JP7552443B2 - Information processing method, program, and information processing device - Google Patents

Description

The present disclosure relates to an information processing method, a program, and an information processing device, and in particular to an information processing method, a program, and an information processing device for providing information regarding a user's level of busyness.

In recent years, interactions between users via networks, such as social networks, have become more common. This trend is accelerating with the spread of remote work.

On a network, a user's status is rarely directly visible to other users. Due to this characteristic, there is a demand for technology that can determine the appropriate timing to interact with users on the network (such as asking questions via chat).

JP 2007-4781 A (Patent Document 1) discloses a technology for controlling the timing of email notifications to users. In this technology, the information notification device calculates, for each combination of time period and user location, a response probability (response history) based on whether or not the user will respond to an email within a given time from receiving the email, and then determines whether or not to notify the user of the email based on the response probability, the current time, and the user's current location.

JP 2007-4781 A

Although the technology disclosed in Patent Document 1 notifies a user of emails addressed to that user at a time appropriate to that user, it is used solely to provide services to that user, and is not used to provide other users with information to predict how busy that user will be.

The present disclosure has been devised in light of the above-mentioned circumstances, and its purpose is to provide a technology that allows other users to predict in real time how busy a given user will be.

According to one aspect of the present disclosure, there is provided an information processing method executed by a computer. The information processing method includes a step of acquiring behavioral information defining a user's behavior, the user being a subject of busyness prediction, and further includes a step of accessing a module for identifying a predicted value of a busyness index corresponding to the behavioral information, the busyness index defining busyness at a timing corresponding to the behavioral information, and further includes a step of acquiring a predicted value of the user's busyness index corresponding to the user's behavioral information as a result of application of the user's behavioral information to the module, and a step of outputting the acquired predicted value of the busyness index.

Preferably, the module includes a predictive model trained to output a predicted value of the busyness index as an estimation result in response to input of behavioral information.

Preferably, the module includes information correlating the behavioral information with the predicted value of the busyness index.

Preferably, the step of obtaining a predicted value of the busyness index is executed in response to a given condition being satisfied in an information processing device operated to input a message to be sent to the user, and the given condition includes detection of an input operation of a message to be sent to the user in the information processing device.

Preferably, the information processing method further includes a step of acquiring user identification information, and the module determines a predicted value of the busyness index corresponding to the identification information in addition to the behavioral information, and the step of acquiring the predicted value of the busyness index includes acquiring a predicted value of the user's busyness index corresponding to the user's behavioral information and the user's identification information.

Preferably, the information processing device further executes a step of acquiring information about a planned sender of a message being created for the user, and the module determines a predicted value of a busyness index that further corresponds to the information about the sender in addition to the behavioral information, and the step of acquiring the predicted value of the busyness index includes acquiring a predicted value of a user's busyness index that corresponds to the user's behavioral information and the information about the sender.

Preferably, the step of outputting the predicted value of the busyness index includes displaying the predicted value of the busyness index as at least one of a character string, a color, and an avatar's facial expression that indicate the degree of busyness, and a predicted value of the time it will take the user to reply from receiving a message.

According to another aspect of the present disclosure, there is provided an information processing method executed by a computer for training a prediction model that outputs a predicted value of a busyness index for a user. The information processing method includes a step of acquiring a learning dataset including behavioral information associated with a busyness index, the busyness index defining busyness at a timing corresponding to an action defined by the behavioral information, and further includes a step of executing machine learning of the prediction model such that a prediction result output by inputting the behavioral information of the learning dataset into the prediction model approaches the busyness index of the learning dataset.

Preferably, the training dataset further includes identification information associated with the busyness index, the identification information identifying each user, and the step of performing machine learning of the predictive model includes performing machine learning of the predictive model such that a predicted value of the busyness index output by inputting the behavioral information and the identification information of the training dataset into the predictive model approaches the busyness index of the training dataset.

Preferably, the busyness index includes a time from receipt of a message to a reply to the message, the training dataset further includes information related to a sender of the message that is associated with the busyness index, and the step of performing machine learning of the predictive model includes performing machine learning of the predictive model such that a predicted value of the busyness index output by inputting the behavioral information and the information related to the sender of the training dataset into the predictive model approaches the busyness index of the training dataset.

Preferably, the information about the sender includes at least one of the sender and an attribute of the sender.

Preferably, the busyness indicator represents the time between receiving a message and replying to it.
Preferably, the behavioral information is based on a manner in which the user operates the device.

Preferably, the device includes a keyboard and the behavioral information includes the number of types per unit time on the keyboard.

Preferably, the device includes a pointing device, and the behavioral information includes a moving distance per unit time of a pointer by the pointing device.

