Prediction method and prediction program

Abstract

A prediction device disclosed herein acquires a series of actions that members of a group performing a collaborative task each performed in order to perform the collaborative task. Next, the prediction device uses the acquired series of actions of each member to calculate, for each member, an action adjustment probability indicating the probability that the member will change the next action in accordance with the action of another member and the probability that the member will not change the next action. Thereafter, on the basis of the calculated action adjustment probability for each member, the prediction device calculates, for each member, a value obtained by multiplying the probability that the member will not change the next action by the probability that another member will change the next action in accordance with the action of the member. Also, the prediction device outputs a value, obtained by adding up the calculated values for the members of the group, as a predicted value of the level of performance by the group.

Description

Prediction method and prediction program

[0001] The present invention relates to a prediction method and a prediction program.

[0002] Conventionally, there is a technology that predicts the performance of a group performing a task (collaborative task) in which multiple people work together, based on the personality traits of each person obtained through a questionnaire (see Non-Patent Document 1).

[0003] Jo E. Hannay et al, Effects of Personality on Pair Programming, IEEE Transactions on Software Engineering 36(1), 61-80 (2010)

[0004] While the above techniques can easily predict the level of performance of a group performing a cooperative task, they have the problem of low prediction accuracy. Therefore, an object of the present invention is to solve the above problem and improve the prediction accuracy of the level of performance of a group performing a cooperative task.

[0005] In order to solve the above-mentioned problems, the present invention provides a prediction method executed by a prediction device, which includes an acquisition step of acquiring a series of actions taken by each member of a group performing a cooperative task in order to perform the cooperative task; a first calculation step of calculating, for each member, a behavior adjustment probability indicating the probability that the member will change his or her next behavior to match the behavior of other members and the probability that the member will not change his or her next behavior, using the acquired series of actions of each member; and a second calculation step of calculating, for each member, a value obtained by multiplying the probability that the member will not change his or her next behavior by the probability that other members will change their next behavior to match the behavior of the member, based on the calculated behavior adjustment probability for each member, and adding up the calculated values ​​for all members of the group to output the sum as a predicted value of the level of performance by the group.

[0006] According to the present invention, it is possible to improve the accuracy of predicting the level of performance of a group performing a cooperative task.

[0007] FIG. 1 is a diagram for explaining an overview of a prediction device according to each embodiment. FIG. 2 is a diagram illustrating an example of the configuration of a prediction device according to a first embodiment. FIG. 3 is a flowchart illustrating an example of a processing procedure executed by the prediction device according to the first embodiment. FIG. 4 is a diagram illustrating an example of a cooperative task. FIG. 5 is a diagram illustrating an example of an implementation result of the cooperative task. FIG. 6 is a diagram illustrating an example of a behavior adjustment amount calculated from the implementation result of the cooperative task shown in FIG. 5. FIG. 7 is a diagram illustrating an example of the configuration of a prediction device according to a second embodiment. FIG. 8 is a flowchart illustrating an example of a processing procedure executed by the prediction device according to the second embodiment. FIG. 9 is a diagram illustrating a cooperative task used in an evaluation experiment of Interaction Score (IS). FIG. 10 is a diagram illustrating the results of the evaluation experiment of Interaction Score (IS). FIG. 11 is a diagram illustrating an example of the configuration of a computer that executes a prediction program.

[0008] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a first embodiment and a second embodiment of the present invention will be described with reference to the accompanying drawings. The present invention is not limited to the respective embodiments.

[0009] First Embodiment Overview An overview of the processing executed by the prediction device of each embodiment will be described below. The prediction device predicts the level of performance of a group performing a task that is a cooperative work (cooperative task).

[0010] The cooperative work referred to here is, for example, a task in which, when a situation (information) is given, each member of a group (also referred to as "performer") performing the cooperative work selects a strategy and performs the work (action). The cooperative work is, for example, a repetitive task in which the next action of the other member is unpredictable. Examples of cooperative work include playing defense in baseball, playing doubles in badminton or tennis, and dividing up work. Note that the above is merely an example, and the cooperative work may be any task performed by multiple people in cooperation (cooperation) with a common goal, such as a task performed by three or more people as a team.

[0011] Here, the prediction device calculates an Interaction Score of the group as a predicted value of the group's performance when performing a task, based on the past task performance results of each member of the group. The Interaction Score is a score that indicates the degree to which each member of the group coordinated their actions with each other when performing a task. The Interaction Score is calculated, for example, as follows.

