as these sensor values, the feature amount is calculated and the feature amount shown in the output S12 is calculated.

Input S13, on the other hand, indicates that the user is taking a test in the Go-NoGo test app. The Go-NoGo test is performed by the user operating a terminal such as communication terminal 200 and activating an application for the Go-NoGo test. A test result is information obtained by operating the application. An output S14 is obtained by performing a scoring process based on the test results. Here, the correct answer rate, efficiency, average Go reaction time, NoGo success rate, etc. of the Go-NoGo test are calculated as the attention function score. Efficiency indicates the difference between the actually measured NoGo success rate and the assumed NoGo success rate calculated from the Go reaction speed at that time.

Then, the estimation model construction unit 105 can construct an attention function score estimation model by performing learning processing of the attention function score estimation model using the feature quantity indicated by the output S12 as an explanatory variable and the attention function score as an objective variable. .

FIG. 3 shows, for each user ID, the date, the attention function score (see output S14 in FIG. 2), the sensor value of each sensor value, and the feature amount of terminal log information (application activation history, screen on/off, etc.) (Fig. 2 (see output S12 of ). The attention function score is indicated by one or a plurality of pieces of information such as the Go-No-Go accuracy rate shown in S14 in FIG. For example, the Go-NoGo accuracy rate is used as this attention function score.

The estimation model construction unit 105 generates the table shown in FIG. 3 based on the feature amount calculated by the usage feature amount calculation unit 102 and the attention function score obtained based on the Go-NoGo test result, An estimation model building process can be performed. In order to generate an attention function score estimation model for each user, the estimation model construction unit 105 extracts feature amounts for each user and constructs them for each user. In addition, when the estimation model is not constructed for each user, the estimation model may be constructed using all the data.

The exclusion list generation unit 108 is a part that generates an exclusion list 107 indicating applications that are not used for attention function estimation, based on the application activation history table. Details will be described later.

Next, functional blocks of communication terminal 200 will be described. Acceleration sensor 201 is a sensor that detects acceleration applied to communication terminal 200 .

Angular velocity sensor 202 is a sensor that detects an angular velocity applied to communication terminal 200 .

The tilt sensor 203 is a sensor that detects the tilt of the communication terminal 200 to determine which posture it is in. FIG.

The screen on/off sensor 204 is a sensor that detects whether the display unit (not shown) of the communication terminal 200 is on or off. For example, the screen on/off sensor 204 detects whether the display unit is in a sleep state and nothing is displayed, or information such as a home screen is displayed.

The application usage sensor 205 is a sensor that detects which application is being used by the user.

GPS 206 is a positioning unit for positioning the position information of communication terminal 200 .

The communication unit 207 is a part for communication connection with the estimation model execution device 100 and other networks.

Control section 208 is a section for controlling communication terminal 200 .

The program 209 is a program for the communication terminal 200 to execute the communication function and other predetermined functions.

FIG. 4 is a diagram showing a sensor value table of each sensor value collected by the communication terminal log collection unit 101 in the estimation model execution device 100. As shown in FIG. These sensor values are information based on information acquired by each sensor of the communication terminal 200 described above.

FIG. 4(a) shows a specific example of an acceleration sensor table (not shown). FIG. 4B is a table of sensor values based on the angular velocity sensor, tilt sensor, and acceleration sensor (after removing the gravity component).

In this embodiment, the communication terminal 200 further includes an air pressure sensor, a battery log sensor, an illuminance sensor, etc., and the estimation model execution device 100 collects these information as sensor values of the motion sensor and stores them in the sensor value table. You can remember. Specific examples are shown in FIGS. 5(a) to 5(c).

FIG. 6 shows specific examples of a screen on/off history table (FIG. 6A) and an earphone connection history table (FIG. 6B). These tables are collectively referred to as a terminal log table in FIG. In addition, the terminal log table may store an operation log table (see FIG. 13), which will be described later. Although communication terminal 200 of the present embodiment does not explicitly include an earphone connection sensor, it may include such a sensor.

FIG. 7 shows specific examples of the positioning history table and the staying point table. These tables are collectively referred to as a position information table in FIG. The positioning history table stores position information obtained by GPS 206 . The stay point table is information obtained by clustering which areas the user frequently stays in based on the position of communication terminal 200 . In the figure, this table associates a stay point ID indicating a stay area ID, a stay point cluster ID, a start date and time of stay at a stay point, an end date and time, and latitude and longitude.