Preferably, the behavioral information is based on a state of the user detected by a sensor.
Preferably, the sensor includes an eye gaze sensor, and the behavioral information includes at least one of whether or not the user is directing his or her gaze at the computer display, and the amount of time per unit time that the user is directing his or her gaze at the computer display.

Preferably, the behavioral information includes information representing a user's participation or non-participation in the collaboration tool.

In accordance with yet another aspect of the present disclosure, a program is provided that, when executed by a processor of a computer, causes the computer to carry out the above-described information processing method.

According to yet another aspect of the present disclosure, there is provided an information processing device comprising a memory for storing the above-mentioned program and a processor for executing the program stored in the memory.

According to the present disclosure, a predicted value of an index (busyness index) defining a user's busyness at a timing corresponding to behavioral information defining the user's behavior is obtained. As a result, a predicted value of a busyness index for a certain user at a certain point in time can be obtained based on the behavioral information of the user at that point in time, and provided to other users.

FIG. 13 is a diagram showing an example of a display of a predicted value of a busyness index. FIG. 1 is a diagram illustrating a hardware configuration of an information processing device 100. FIG. 2 is a diagram illustrating a hardware configuration of a server 200. 1 is a diagram illustrating an example of a functional configuration of an information processing system for displaying a predicted value of an index of a user's busyness in an information processing device 100. FIG. FIG. 1 is a diagram for explaining generally the flow of information for machine learning of a predictive model. 2 is a flowchart of a process performed in server 200 for machine learning of a predictive model 210. 10 is a flowchart of a process executed in information processing device 100A for providing a predicted value.

Below, an embodiment of the present disclosure will be described with reference to the drawings. In the following description, the same parts and components are given the same reference numerals. Their names and functions are also the same. Therefore, their description will not be repeated.

[1. Example of predicted value display]
FIG. 1 is a diagram showing an example of a display of a predicted value of a busyness index displayed by an information processing device.

Screen 10 in FIG. 1 includes two messages 11, 12, and a message editing field 13. Each of messages 11, 12 is a message posted to a thread in a chat tool. Message editing field 13 is displayed when an operation to start inputting a message in reply to message 12 is performed (for example, clicking a "Reply" button displayed in relation to message 12), and includes input field 13A, send button 13B, and display field 13C. That is, message editing field 13 is displayed for editing a message to be sent to the user (called "Ken" in the example of FIG. 1) who posted message 12. Display field 13C includes the string "Ken's expected response time: 45 minutes."

The "45 minutes" in the string "Ken's predicted response time: 45 minutes" represents the predicted busyness index value obtained for the user "Ken." This predicted value represents the predicted time it will take for the target user to reply to a message from the time they receive the message.

The information processing device does not need to display the predicted value itself as a numerical value. Instead of or together with the predicted value, the information processing device may display information (such as a character string, a color, an avatar's facial expression, etc.) that indicates the degree of busyness obtained from the predicted value.

The degree of busyness is determined, for example, by identifying the group into which the user's predicted value falls out of two or more groups into which predicted values are classified according to a predetermined threshold. In one example, the predicted values can be classified into the following groups (1) to (4). Group (1) corresponds to the busiest state, group (2) corresponds to the second busiest state after group (1), group (3) corresponds to the second busiest state after group (2), and group (1) corresponds to the least busy state.

- Group (1): 2 hours or more - Group (2): 30 minutes or more but less than 2 hours - Group (3): 5 minutes or more but less than 30 minutes - Group (4): Less than 5 minutes For example, when the information processing device expresses the degree of busyness as a string of characters, it may display “very busy” when the predicted value is classified into group (1), “busy” when the predicted value is classified into group (2), “slightly busy” when the predicted value is classified into group (3), and “normal” when the predicted value is classified into group (4).

When the information processing device expresses the degree of busyness using color, the longer the time period that the group into which the predicted value is classified corresponds to, the darker the color of the information (for example, the predicted value itself or a shape such as a circle) that is displayed.

When expressing the degree of busyness with the facial expression of an avatar, the information processing device may display a "crying hard" facial expression when the predicted value is classified into group (1), a "crying" facial expression (a milder facial expression than the "crying hard" facial expression) when the predicted value is classified into group (2), a "sweating" facial expression when the predicted value is classified into group (3), and a "smiling" facial expression when the predicted value is classified into group (4).

The information processing device may display the predicted busyness level not only when creating a message using a chat tool, but also when creating a message using other types of tools, such as short message service, email, webmail, or social networking services.

In the example of FIG. 1, a predicted value is displayed when inputting a message to reply to an already displayed message (message 12). Note that the information processing device may also display a predicted value when inputting a new message. The predicted value displayed at this time is the predicted value for the user specified as the destination of the newly created message.

2. Hardware Configuration
In this embodiment, the information processing device displays the predicted value in cooperation with the server. The hardware configurations of the information processing device and the server will be described below.