[0012] Here, an example will be described in which the members of a group performing a cooperative task are Mr. A and Mr. B. In this case, the prediction device acquires the results of the tasks performed by Mr. A and Mr. B, for example, as shown in FIG. 1. The results of the tasks performed by Mr. A and Mr. B are represented by a matrix a A ,a B ∈R^(A×N) and a matrix a showing the results of the actions of A and B. A_part ,a B_part ∈R^(A×N), where A is the number of dimensions of each action, N is the number of attempts of the action, and R represents a real number.

[0013] Next, the prediction device calculates the behavioral adjustment amounts for the tasks of Persons A and B from the execution results of the tasks of Persons A and B. For example, the prediction device calculates a matrix indicating the behavioral adjustment amounts for the tasks of Persons A and B.

[0014] The behavior adjustment amount is a value indicating the degree to which a member performing a task has changed his or her own behavior based on the behavior of the other member (details will be described later). The behavior adjustment amount is expressed, for example, by the probability distribution of each behavior.

[0015] For example, if the behavior of a member is classified into three types: (1) not changing the next behavior to match the behavior of other members, (2) changing the next behavior to match the behavior of other members, and (3) taking other actions, the amount of behavior adjustment can be expressed as a three-dimensional probability distribution. For example, the matrix showing the amount of behavior adjustment (behavior adjustment probability) between Mr. A and Mr. B is a A_adjust ,a B_adjust ∈R^3.

[0016] Thereafter, the prediction device inputs the behavioral adjustment amounts of A and B into a determination function for the Interaction Score (details to be described later) to calculate the Interaction Score. As a result, the prediction device calculates the Interaction Score IS for A and B. AB ∈R.

[0017] In this way, the prediction device extracts the behavioral adjustment amount for each member of the group from the past task execution results of each member of the group.The prediction device then calculates the Interaction Score of the group using the extracted behavioral adjustment amount for each member of the group.

[0018] This Interaction Score can be used to predict the performance of each group. For example, the prediction device calculates the Interaction Score for each of multiple groups with different members and determines the group with the highest Interaction Score. This allows users of the prediction device to know the combination of members that can achieve high performance (i.e., a group composed of members who get along well).

[0019] [Configuration Example] Next, a configuration example of the prediction device 10 will be described with reference to Fig. 2. The prediction device 10 is realized by a general-purpose computer such as a personal computer, and includes a communication unit 110, a storage unit 120, and a control unit 130.

[0020] In addition, the prediction device 10 may have an input unit (e.g., a keyboard, a mouse, etc.) that accepts various operations from a user of the prediction device 10, an output unit such as a display unit (e.g., a liquid crystal display, etc.) for displaying information, and an audio output unit (e.g., a speaker, etc.) for outputting information aloud.

[0021] The communication unit 110 is realized by, for example, a network interface card (NIC), etc. The communication unit 110 is connected to a predetermined network such as the Internet via a wired or wireless connection, and transmits and receives information to and from external devices.

[0022] The storage unit 120 is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 120 stores data and programs used when the control unit 130 executes various processes. For example, the storage unit 120 stores the results of task execution for each group.

[0023] The control unit 130 is responsible for overall control of the prediction device 10. The functions of the control unit 130 are realized, for example, by a central processing unit (CPU) executing a program stored in the storage unit 120.

[0024] The control unit 130 includes an acquisition unit 131 , a first calculation unit 132 , and a second calculation unit 133 .

[0025] The acquisition unit 131 acquires information indicating a series of actions (task execution results) taken by each member of the group to execute a cooperative task.

[0026] The first calculation unit 132 calculates a behavior adjustment amount (behavior adjustment probability) for each member using the information indicating a series of behaviors of each member of the group acquired by the acquisition unit 131 .

[0027] For example, when a member takes a next action after an action that has not yet produced the best performance, the probability distribution of whether the member takes one of the following actions is calculated: (1) (assuming that the other member will try to match him / her) not changing the action, (2) (assuming that the other member will not try to match him / her) taking an action that will result in the best performance if the other member continues to take the same action, or (3) taking an action other than (1) or (2) above, and the calculated probability distribution is set as the action adjustment probability. Note that the above (3) could, for example, be changing the next action so as not to match the action of the other member.

[0028] The first calculation unit 132, for example, uses information indicating a series of actions of each member of the group to calculate, for each member, the proportion (probability distribution) of the members that will take the above actions (1) to (3) in their next action.

[0029] The second calculation unit 133 calculates an Interaction Score of the group based on the behavior adjustment probability of each member of the group calculated by the first calculation unit 132.