FIG. 8 shows specific examples of the application activation history table and the application information table 101b. Since the application information stored in the application information table 101b is not log information, it is not collected and is prepared in advance. The application launch history table associates launch dates and times with their package names (application IDs). The application information table associates package names, category numbers indicating application categories, and tags indicating communication functions. A tag is information added later, and is added here as information indicating a communication function.

Next, the operation of estimation model execution device 100 will be described. FIG. 9 is a flowchart showing a process of constructing an attention function score estimation model. The communication terminal log collection unit 101 collects sensor values and terminal log information from one or more communication terminals 200 possessed by one or more users (S101). The used feature quantity calculation unit 102 calculates a feature quantity based on the sensor values and the terminal log information (S102). The estimation model construction unit 105 inputs the feature amount of the communication terminal 200 as an explanatory variable (S103).

On the other hand, each user takes a test based on the Go-NoGo application, and the estimation model construction unit 105 acquires the test result (S105). Then, the estimation model construction unit 105 inputs the index of the attention function based on the Go-NoGo test result as the objective variable (S106).

Then, the estimated model construction unit 105 performs attention function score estimation model construction processing (S104), and stores the attention function score estimation model in the estimation model storage unit 106 (S107).

By this processing, an attention function score estimation model can be constructed.

FIG. 10 is a flow chart in which estimation model execution device 100 performs estimation processing of a user's attention function using an estimation model. The communication terminal log collection unit 101 collects sensor values and log information from one communication terminal 200 and stores them in each table (S201). The utilization feature amount calculation unit 102 acquires the sensor values and the terminal log information in the designated time unit, and calculates the feature amount based on the sensor values and the terminal log information (S202). The attention function estimation unit 103 applies the calculated feature amount to the attention function score estimation model (S203), and estimates the attention function score (S204). This makes it possible to estimate the attention function of the user who owns one communication terminal 200 .

Next, a method of calculating a feature amount generated from an application activation log will be described. FIG. 11 is a flow chart showing processing for calculating the feature amount. This feature amount generation processing is executed when estimating the attentional function score and constructing an attentional function score estimation model.

The usage feature amount calculation unit 102 acquires the application activation log of the communication terminal 200 of the user to be calculated for a predetermined date and a predetermined period from the log table 101a (S301).

The usage feature amount calculation unit 102 refers to the application information table 101b and extracts an application activation log for each category number (S302).

Then, the usage feature amount calculation unit 102 checks whether or not each category number corresponds to an application described in the exclusion list, and furthermore, the tag added to the application activation log is CALL (call function) or SMS. (S304-S305).

Then, the usage feature amount calculation unit 102 excludes the usage of an application whose tag is CALL or SMS and whose activation time is equal to or less than a threshold value (S306), is calculated (S307).

In addition, the usage feature amount calculation unit 102 calculates the number of times of activation, the activation time, and the total/average of the activation time for applications other than CALL or SMS (S308).

The utilization feature amount calculation unit 102 uses these calculated results as feature amounts (S309), checks whether or not there is an unprocessed category number, and repeats for all applications until there is no unprocessed category number. process (S303).

In the above processing, the usage feature amount calculation unit 102 acquires the application activation log in a plurality of time units such as every hour, every 6 hours, and every day. Further, the utilization feature amount calculation unit 102 extracts the category ID from the package name based on the application activation history table and the application information table 101b shown in FIG. 7, and calculates the activation time and the like for each category ID. For apps with a CALL (call) tag in the app information table, the number of outgoing calls, the number of incoming calls, the total number, the total call time (outgoing/incoming/answering), and the average call Time (outgoing/incoming/response) may be calculated. In addition, the history of less than 5 seconds may be removed as preprocessing.

Similarly, for SMS tags, the number of SMS activations, the total activation time, and the average activation time may be calculated for each tag. In addition, the history of less than 3 seconds may be removed as preprocessing.

Next, a feature amount generation process from terminal log information based on application operations will be described. FIG. 12 is a diagram showing changes in screen history for showing an example of an operation log showing operations from turning on the screen of communication terminal 200 to starting an application.

Screen G1 is a screen after a screen-on event (the user performs an operation to turn on the screen) has occurred from the standby state of communication terminal 200 at time t1.

The screen G2 is a screen showing a state of receiving an operation for unlocking after that.