(Information processing device)
2 is a diagram showing a hardware configuration of the information processing device 100. The information processing device 100 may be a mobile terminal such as a smartphone, or may be a built-in terminal such as a personal computer.

As shown in FIG. 2, the information processing device 100 includes a CPU (Central Processing Unit) 101, a RAM (Random Access Memory) 102, a storage 103, a display 104, a keyboard 105, a mouse 106, a NIC (Network Interface Card) 107, and a camera 108. Each element shown in FIG. 2 is connected to each other by an internal bus.

The CPU 101 controls the operation of the information processing device 100. The RAM 102 functions as a work area when the CPU 101 is executing processes. The storage 103 is a non-volatile storage device that stores various programs executed by the CPU 101 and data used to execute the programs. The storage 103 is realized, for example, by a flash EEPROM (Electrically Erasable Programmable Read-Only Memory), a flash ROM (Read Only Memory), a HDD (Hard Disk Drive), and/or an SSD (Solid State Drive).

The program according to the present disclosure may execute processing by calling necessary modules in a predetermined sequence at a predetermined timing among program modules provided as part of a computer's operating system (OS). In such a case, the program itself does not include the above modules, and executes processing in cooperation with the OS. Programs that do not include the above modules may also be included in the program according to the present disclosure.

The programs according to the present disclosure may be provided as part of other programs. In such cases, the programs themselves do not include modules included in the other programs, and the programs execute processing in cooperation with the other programs. Such programs incorporated in other programs may also be included in the programs according to the present disclosure.

The provided program product is installed and executed in a program storage unit such as a hard disk. The program product includes the program itself and a recording medium on which the program is recorded.

Display 104 is an example of a "display of information processing device 100" and displays an image showing the processing result of a program executed by CPU 101. Information processing device 100 may display the image on a display external to information processing device 100. In this case, the external display is another example of a "display of information processing device 100".

The keyboard 105 is realized by a hardware resource having one or more keys. The information processing device 100 may include an element that realizes a software keyboard as the keyboard 105 instead of or in addition to the hardware resource.

The mouse 106 is an example of a pointing device. In the information processing device 100, the mouse 106 can be operated to move a pointer on the screen displayed on the display.

NIC 107 is a communication interface that enables the information processing device 100 to communicate with other devices (such as the server 200 described below) via a network.

(server)
3 is a diagram showing a hardware configuration of the server 200. The server 200 is realized by, for example, a general-purpose computer.

As shown in FIG. 3, the server 200 includes a CPU 201, a RAM 202, a storage 203, a display 204, an input device 205, and a NIC 207. Each element shown in FIG. 3 is connected to each other by an internal bus.

The CPU 201 controls the operation of the server 200. The RAM 202 functions as a work area when the CPU 201 executes processes. The storage 203 is a non-volatile storage device that stores various programs executed by the CPU 201 and data used to execute the programs.

The display 204 is realized by a display device such as a liquid crystal display. The input device 205 is realized by a general-purpose input device such as a keyboard and/or a mouse. The NIC 207 is a communication interface that allows the server 200 to communicate with other devices (such as the information processing device 100) via a network.

[3. Functional configuration]
Fig. 4 is a diagram showing an example of a functional configuration of an information processing system for displaying a predicted value of an index of a user's busyness in the information processing device 100. The information processing system shown in Fig. 4 includes an information processing device 100A, an information processing device 100B, and a server 200. Each of the information processing device 100A and the information processing device 100B can be realized by the information processing device 100 shown in Fig. 2.

The information processing device 100A and the information processing device 100B send and receive messages to each other via a chat tool. The server 200 provides information such as the above-mentioned predicted values to the information processing device 100A and the information processing device 100B. The functional configuration of each device will be described below.

(Information processing device 100)
The information processing device 100 (each of the information processing device 100A and the information processing device 100B) includes, as its functions, a busyness prediction unit 110, a chat tool 150, and a collaboration tool 160. The busyness prediction unit 110 includes various functions for displaying a predicted value of a busyness index, and is realized, for example, by the CPU 101 executing a given program. The detailed functions of the busyness prediction unit 110 will be described later.

The chat tool 150 has a function for sending and receiving messages using the chat tool, and is realized, for example, by the CPU 101 executing an application program for the chat tool. The collaboration tool 160 has a function for communication using a collaboration tool such as a web conference system, and is realized, for example, by the CPU 101 executing a program for a collaboration application. The program may be installed in the information processing device 100, or may be installed in a device different from the information processing device 100.

The busyness prediction unit 110 detects the sending and receiving of messages in the chat tool 150. The busyness prediction unit 110 also includes, as its functions, a behavioral information acquisition unit 120, a response time detection unit 130, and a predicted response time display unit 140.