[0030] For example, the second calculation unit 133 calculates, for each member of the group, a value obtained by multiplying the probability ((1)) that the member will not change his or her next behavior by the probability ((2)) that other members will change their next behavior to match the behavior of the member in question, based on the behavior adjustment probability of each member of the group.The second calculation unit 133 then adds up the calculated values ​​for all members of the group and outputs the sum as the Interaction Score.

[0031] For example, if a group has two members (a first member and a second member), the Interaction Score is calculated using the following formula (1): In the following formula (1), the first member in the group is referred to as "you" and the second member is referred to as "the other person."

[0032] (1. Probability of not changing your behavior) x (2. Probability of the other person adjusting their behavior) + (1. Probability of the other person not changing their behavior) x (2. Probability of you adjusting your behavior) ... Equation (1)

[0033] The second calculation unit 133 may calculate the Interaction Score by taking into account the above-mentioned probability (3) among the behavior adjustment probabilities of each member of the group. For example, the second calculation unit 133 may calculate the Interaction Score by subtracting the probability ((3)) that each member of the group will change their next behavior so as not to match the behavior of other members from the value calculated by the above formula (1). In this case, the Interaction Score is calculated, for example, by the following formula (2).

[0034] (1. Probability that you will not change your behavior) x (2. Probability that the other person will change their behavior to match) + (1. Probability that the other person will not change their behavior) x (2. Probability that you will change your behavior to match) - (3. Probability that you will change your behavior in a way that does not match the other person) - (3. Probability that the other person will change their behavior in a way that does not match yours) ... Equation (2)

[0035] [Example of Processing Procedure] Next, an example of processing procedure executed by the prediction device 10 will be described with reference to FIG. 3 . First, the acquisition unit 131 of the prediction device 10 acquires the results of the task performed by the group (S1). Next, the first calculation unit 132 calculates the behavior adjustment amount (behavior adjustment probability) of each member of the group based on the results of the task performed by the group acquired in S1 (S2). Thereafter, the second calculation unit 133 calculates and outputs an Interaction Score based on the behavior adjustment amount of each member calculated in S2 (S3).

[0036] According to the prediction device 10 described above, it is possible to accurately predict the level of performance of a group performing a cooperative task.

[0037] [Example of Processing Executed by the Prediction Device] An example of processing executed by the prediction device 10 will be described below. Here, an example will be described in which the cooperative task is a task in which two people repeatedly select one card at the same time and select the same card (see FIG. 4). FIG. 5 shows examples of cards selected by two people (person A and person B). If the series of actions (selections) of person A and person B shown in FIG. 5 are classified into (1. Do not change behavior, 2. Change behavior to match the other person, 3. Other), the result will be as shown in FIG. 6.

[0038] Then, when the prediction device 10 calculates the proportion (behavior adjustment probability) of the behaviors of Mr. A and Mr. B shown in Figure 6 (1. Do not change behavior, 2. Change behavior to suit the other person, 3. Other), the results are Mr. A: (0.57, 0.43, 0) and Mr. B: (0.43, 0.29, 0.29).

[0039] When the prediction device 10 calculates the Interaction Score (IS) based on the behavioral adjustment probability between Mr. A and Mr. B, the result is IS = 0.57 × 0.29 + 0.43 × 0.43 - 0 - 0.29 = 0.06. For example, the prediction device 10 can similarly calculate the Interaction Scores for groups with other member configurations and compare the Interaction Scores to determine the member configuration of a group predicted to have high performance.

[0040] Second Embodiment An embodiment in which the prediction device 10 determines the member configuration of a group predicted to have high performance by comparing Interaction Scores as described above will be described as the second embodiment. The same components as those in the first embodiment are denoted by the same reference numerals, and descriptions thereof will be omitted.

[0041] [Configuration] An example configuration of the prediction device 10a according to the second embodiment will be described with reference to Fig. 7. As shown in Fig. 7, the prediction device 10a further includes a group determination unit 134.

[0042] Furthermore, the prediction device 10a acquires information indicating the results of tasks performed by each group having different constituent members, and stores the information in the storage unit 120. For example, the storage unit 120 stores information indicating the results of tasks performed by group A, information indicating the results of tasks performed by group B, information indicating the results of tasks performed by group C, and so on.

[0043] The group determination unit 134 determines and outputs the group with the highest Interaction Score based on the Interaction Score calculated based on the results of the tasks performed by each group. For example, the group determination unit 134 outputs the members of the group with the highest Interaction Score.