Screen G3 shows the home screen after the unlock event (completion of password input) is completed at time t2. The time Ton_unlock is the time from when the screen is displayed until the screen is unlocked. Time Ton_unlock=Time t2-Time t1.

Screen G4 shows the screen of the application that the user tapped on the application icon at time t3 and started. Time Ths_tap indicates the time during which the user has operated from unlocking to application activation, and time Ths_tap=time t3−time t2.

Time Ton_unlock and time Ths_tap are feature amounts based on the operation log.

In this way, by using the on/off of the screen of the communication terminal 200 and the application activation log, the time from the screen on to the application activation is calculated, and the activation time is totaled, thereby calculating the efficiency of the user operation. can. In other words, if the user can operate efficiently, it can be determined that the user has a warning function.

FIG. 13 is a specific example of an operation log table (terminal log table) that stores the above operation logs. As shown in FIG. 13, time, event, and value are described for each user ID (UID). The event indicates a user operation event, as described above. Value indicates an application.

A time is attached to each event, and the time Ton_unlock and the time Ths_tap can be calculated based on this time.

FIG. 14 is a flow chart showing processing when calculating a feature amount from an operation log in the estimation model execution device 100. As shown in FIG.

The usage feature amount calculation unit 102 acquires the operation log of the calculation target user for a predetermined period from the operation log table (S401). Then, the usage feature quantity calculation unit 102 extracts a pair of Screen On and App Launch (S402).

The usage feature amount calculation unit 102 calculates the required time from the TimeStamp (time) of all the extracted pairs of Screen On and App Launch (S403).

The usage feature amount calculation unit 102 calculates the average value of the required time (S404), and calculates the efficiency score based on the average value (S405). The efficiency score can be calculated by substituting the average value for X in any of the following formulas.

Y=AX+B
Y=exp(AX+B)
Y=log(AX+B)
Also, the formula is not limited to this, and any formula in which Y monotonically increases with respect to X may be used. Also, although A and B are arbitrarily set values in advance, they are not limited to this, and a questionnaire may be given to users to evaluate efficiency on a scale of 5, and coefficients may be set by machine learning. For this questionnaire, for example, a method based on a check table in which the user himself/herself can input the efficiency evaluation of terminal operation such as a smartphone can be considered.

Note that this efficiency score may be used as a feature amount, or the average value of the time Ton_unlock and the time Ths_tap, which are the source information of the efficiency score, may be used as a feature amount.

Next, generation of the exclusion list explained in FIG. 11 will be explained. The exclusion list is an application list for excluding applications, such as home screens and launchers, that were not started by the user intentionally. By using this, it is possible to generate a history of applications that the user has attempted to operate.

FIG. 15(a) shows an exclusion list that associates the average number of daily activations, number of days of activation, TF (Term Frequency), IDF (Inverse Document Frequency), and TF/IDF for each application (aggregated by category number). shows a management table for Here, TF indicates the average number of activations of each application in one day, and IDF indicates Log (number of days in aggregation unit/number of days of activation+1) in each application. The number of days in aggregation units is information for determining the degree of importance of the application. For example, if the aggregation unit is 30 days, it will be Log(30/startup days+1). The number of days is set arbitrarily. Note that the TF is not limited to the average number of activations, and may be information indicating the frequency of use in a day. For example, the average activation time, the total number of activations in a predetermined period, or the like may be used. Also, the IDF is not limited to the above formula, as long as it is a formula that indicates how much the application is used in the number of days in the aggregate unit.

By calculating the TF/IDF score for each category number, it is possible to judge important applications by weighting applications that are frequently used but not used every day.

Taking the management table of FIG. 15(a) as an example, the exclusion list generation unit 108 obtains TF/IDF based on the application activation history table, and excludes applications with category numbers whose TF/IDF is equal to or less than the threshold. Subscribe to the list. FIG. 15(b) is a specific example of the exclusion list, in which applications with category numbers for which exclusion is set to true are defined as the exclusion list. The threshold may be a predetermined numerical value, but may be a numerical value obtained by taking the average value or the median value of TF/IDF for all applications.

Next, the effects of the estimation model execution device 100 of this embodiment will be described. This estimation model execution device 100 is an estimation processing device for estimating the attention function of the user, and is generated based on the sensor values of each sensor (motion sensor) of the communication terminal 200, which is the information processing device possessed by the user. An attention function estimating unit 103 is provided for estimating the user's attention function using the obtained feature amount.