The behavioral information acquisition unit 120 acquires operation information from devices such as the keyboard 105, thereby generating behavioral information (information defining the user's behavior) of the user of the information processing device 100. More specifically, in the behavioral information acquisition unit 120, the detection unit 121 uses the operation information acquired from the keyboard 105 to generate the number of typing operations on the keyboard 105 by the user per unit time. The detection unit 122 uses the operation information acquired from the mouse 106 to generate the distance the user moves the pointer using the mouse 106 per unit time. The "user of the information processing device 100" is, for example, a user who is logged in to at least one application program executed in the information processing device 100 in order to use the information processing device 100.

The behavioral information acquisition unit 120 acquires the state of the user detected by a sensor such as the camera 108, thereby generating behavioral information (information defining the user's behavior) of the user of the information processing device 100. More specifically, in the behavioral information acquisition unit 120, the detection unit 123 uses an image from the camera 108 to identify the position where the user's gaze is directed, and thereby generates information indicating whether the user is currently directing his/her gaze at the display 104 and/or the time per unit time that the user directed his/her gaze at the display 104. In this sense, the program elements (software elements) constituting the detection unit 123 are an example of a gaze sensor. Note that the gaze sensor may be realized by an independent hardware element, or may be realized by a combination of a hardware element and a software element.

In addition, in the behavioral information acquisition unit 120, the detection unit 124 acquires the login status of the user in the collaboration tool 160 (whether they are logged in, logged out, etc.), and thereby generates information indicating whether the user is participating or not in the collaboration tool.

As a result of the above, the detection units 121 to 124 can generate five types of information, indicated as A1 to A5 below. Each of these five types is an example of behavioral information.

A1: The number of types of keyboard 105 entered by the user per unit time. A2: The distance the user moves the pointer using the mouse 106 per unit time. A3: Whether the user is directing his/her gaze at the display 104. A4: The amount of time the user directs his/her gaze at the display 104 per unit time. A5: The user's participation (or non-participation) in the collaboration tool.
Of the above five types, A1, A2, and A4 are calculated at regular time intervals, and the latest calculated value may be used to obtain a predicted value of the busyness index. For example, for A1, the information processing device 100 may generate the "number of typings on the keyboard 105 by the user per unit time" every 10 seconds by using the number of typing operations detected on the keyboard 105 during the latest 10 seconds. Then, the information processing device 100 may use the latest "number of typings on the keyboard 105 by the user per unit time" to obtain the predicted value.

The response time detection unit 130 generates information (response time) in the information processing device 100 that represents the time from when a user receives a message on the chat tool 150 to when the user sends a reply to the message.

It is assumed that when a user is busy, it takes a long time from receiving a message to sending a reply to that message, and when a user is not so busy, it does not take as long to send a reply to that message. In this sense, response time (the time from receiving a message to replying) can be said to be an example of an indicator of how busy a user is.

It is assumed that when a user is busy, the reply to a received message will be short (the reply will contain less content), and when the user is not so busy, the reply to the message will be longer (the reply will contain more content). In this sense, the length of the reply message can be said to be another example of an indicator of how busy a user is.

The predicted response time display unit 140 has a function of outputting a predicted value of the busyness index acquired by the information processing device 100. In one example, the busyness prediction unit 110 may be realized as a plug-in for an application for a chat tool, and the predicted response time display unit 140 may output a predicted value of the busyness index to the chat tool 150. In this case, the chat tool 150 displays the predicted value on the display 104, as shown as display field 13C in FIG. 2.

(Server 200)
The server 200 includes, as its functions, a predictive model 210, a learning engine 220, and an analysis engine 230.

In response to the input of a user's behavioral information, the prediction model 210 outputs a predicted value of a busyness index at a timing corresponding to the behavioral information. The prediction model 210 includes a program executed by the CPU 201 (to realize a machine learning model according to a given algorithm) and parameters applied to the program. The program and the program may be stored in the storage 203, or may be stored in a storage device outside the server 200 that is accessible by the CPU 201.

The learning engine 220 uses the learning dataset and executes machine learning processing of the prediction model 210 according to a machine learning algorithm. This allows the prediction model 210 to be updated. The learning engine 220 is realized, for example, by the CPU 201 executing a given program.

The analysis engine 230 applies the user's behavioral information to the prediction model 210 to obtain a predicted value of the busyness index at the timing corresponding to the behavioral information. The analysis engine 230 is realized, for example, by the CPU 201 executing a given program.

[4. Overview of obtaining predicted values]
4, a description will be given of an example of how a predicted value is acquired by the information processing device 100. In this description, a scene will be assumed in which an input operation of a message addressed to a user of the information processing device 100B is being performed in the information processing device 100A.