[0044] An example of a processing procedure executed by the prediction device 10a will be described with reference to FIG. 8 . For example, the acquisition unit 131 of the prediction device 10a acquires information indicating the results of task execution for each group from the storage unit 120 (S11). The processing of S12 and S13 is similar to S2 and S3 of FIG. 3 , so description thereof will be omitted, and the description will begin with S14 of FIG. 8 . After S13, if there is a group for which the Interaction Score has not yet been calculated (Yes in S14), the process returns to S11. On the other hand, if the prediction device 10a has calculated the Interaction Score for all groups (No in S14), the group determination unit 134 determines and outputs the group with the highest Interaction Score based on the Interaction Score of each group calculated in S13 (S15).

[0045] This allows the prediction device 10a to determine a group that is predicted to have high performance.

[0046] The group determination unit 134 may output a predetermined number of groups in descending order of the calculated Interaction Score, or may output groups whose Interaction Score is equal to or greater than a predetermined value.

[0047] [Other Embodiments] The behavior adjustment amount of a certain member when performing a cooperative task may be defined as in the following formula (3): Formula (3) shows a definition example when the group performing the cooperative task has two members (self and partner).

[0048] (Your current action - your previous action) ÷ (Your action that will maximize your cooperative task performance if the other person takes the same action as your previous action - your previous action) ... Equation (3)

[0049] In other words, the amount of behavioral adjustment may be defined as the degree to which a certain member (oneself) has changed their behavior based on the previous behavior of the other member to maximize the performance of the cooperative task.

[0050] Furthermore, the behavioral adjustment amount may take into account the personality traits of the members. For example, if a member has a stubborn personality, the value of "(1) Do not change behavior" may be added by 0.1 to the behavioral adjustment amount ((1) to (3) above), and the value of "(2) Take an action that will result in the best performance if the other member continues to take the same action" may be subtracted by 0.1.

[0051] The Interaction Score output by the prediction device 10, 10a may be used as a predicted value of the performance level of the same task as the cooperative task used to calculate the behavioral adjustment amount of each member, or may be used as a predicted value of the performance level of a different task.

[0052] For example, consider a case where the prediction device 10, 10a calculates an Interaction Score based on the behavior adjustment amount for the cooperative task of "playing the same card in a deck of cards." In this case, the Interaction Score may be used as a predicted value of the level of performance when each of the members performs the cooperative task of "playing the same card in a deck of cards" (the same cooperative task), or may be used as a predicted value of the level of performance when performing another cooperative task.

[0053] [Experimental Results] Next, we will explain the results of the above-mentioned Interaction Score (IS) evaluation experiment. The cooperative tasks used in this experiment were Blackjack and MonsterSlayer (see Figure 9). The Blackjack task involves playing Blackjack using the opponent's hand, your own hand, and the cards on the table. The MonsterSlayer task is successful if the opponent's hand and your own hand are equal to or greater than the cards on the table in all three parameters.

[0054] In this experiment, we predicted the performance of a cooperative task for 40 pairs of 80 subjects. The 40 pairs of subjects were divided into two groups of 20. The first group performed the Blackjack task before the MonsterSlayer task, and the second group performed the MonsterSlayer task before the Blackjack task.

[0055] The comparative examples are (A) a method for predicting the performance of a collaborative task from the personality traits of each player (member) (Questionnaire), and (B) a method for predicting the performance of a collaborative task from the performance of each player (member) in another task (Performance).To predict the performance of the collaborative task, we used a regression model and leave-one-out cross validation.

[0056] Figure 10 shows the RMSE (Root Mean Squared Error) of the predicted results for the collaborative tasks (Blackjack, MonsterSlayer) by each method. As shown in Figure 10, it was confirmed that Interaction Score had higher prediction accuracy than the comparative example for both the Blackjack and MonsterSlayer tasks.

[0057] From this, it was confirmed that the level of group performance can be predicted more accurately by using a value (Interaction Score) calculated from the empirical rules of behavioral adjustment (amount of behavioral adjustment) of each group member when performing a collaborative task, rather than using the personality traits of each group member.

[0058] [System Configuration, etc.] The components of each unit shown in the figure are conceptual functional units and do not necessarily have to be physically configured as shown. In other words, the specific form of distribution and integration of each device is not limited to that shown, and all or part of them can be functionally or physically distributed and integrated in any unit depending on various loads, usage conditions, etc. Furthermore, all or any part of the processing functions performed by each device can be realized by a CPU and a program executed by the CPU, or can be realized as hardware using wired logic.