In the above-described embodiment, the estimation model execution device 100 includes the communication terminal log collection unit 101 that acquires sensor values (or feature amounts) transmitted from the communication terminal 200, and the use feature amount that converts the acquired sensor values into feature amounts. Although the calculation unit 102 and the estimation model execution device 100 generate feature quantities, the present invention is not limited to this. The communication terminal 200 may have a configuration equivalent to the usage feature quantity calculation unit 102, generate a feature quantity from the sensor values acquired by itself, and the estimation model execution device 100 may receive it. The same applies to the following description.

With this configuration, the attention function of the user can be estimated from the sensor values of the sensors of communication terminal 200 . Therefore, the attention function of the user can be estimated from the movement of the communication terminal 200 held by the user without detecting the user's operation. Since the user's operation is not required, the attention function can be estimated without making the user aware of it.

In addition, the estimation model execution device 100 uses an attention function score estimation model constructed using sensor values and attention function indices (for example, Go-NoGo test results) of each of a plurality of users. Judge the function.

This attention function score estimation model may be built by itself using the estimation model building unit 105, or may use a model built in another server (which has a function corresponding to the estimation model building unit 105). .

With this configuration, the user's attention function can be estimated using the learning model. Therefore, the estimation process can be realized with a simple process.

Further, in the estimation model execution device 100, the usage feature amount calculation unit 102 generates feature amounts by dividing the sensor values acquired by the communication terminal log collection unit 101 into predetermined time units.

With this configuration, it is possible to estimate the attention function of the user on a predetermined time basis. By setting both the fine time unit and the rough time unit as the predetermined time unit, it is possible to accurately estimate the attention function of the user.

Communication terminal log collection unit 101 also acquires logs of one or more motion sensors installed in communication terminal 200 . In particular, by using a plurality of motion sensors, it is possible to obtain various feature amounts, and it is possible to accurately estimate the attention function of the user.

In addition, the estimation model execution device 100 of the present embodiment further includes a program management unit that acquires the activation status of one or more programs that can be executed on the communication terminal 200 as an operation log (including an application activation log). This program management unit can be realized by the communication terminal log collection unit 101 and the terminal log table. The utilization feature quantity calculation unit 102 converts the operation log into a feature quantity in addition to the sensor value of the motion sensor. The activation state of the program indicates the activation time during which the program is running or the activation start time from a predetermined state until the program is activated.

With this configuration, it is possible to estimate the attentional function based not only on the motion sensor, but also on the basis of the user's activation operation for the application. Therefore, the estimation accuracy can be improved.

In addition, the usage feature amount calculation unit 102 excludes, from the feature amount calculation targets, usage logs for programs classified into categories defined by a predetermined method. Here, programs of categories described in the exclusion list 107 are excluded.

With this configuration, by excluding usage logs for programs that do not affect the attention function estimation of the user, the estimation accuracy is improved.

For example, as an exclusion method, categories to be excluded are determined based on TF-IDF. In the present embodiment, TF is the average number of activations per day for each application, and IDF is Log (number of days in aggregate unit/number of days of activation+1) in each application. However, as long as the TF is an index representing the frequency of application activation, and the IDF is an index representing the degree to which the application is commonly used on different days or on a per-person basis, the present invention is not limited to the above.

Further, the activation state of the program includes activation start time from a predetermined state to activation of the program. In addition, the program activation state includes the activation count, activation time, and communication time of the program used for communication among the one or more activated programs.

Since the user's attention function depends on the agility of the user's operation or the frequency with which the user is trying to make a call, its estimation accuracy can be improved.

Although the estimation model execution device 100 is configured to acquire sensor values and the like from the communication terminal 200, it is not limited to this. Communication terminal 200 may have the main functions of estimation model execution device 100 (estimation model storage unit 106 and attention function estimation unit 103), and communication terminal 200 may perform estimation processing using the estimation model. In this case, the estimated model may be obtained from estimated model execution device 100 or another model building device (not shown), or communication terminal 200 may build the estimated model by itself.

The estimation model execution device 100 described above may function as an estimation model construction device. When functioning as an estimation model construction device, estimation model execution device 100 includes at least estimation model construction section 105 and estimation model storage section 106 . Configuration requirements for performing attentional function estimation processing may be excluded.