In the information processing device 100A, the busyness prediction unit 110 monitors the state of the chat tool 150, and when an operation to input a message is performed and a destination of the message is selected, it sends a request to the server 200 for a predicted value of the busyness index for the destination user (prediction value request). In one example, the request may include identification information of the destination user (user of the information processing device 100B).

The analysis engine 230 of the server 200 acquires behavioral information of the user from the information processing device 100B in response to a prediction value request from the information processing device 100A. The server 200 may store information associating the identification information of the user with the information processing device 100B, and the analysis engine 230 may refer to the information to identify the information processing device 100B as the device corresponding to the identification information and request behavioral information from the information processing device 100B. The information processing device 100B may transmit the behavioral information to the server 200 in response to a request from the server 200.

The behavioral information to be transmitted includes, for example, one or more of the above A1 to A5.
When the analysis engine 230 obtains the behavioral information, it applies the behavioral information to the prediction model 210 to obtain a predicted value of the busyness index.

In one example, the greater the "number of keyboard 105 types by the user per unit time," the busier the user will be (e.g., a longer response time) as the predicted value of the busyness index, which is the prediction result, indicates.

In one example, the longer the "distance that the user moves the pointer using the mouse 106 per unit time," the busier the predicted value of the busyness index, which is the prediction result, indicates that the user is busier.

In one example, with respect to "whether the user is directing their gaze at the display 104," if the user is "directing their gaze at the display 104," the predicted value of the busyness index, which is the prediction result, indicates that the user is busier than if the user is "directing their gaze at the display 104."

In one example, the longer the "time that the user directs their gaze at the display 104 per unit time," the busier the predicted value of the busyness index, which is the prediction result, indicates the busier the user is.

In one example, for "user participation (or non-participation) in a collaboration tool," the predicted value of the busyness index, which is the prediction result, indicates that the user is busier in the case of "participation" than in the case of "non-participation." On the other hand, there may be cases where a user is unable to participate in a collaboration tool due to being busy. Therefore, there may be cases where the predicted value of the busyness index, which is the prediction result, indicates that the user is busier in the case of "non-participation" than in the case of "participation."

Then, the analysis engine 230 transmits the predicted value of the busyness index obtained as described above to the information processing device 100A.

In the information processing device 100A, the predicted response time display unit 140 outputs the predicted value of the busyness index sent from the analysis engine 230.

In the example described with reference to FIG. 1, the busyness prediction target is the destination of the message being created in information processing device 100A (the user of information processing device 100B). Note that the busyness prediction target is not limited to one person. For example, when the information processing device 100A displays a list of candidate destinations in a pull-down menu during message input, it may obtain and display a predicted value of the busyness index for each destination in the list.

In response to a prediction value request from the information processing device 100A, the analysis engine 230 acquires behavioral information of the user to be predicted, and applies the behavioral information to the prediction model 210 to acquire a prediction value of the busyness index for the user to be predicted. Note that acquisition of the prediction value does not need to be triggered by a prediction value request. The analysis engine 230 may acquire behavioral information of each of one or more users at regular intervals, for example, and acquire a prediction value of the busyness index for each of the one or more users.

[5. Overview of machine learning for predictive models]
Fig. 5 is a diagram for explaining the information flow for machine learning of a predictive model. Fig. 5 shows the same functional configuration as shown in Fig. 4.

The information processing device 100 (each of the information processing device 100A and the information processing device 100B) generates a training dataset and provides the generated training dataset to the server 200. The generation of the training dataset in the information processing device 100 will be described later with reference to FIG. 6.

The learning engine 220 of the server 200 uses the provided learning dataset to perform machine learning of the predictive model 210.

The information processing device 100 may provide the server 200 with a learning dataset each time a reply message is sent from the information processing device 100, or may provide the learning dataset when a given number of learning datasets are generated in the information processing device 100, or may provide the learning dataset at regular intervals.

[6. Processing flow (optimization of predictive model)]
6 is a flowchart of a process executed in the server 200 for machine learning of the prediction model 210. Hereinafter, the contents of the processes of the information processing device 100 and the server 200 shown in FIG. 6 will be described.

In step S100, the information processing device 100 determines whether or not a message has been received by the chat tool 150.

In step S102, the information processing device 100 stores the "behavioral information" at that time in a given memory area (for example, in the storage 103). The behavioral information is, for example, one or more types of information from A1 to A5 above.

In step S104, the information processing device 100 starts a timer.
In step S106, information processing device 100 determines whether a reply to the message determined to have been received in step S100 has been sent. Information processing device 100 repeats the control of step S106, for example, at regular time intervals, until it determines that a reply has been sent (NO in step S106), and when it determines that a reply has been sent (YES in step S106), it proceeds to control to step S108.

In step S108, the information processing device 100 ends the timer that was started in step S102.