[0059] Furthermore, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically using a known method.In addition, the information including the processing procedures, control procedures, specific names, various data and parameters shown in the above documents and drawings can be changed as desired unless otherwise specified.

[0060] [Program] The prediction device 10 can be implemented by installing a program (prediction program) as package software or online software on a desired computer. For example, by executing the program on an information processing device, the information processing device can function as the prediction device 10. The information processing device referred to here includes mobile communication terminals such as smartphones, mobile phones, and PHS (Personal Handyphone Systems), as well as terminals such as PDAs (Personal Digital Assistants).

[0061] 11 is a diagram showing an example of a computer that executes a prediction program. The computer 1000 includes, for example, a memory 1010 and a CPU 1020. The computer 1000 also includes a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0062] The memory 1010 includes a read-only memory (ROM) 1011 and a random access memory (RAM) 1012. The ROM 1011 stores a boot program such as a basic input / output system (BIOS). The hard disk drive interface 1030 is connected to a hard disk drive 1090. The disk drive interface 1040 is connected to a disk drive 1100. A removable storage medium such as a magnetic disk or optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to a mouse 1110 and a keyboard 1120, for example. The video adapter 1060 is connected to a display 1130, for example.

[0063] The hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. That is, the programs that define the processes executed by the prediction device 10 are implemented as program modules 1093 in which computer-executable code is written. The program modules 1093 are stored, for example, in the hard disk drive 1090. For example, the program modules 1093 for executing processes similar to those of the functional configuration of the prediction device 10 are stored in the hard disk drive 1090. Note that the hard disk drive 1090 may be replaced by an SSD (Solid State Drive).

[0064] Data used in the processing of the above-described embodiment is stored as program data 1094, for example, in the memory 1010 or the hard disk drive 1090. The CPU 1020 then reads the program module 1093 or the program data 1094 stored in the memory 1010 or the hard disk drive 1090 into the RAM 1012 as necessary and executes them.

[0065] The program module 1093 and program data 1094 may not necessarily be stored in the hard disk drive 1090, but may also be stored in a removable storage medium and read by the CPU 1020 via the disk drive 1100 or the like. Alternatively, the program module 1093 and program data 1094 may be stored in another computer connected via a network (such as a local area network (LAN) or a wide area network (WAN)). The program module 1093 and program data 1094 may then be read by the CPU 1020 from the other computer via the network interface 1070.

[0066] REFERENCE SIGNS LIST 10, 10a Prediction device 110 Communication unit 120 Storage unit 130 Control unit 131 Acquisition unit 132 First calculation unit 133 Second calculation unit 134 Group determination unit

Claims

1. A prediction method executed by a prediction device, comprising: an acquisition step of acquiring a series of actions taken by each member of a group performing a cooperative task to perform the cooperative task; a first calculation step of calculating, for each member, a behavior adjustment probability indicating the probability that the member will change their next behavior to match the behavior of other members and the probability that they will not change their next behavior, based on the acquired series of actions of each member; and a second calculation step of calculating, for each member, a value obtained by multiplying the probability that the member will not change their next behavior by the probability that other members will change their next behavior to match their next behavior, based on the calculated behavior adjustment probability for each member, and adding up the calculated values ​​for all members of the group, and outputting the sum as a predicted value of the level of performance by the group.

2. The prediction method described in claim 1, characterized in that the behavior adjustment probability further includes, for each member, a probability that the member will change their next behavior so as not to match the behavior of the other members, and the second calculation step outputs a value obtained by subtracting the probability that each member of the group will change their next behavior so as not to match the behavior of the other members from the added value based on the behavior adjustment probability as a predicted value of the level of performance by the group.

3. The prediction method described in claim 1, characterized in that the prediction device further includes a step of comparing predicted values ​​of performance levels for each group obtained by executing the acquisition step, the first calculation step, and the second calculation step for each group with different member composition, and outputting the group with the highest predicted value.

4. A prediction program for causing a computer to execute the following steps: an acquisition step of acquiring a series of actions taken by each member of a group performing a cooperative task to perform the cooperative task; a first calculation step of calculating, for each member, a behavior adjustment probability indicating the probability that the member will change his or her next behavior to match the actions of other members and the probability that the member will not change his or her next behavior, using the acquired series of actions of each member; and a second calculation step of calculating, for each member based on the calculated behavior adjustment probability of each member, a value obtained by multiplying the probability that the member will not change his or her next behavior by the probability that other members will change their next behavior to match the action of the member, and adding up the calculated values ​​for all members of the group to output the sum as a predicted value of the level of performance by the group.

Images