The block diagrams used in the description of the above embodiments show blocks for each function. These functional blocks (components) are implemented by any combination of at least one of hardware and software. Also, the method of realizing each functional block is not particularly limited. That is, each functional block may be implemented using one device physically or logically coupled, or directly or indirectly using two or more physically or logically separated devices (e.g. , wired, wireless, etc.) and may be implemented using these multiple devices. A functional block may be implemented by combining software in the one device or the plurality of devices.

Functions include judging, determining, determining, calculating, calculating, processing, deriving, examining, searching, checking, receiving, transmitting, outputting, accessing, resolving, selecting, choosing, establishing, comparing, assuming, expecting, assuming, Broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, assigning, etc. can't For example, a functional block (component) that makes transmission work is called a transmitting unit or a transmitter. In either case, as described above, the implementation method is not particularly limited.

For example, estimation model execution device 100 and communication terminal 200 according to an embodiment of the present disclosure may function as a computer that performs processing of the estimation model execution method and the estimation model construction method of the present disclosure. FIG. 16 is a diagram illustrating an example of a hardware configuration of estimation model execution device 100 and communication terminal 200 according to an embodiment of the present disclosure. The estimation model execution device 100 and the communication terminal 200 described above are physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like. good too.

Note that in the following description, the term "apparatus" can be read as a circuit, device, unit, or the like. The hardware configuration of estimation model execution device 100 and communication terminal 200 may be configured to include one or more of each device shown in the figure, or may be configured without some of the devices.

Each function of estimation model execution device 100 and communication terminal 200 is performed by loading predetermined software (program) onto hardware such as processor 1001 and memory 1002 so that processor 1001 performs calculations and communication by communication device 1004 is performed. and at least one of reading and writing data in the memory 1002 and the storage 1003 .

The processor 1001, for example, operates an operating system to control the entire computer. The processor 1001 may be configured by a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic device, registers, and the like. For example, the utilization feature amount calculation unit 102 , the attention function estimation unit 103 , the estimation model construction unit 105 and the like described above may be implemented by the processor 1001 .

The processor 1001 also reads programs (program codes), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 to the memory 1002, and executes various processes according to these. As the program, a program that causes a computer to execute at least part of the operations described in the above embodiments is used. For example, the usage feature amount calculation unit 102 and the like may be implemented by a control program stored in the memory 1002 and operating on the processor 1001, and other functional blocks may be implemented in the same way. Although it has been explained that the above-described various processes are executed by one processor 1001, they may be executed simultaneously or sequentially by two or more processors 1001. FIG. Processor 1001 may be implemented by one or more chips. Note that the program may be transmitted from a network via an electric communication line.

The memory 1002 is a computer-readable recording medium, and is composed of at least one of, for example, ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), and RAM (Random Access Memory). may be The memory 1002 may also be called a register, cache, main memory (main storage device), or the like. The memory 1002 can store executable programs (program codes), software modules, etc. for implementing an estimation model execution method and an estimation model construction method according to an embodiment of the present disclosure.

The storage 1003 is a computer-readable recording medium, for example, an optical disc such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disc, a magneto-optical disc (for example, a compact disc, a digital versatile disc, a Blu-ray disk), smart card, flash memory (eg, card, stick, key drive), floppy disk, magnetic strip, and/or the like. Storage 1003 may also be called an auxiliary storage device. The storage medium described above may be, for example, a database, server, or other suitable medium including at least one of memory 1002 and storage 1003 .

The communication device 1004 is hardware (transmitting/receiving device) for communicating between computers via at least one of a wired network and a wireless network, and is also called a network device, a network controller, a network card, a communication module, or the like. The communication device 1004 includes a high-frequency switch, a duplexer, a filter, a frequency synthesizer, and the like, for example, in order to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). may consist of For example, the communication terminal log collection unit 101 described above may be implemented by the communication device 1004 . Also, an implementation may be made in which the transmission and reception functions are physically or logically separated.

The input device 1005 is an input device (for example, keyboard, mouse, microphone, switch, button, sensor, etc.) that receives input from the outside. The output device 1006 is an output device (eg, display, speaker, LED lamp, etc.) that outputs to the outside. Note that the input device 1005 and the output device 1006 may be integrated (for example, a touch panel).

Devices such as the processor 1001 and the memory 1002 are connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between devices.