In step S110, the information processing device 100 acquires the time measured by the timer from when it is started in step S104 to when it is ended in step S108 as the "response time" and stores the "response time" in a given memory area.

In step S112, the information processing device 100 generates a learning dataset by labeling the "behavioral information" stored in step S102 with the "response time" stored in step S110, and transmits the learning dataset to the server 200. In this specification, "labeling" means an example of "associating" two or more types of data with each other. Thereafter, the information processing device 100 returns control to step S100.

In response to receiving the learning dataset from the information processing device 100, the server 200 starts the processing shown in FIG. 6. The processing of the server 200 in FIG. 6 is realized, for example, as a function of the learning engine 220.

In step S200, the server 200 receives the learning dataset sent from the information processing device 100.

In step S202, the server 200 uses the training data set received in step S200 to perform machine learning on the prediction model 210, thereby updating the prediction model 210 (the parameters of the prediction model 210). In other words, the prediction model 210 is updated so that the prediction result output by inputting the behavioral information of the training data set into the prediction model 210 approaches the "response time" (an index of busyness) of the training data set. Thereafter, the server 200 returns control to step S200.

[7. Processing flow (Providing predicted values)]
Fig. 7 is a flowchart of a process executed for providing a predicted value in the information processing device 100A. In the example of Fig. 7, a predicted value is provided to a user of the information processing device 100A who creates a message to be sent to a user of the information processing device 100B. The contents of the processes of the information processing device 100A, the server 200, and the information processing device 100B shown in Fig. 7 will be described below.

In step S120, information processing device 100A determines whether an operation has been performed to create a message to be sent to the user of information processing device 100B. An example of such an operation may be an operation of inputting information specifying the user of information processing device 100B as the destination of the message, or an operation of specifying a reply to a message from the user of information processing device 100B (clicking a reply button). Information processing device 100A repeats the control of step S120, for example, at regular intervals until it determines that the above operation has been performed (NO in step S120), and when it determines that the above operation has been performed (YES in step S120), it proceeds to step S122.

In step S122, the information processing device 100A sends a request for a predicted value regarding the user of the information processing device 100B to the server 200.

The processing of the server 200 in FIG. 7 is realized, for example, as a function of the analysis engine 230.

In step S220, the server 200 acquires a request for a predicted value from the information processing device 100A, and in step S222, sends a request for behavioral information of the user of the information processing device 100B to the information processing device 100B.

In step S130, the information processing device 100B receives a request from the server 200, in step S132 generates behavioral information of the user of the information processing device 100B, and in step S134 transmits the generated behavioral information to the server 200.

In step S224, the server 200 acquires behavioral information from the information processing device 100B, and in step S226, the server 200 applies the behavioral information to the prediction model 210 to acquire a predicted value (e.g., a response time) corresponding to the behavioral information. The acquired predicted value is a predicted value related to the information processing device 100B. Then, in step S228, the server 200 transmits the predicted value to the information processing device 100A.

In step S124, the information processing device 100A acquires the predicted value for the information processing device 100B transmitted from the server 200. Then, in step S126, the information processing device 100A displays the acquired predicted value.

8. Modifications
Some of the possible variations of the above-described embodiment will be described below.

(1) Form of Module In the example shown in FIG. 4 etc., the prediction model 210 has been described as an example of a module used to obtain a predicted value. Note that it is not necessary for a machine learning model to be used to obtain a predicted value. That is, related information (such as a table) that associates the predicted value of the "busyness index" with the "behavioral information" may be stored in the storage 103 of the information processing device 100, and the information processing device 100 may use the related information as a module for obtaining a predicted value. More specifically, the information processing device 100 may obtain a predicted value corresponding to the behavioral information of the user to be predicted by referring to the related information.

(2) Information Used to Acquire Predicted Value of Busyness Index In the above-described embodiment, behavioral information is used to acquire a predicted value of a busyness index. That is, the learning dataset used for the machine learning of the prediction model 210 is generated by labeling behavioral information with a busyness index, and the behavioral information is provided to the prediction model 210 to acquire a predicted value of the busyness index.

In this regard, in addition to behavioral information, information about the source of the messages used to generate the response time may be used to obtain a predicted value for the busyness index.

More specifically, the learning request dataset used for the machine learning of the prediction model 210 may be generated by labeling information about the message sender with a busyness index in addition to the behavioral information. Even more specifically, in step S102 of FIG. 6, the information processing device 100 may store information about the sender of the message determined to have been received in step S100 in a given storage area in addition to the behavioral information. Then, in step S112 of FIG. 6, the information processing device 100 may generate a learning dataset by labeling the behavioral information and information about the message sender stored in step S102 with a response time (busyness index).

The information regarding the sender of the message may be information identifying the sender user itself, or may be the attributes of the sender user (position in the organization, gender, language used, age, etc.), or a combination of these. The attributes of each user may be registered in the information processing device 100 as data related to the chat tool 150. In step S102, the information processing device 100 may obtain the attributes of the sender by referring to the data.

In addition, to obtain a predicted value, the information processing device 100 may apply information about the user who is the source of the message, in addition to behavioral information about the user who is the destination of the message, to the prediction model 210 to obtain the predicted value.

When the related information described above is used as a prediction module, the related information may store a predicted value of the busyness index in association with information about the destination user and behavioral information. The information processing device 100 may obtain a predicted value corresponding to the information about the destination user and behavioral information by referring to the related information.

(3) User-Specific Modules The modules used to obtain predicted values may be set for each user.

For example, the training dataset may be labeled with not only the busyness indicator but also the user's identification information, and the predictive model 210 may be generated and machine-learned for each identification information. Also, the above-mentioned related information may be prepared for each identification information.

(4) Arrangement of Functions in Hardware Resources In the embodiment described above, the server 200 includes a predictive model 210, a learning engine 220, and an analysis engine 230.

On the other hand, each information processing device 100 may include an analysis engine 230. For example, an application program for the analysis engine 230 may be installed in each information processing device 100, and the analysis engine 230 may be realized by the CPU 101 executing the application program.

Each information processing device 100 may include a learning engine 220 and a prediction model 210. For example, an application program for the learning engine 220 may be installed in each information processing device 100, and the learning engine 220 may be realized by the CPU 101 executing the application program. Also, a program for realizing the prediction model 210 may be installed in each information processing device 100, and parameters that define the prediction model 210 may be stored. The prediction model 210 may be realized by the CPU 101 executing the program using the parameters. In this case, a process for aligning parameters used in the prediction model 210 between multiple information processing devices 100 (such as a notification of parameters from one information processing device 100 to another information processing device 100) may be performed periodically and/or when a given condition (such as an instruction from a user or a system administrator) is met.

(5) Timing Corresponding to Behavioral Information The above-described "prediction value" corresponds to the same timing as the "behavioral information" used to obtain the "prediction value." In other words, when "behavioral information" at a certain point in time is input, the module for obtaining the prediction value outputs a prediction value of the time required from receipt to reply if the user receives a message at that point in time.

Note that the "predicted value" may correspond to a different timing than the "behavioral information" used to obtain the "predicted value." In other words, when "behavioral information" at a certain point in time is input, the module for obtaining the predicted value may output a predicted value of the time required from receipt to reply when the user receives a message a certain time later (e.g., one minute later) from that point in time. When the module for obtaining the predicted value is realized by a machine learning model, in the learning dataset used for machine learning of the machine learning model, the behavioral information at a given timing may be labeled with a "busyness index" for a message received a certain time later from the given timing.

In the information processing system, both of the above two types of predicted values (a predicted value corresponding to the same timing as the behavioral information, and a predicted value corresponding to a certain time after the behavioral information) may be provided as "predicted values." This allows a person who sees both predicted values to determine whether it would be better to send a message to a user at that time or after a certain time has passed.

For example, if the "predicted value corresponding to the same timing as the behavioral information" indicates that the user is busier than the "predicted value corresponding to a certain time after the behavioral information," someone who sees both predicted values can postpone sending a message to that user until a certain time later, thereby adjusting the timing of sending the message to a time when the user is less busy.

(6) Timing of outputting predicted value The information processing device 100 acquires and displays a predicted value of a busyness index of a certain user while the information processing device 100 is creating a message for the certain user. The timing of acquiring and displaying the predicted value may be when the user starts creating a message (for example, when the user sends a "reply button" for a message from a certain user in a chat system), when the user starts inputting text that constitutes a message, or when the certain user is designated as the destination of the message (when the user's address is entered or selected as the destination of the message in an e-mail). The information processing device 100 may also acquire a predicted value at regular intervals while the message for the certain user is being created, and update the display of the predicted value.

The embodiments disclosed herein should be considered in all respects as illustrative and not restrictive. The scope of the present invention is indicated by the claims, not by the above description, and is intended to include all modifications within the meaning and scope of the claims. Furthermore, the inventions described in the embodiments and each modified example are intended to be implemented, as far as possible, either alone or in combination.

10 screen, 11, 12 message, 13 message editing field, 13A input field, 13B send button, 13C display field, 100, 100A, 100B information processing device, 102, 202 RAM, 103, 203 storage, 104, 204 display, 105 keyboard, 106 mouse, 108 camera, 110 busyness prediction unit, 120 behavioral information acquisition unit, 121-124 detection unit, 130 response time detection unit, 140 predicted response time display unit, 150 chat tool, 160 collaboration tool, 200 server, 205 input device, 210 prediction model, 220 learning engine, 230 analysis engine.