In addition, the estimation model execution device 100 and the communication terminal 200 are microprocessors, digital signal processors (DSPs), ASICs (Application Specific Integrated Circuits), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), and the like. hardware, and part or all of each functional block may be realized by the hardware. For example, processor 1001 may be implemented using at least one of these pieces of hardware.

Notification of information is not limited to the aspects/embodiments described in this disclosure, and may be performed using other methods. For example, notification of information includes physical layer signaling (e.g., DCI (Downlink Control Information), UCI (Uplink Control Information)), higher layer signaling (e.g., RRC (Radio Resource Control) signaling, MAC (Medium Access Control) signaling, It may be implemented by broadcast information (MIB (Master Information Block), SIB (System Information Block)), other signals, or a combination thereof. RRC signaling may also be called an RRC message, and may be, for example, an RRC connection setup message, an RRC connection reconfiguration message, or the like.

Each aspect/embodiment described in the present disclosure includes LTE (Long Term Evolution), LTE-A (LTE-Advanced), SUPER 3G, IMT-Advanced, 4G (4th generation mobile communication system), 5G (5th generation mobile communication system), FRA (Future Radio Access), NR (new Radio), W-CDMA (registered trademark), GSM (registered trademark), CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi (registered trademark) )), IEEE 802.16 (WiMAX®), IEEE 802.20, UWB (Ultra-WideBand), Bluetooth®, other suitable systems, and extended It may be applied to at least one of the next generation systems. Also, a plurality of systems may be applied in combination (for example, a combination of at least one of LTE and LTE-A and 5G, etc.).

The processing procedures, sequences, flowcharts, etc. of each aspect/embodiment described in this disclosure may be rearranged as long as there is no contradiction. For example, the methods described in this disclosure present elements of the various steps using a sample order, and are not limited to the specific order presented.

Information and the like may be input and output via multiple network nodes that may be output from higher layers (or lower layers) to lower layers (or higher layers).

Input/output information and the like may be stored in a specific location (for example, memory), or may be managed using a management table. Input/output information and the like can be overwritten, updated, or appended. The output information and the like may be deleted. The entered information and the like may be transmitted to another device.

The determination may be made by a value represented by one bit (0 or 1), by a true/false value (Boolean: true or false), or by numerical comparison (for example, a predetermined value).

Each aspect/embodiment described in the present disclosure may be used alone, may be used in combination, or may be used by switching according to execution. In addition, the notification of predetermined information (for example, notification of “being X”) is not limited to being performed explicitly, but may be performed implicitly (for example, not notifying the predetermined information). good too.

Although the present disclosure has been described in detail above, it should be apparent to those skilled in the art that the present disclosure is not limited to the embodiments described in this disclosure. The present disclosure can be practiced with modifications and variations without departing from the spirit and scope of the present disclosure as defined by the claims. Accordingly, the description of the present disclosure is for illustrative purposes and is not meant to be limiting in any way.

Software, whether referred to as software, firmware, middleware, microcode, hardware description language or otherwise, includes instructions, instruction sets, code, code segments, program code, programs, subprograms, and software modules. , applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, and the like.

Software, instructions, information, etc. may also be sent and received over a transmission medium. For example, the software uses wired technology (coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), etc.) and/or wireless technology (infrared, microwave, etc.) to create websites, Wired and/or wireless technologies are included within the definition of transmission medium when sent from a server or other remote source.

Information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may refer to voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, light fields or photons, or any of these. may be represented by a combination of

The terms explained in this disclosure and the terms necessary for understanding the present disclosure may be replaced with terms having the same or similar meanings. For example, the channel and/or symbols may be signaling. A signal may also be a message. A component carrier (CC) may also be called a carrier frequency, a cell, a frequency carrier, or the like.

As used in this disclosure, the terms "system" and "network" are used interchangeably.

In addition, the information, parameters, etc. described in the present disclosure may be expressed using absolute values, may be expressed using relative values from a predetermined value, or may be expressed using other corresponding information. may be represented. For example, radio resources may be indexed.

The names used for the parameters described above are not limiting names in any way. Further, the formulas, etc., using these parameters may differ from those expressly disclosed in this disclosure. Since the various channels (e.g., PUCCH, PDCCH, etc.) and information elements can be identified by any suitable names, the various names assigned to these various channels and information elements are in no way restrictive names. isn't it.