Claims (18)

1. A computer-implemented information processing method, comprising:
obtaining behavioral information defining a behavior of a user, the user being a subject of busyness prediction;
accessing a module for determining a predicted busyness index corresponding to the behavioral information, the busyness index defining a busyness at a time corresponding to the behavioral information;
obtaining, as a result of application of the user behavior information to said module, a predicted value of said user busyness index corresponding to said user behavior information;
and outputting the obtained predicted value of the busyness index.
the step of acquiring a predicted value of the busyness index is executed in response to a given condition being satisfied in an information processing device operated for inputting a message to be sent to the user;
The information processing method, wherein the given condition includes detection, in the information processing device, of an input operation for a message to be sent to the user. 1. A computer-implemented information processing method, comprising:
obtaining behavioral information defining a behavior of a user, the user being a subject of busyness prediction;
accessing a module for determining a predicted busyness index corresponding to the behavioral information, the busyness index defining a busyness at a time corresponding to the behavioral information;
obtaining, as a result of application of the user behavior information to said module, a predicted value of said user busyness index corresponding to said user behavior information;
outputting the obtained predicted value of the busyness index;
obtaining information regarding a proposed source of a message being generated for the user ;
the module determines a predicted busyness metric value that is further responsive to the source information in addition to the behavioral information;
An information processing method, wherein the step of obtaining a predicted value of the busyness index includes obtaining a predicted value of the user's busyness index corresponding to the user's behavioral information and information related to the sender. The information processing method according to claim 1 or 2 , wherein the module includes a prediction model trained to output a predicted value of a busyness index as an estimation result in response to input of behavioral information. The information processing method according to claim 1 , wherein the module includes information that correlates behavioral information with a predicted value of a busyness index. obtaining an identity of the user;
the module determines a predicted value of a busyness metric corresponding to the behavioral information plus the identification information;
The information processing method according to any one of claims 1 to 4, wherein the step of obtaining a predicted value of the busyness index includes obtaining a predicted value of the busyness index of the user corresponding to the user's behavioral information and the user's identification information. The information processing method according to any one of claims 1 to 5, wherein the step of outputting the predicted value of the busyness index includes displaying the predicted value of the busyness index as at least one of a character string, a color, and an avatar 's facial expression indicating a degree of busyness, and a predicted value of the time it will take the user from receiving the message to replying. 1. An information processing method for training a predictive model that outputs a predicted value of a user's busyness index, the method comprising:
acquiring a learning dataset including behavioral information associated with a busyness index, the busyness index defining a busyness at a timing corresponding to a behavior defined by the behavioral information;
The method further includes a step of executing machine learning of the predictive model so that a prediction result output by inputting the behavioral information of the learning dataset into the predictive model approaches the busyness index of the learning dataset;
The busyness metric includes a time between receiving a message and replying to the message;
The training data set further includes information about a source of a message associated with the busyness index;
An information processing method, wherein the step of performing machine learning of the predictive model includes performing machine learning of the predictive model so that a predicted value of a busyness index output by inputting the behavioral information and information regarding the sender of the learning dataset into the predictive model approaches the busyness index of the learning dataset. The training data set further includes identification information associated with the busyness index;
The identification information identifies each user;
The information processing method according to claim 7, wherein the step of performing machine learning of the predictive model includes performing machine learning of the predictive model so that a predicted value of a busyness index output by inputting the behavioral information of the learning dataset into the predictive model approaches the busyness index of the learning dataset. 8. The information processing method according to claim 2 , wherein the information relating to the sender includes at least one of the sender and an attribute of the sender. The information processing method according to any one of claims 1 to 9 , wherein the busyness index indicates a time from receipt of a message to a reply. The information processing method according to any one of claims 1 to 10 , wherein the behavioral information is based on a manner in which the user operates a device. the device includes a keyboard;
The information processing method according to claim 11 , wherein the behavioral information includes a number of types on a keyboard per unit time. the device includes a pointing device;
The information processing method according to claim 11 , wherein the behavioral information includes a moving distance per unit time of a pointer moved by a pointing device. The information processing method according to any one of claims 1 to 13 , wherein the behavioral information is based on a state of the user detected by a sensor. The sensor includes a gaze sensor;
The information processing method according to claim 14 , wherein the behavioral information includes at least one of whether the user is directing his/her gaze at the display of the computer and the time per unit time that the user directs his/her gaze at the display of the computer . The information processing method according to any one of claims 1 to 15 , wherein the behavioral information includes information indicating whether the user is participating or not in a collaboration tool. A program that, when executed by a processor of a computer, causes the computer to carry out the information processing method according to any one of claims 1 to 16 . A memory for storing the program according to claim 17 ;
and a processor that executes a program stored in the memory.

Images