As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Judgement", "determining" are, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (eg, lookup in a table, database, or other data structure); Also, "judgment" and "determination" are used for receiving (e.g., receiving information), transmitting (e.g., transmitting information), input, output, access (accessing) (for example, accessing data in memory) may include deeming that a "judgement" or "decision" has been made. In addition, "judgment" and "decision" are considered to be "judgment" and "decision" by resolving, selecting, choosing, establishing, comparing, etc. can contain. In other words, "judgment" and "decision" can include considering that some action is "judgment" and "decision". Also, "judgment (decision)" may be read as "assuming", "expecting", "considering", or the like.

The terms "connected", "coupled", or any variation thereof, mean any direct or indirect connection or coupling between two or more elements, It can include the presence of one or more intermediate elements between two elements being "connected" or "coupled." Couplings or connections between elements may be physical, logical, or a combination thereof. For example, "connection" may be read as "access". As used in this disclosure, two elements are defined using at least one of one or more wires, cables, and printed electrical connections and, as some non-limiting and non-exhaustive examples, in the radio frequency domain. , electromagnetic energy having wavelengths in the microwave and optical (both visible and invisible) regions, and the like.

As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly specified otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

Where "include," "including," and variations thereof are used in this disclosure, these terms are inclusive, as is the term "comprising." is intended. Furthermore, the term "or" as used in this disclosure is not intended to be an exclusive OR.

In this disclosure, where articles have been added by translation, such as a, an, and the in English, the disclosure may include the plural nouns following these articles.

In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean that "A and B are different from C". Terms such as "separate," "coupled," etc. may also be interpreted in the same manner as "different."

DESCRIPTION OF SYMBOLS 100... Estimation model execution apparatus 101... Communication terminal log collection part 101a... Log table 101b... Application information table 102... Utilization feature amount calculation part 103... Attention function estimation part 104... Estimation result display part 105... Estimation model construction unit 106 Estimation model storage unit 107 Exclusion list 108 Exclusion list generation unit 200 Communication terminal 201 Acceleration sensor 202 Angular velocity sensor 203 Tilt sensor 204 Screen on/off Sensor 205 Application use sensor 207 Communication unit 208 Control unit 209 Program.

Claims (10)

an attention function estimating unit for estimating a user's attention function using a feature amount generated based on a sensor value of a motion sensor of an information processing device possessed by the user;
with
The attention function estimation unit
Determining the user's attention function using an estimation model constructed using the sensor values and attention function indicators of each of a plurality of users ,
The feature quantity is generated based on usage logs obtained from the running state of a plurality of programs executable by the information processing device in addition to the sensor values,
The activation state of the program indicates the activation time during which the program is running or the activation start time from a predetermined state to the activation of the program,
Estimation processing unit. further comprising an estimation model building unit that builds the estimation model using the sensor values and attention function indicators of each of the plurality of users;
The estimation processing device according to claim 1. The attention function estimating unit estimates using a feature amount generated by dividing the sensor value into predetermined time units,
The estimation processing device according to claim 1 or 2. The feature amount is generated based on logs of one or more motion sensors mounted on the information processing device,
The estimation processing device according to any one of claims 1 to 3. The estimation process according to any one of claims 1 to 4, wherein the feature amount is generated by excluding usage logs for programs classified into categories defined by a predetermined method among the programs. Device. 6. The estimation processing device according to claim 5, wherein, as said predetermined method, a category to be excluded is determined based on the TF-IDF calculated for each category of said program. The estimation processing device according to any one of claims 1 to 6 , wherein the activation state of the program includes activation start time from a predetermined state to activation of the program. The activation state of the program includes the activation count, activation time, and communication time of the program used for communication among the one or more activated programs,
The estimation processing device according to any one of claims 1 to 7 . The estimation processing device according to any one of claims 1 to 8 , further comprising an attention function presentation unit that presents the estimated attention function. One or more feature quantities generated based on sensor values of motion sensors of one or more information processing devices possessed by one or more users, and an indicator of attention function of the one or more users Acquired,
By machine learning using the feature quantity as an explanatory variable and the index of the attention function as teacher data,
Construct an estimation model to estimate the user's attention function ,
The feature amount is generated based on a usage log obtained from the running state of a plurality of programs executable by the information processing device,
The activation state of the program indicates the activation time during which the program is running or the activation start time from a predetermined state to the activation of the program,
Estimation model building device.

Images