Evaluation device and program

Abstract

To provide an evaluation apparatus configured to visualize a state of a person, and a program.SOLUTION: An evaluation apparatus 1 for evaluating a state of participants participating in a community includes: a target information acquisition unit 11 which acquires identification information identifying each of the participants and at least one of voice information and image information on a conversation of the participants associated with the identification information, as target information; an evaluation unit 13 which evaluates a state of the participants regarding a predetermined index, on the basis of the acquired target information; a detection unit 17 which detects an evaluation result beyond a predetermined threshold, out of results of the evaluation; and an output unit 14 which outputs the detected evaluation result.SELECTED DRAWING: Figure 11

Description

[Technical Field]

[0001] The present disclosure relates to an evaluation device and a program. [Background technology]

[0002] The status of people who belong to an organization is very important when considering the condition of the organization. For example, when considering the condition of an organization, it is very important to maintain the motivation of people who belong to the organization and good relationships between people. As a system that considers people's motivation in this way, a personnel management system that can consider people's life events has been proposed (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6845595 Summary of the Invention [Problem to be solved by the invention]

[0004] However, it is difficult to quantitatively visualize the motivation of individuals and the relationships between individuals. Therefore, it is also difficult to visualize the condition of an organization. As a result, even if the condition or relationships of individuals in an organization deteriorate, it often goes unnoticed, which can have a negative impact on the organizational condition. Therefore, it would be ideal if it were possible to easily judge the condition of individuals.

[0005] The present invention has been made in view of the above-described conventional situation, and has an object to provide an evaluation device and a program that can easily judge the state of a person. [Means for solving the problem]

[0006] The present invention relates to an evaluation device that evaluates the status of participants in a community, and includes a target information acquisition unit that acquires, as target information, identification information that identifies each of the participants and at least one of audio information and image information related to the participants' conversations linked to the identification information; an evaluation unit that evaluates the status of the participants with respect to predetermined indicators based on the acquired target information; a detection unit that detects, from among the evaluation results, evaluation results that deviate from a predetermined threshold; and an output unit that outputs the detected evaluation results.

[0007] It is also preferable that the evaluation unit evaluates the relationship between the participants using evaluation values ​​between the participants.

[0008] It is also preferable that the evaluation unit evaluates the state of the participant himself.

[0009] Furthermore, it is preferable that the target information acquisition unit further acquires the date and time when the participants' conversation took place as the target information, the evaluation unit evaluates the interval at which the participants' conversation took place, and the detection unit detects an interval that is greater than or less than a predetermined value.

[0010] It is also preferable that the evaluation unit evaluates the words contained in the target information, the detection unit detects terms set as inappropriate from the detected words, and the output unit outputs the evaluation result based on the detected terms.

[0011] It is also preferable that the evaluation unit evaluates the actions of the participants included in the target information, the detection unit detects actions set as inappropriate from the detected actions, and the output unit outputs an evaluation result based on the detected actions.

[0012] Furthermore, it is preferable that the evaluation device further includes an output method setting unit that sets an output method and timing of the evaluation result, and the output unit outputs the evaluation result using the output method and timing set by the output method setting unit.

[0013] It is also preferable that the evaluation unit evaluates the transition of the state of the participant at predetermined intervals.

[0014] It is also preferable that the output unit outputs the evaluation results by superimposing them on an organizational chart.

[0015] The present invention also relates to a program that operates a computer as an evaluation device that evaluates the status of participants in a community, and that causes the computer to function as a target information acquisition unit that acquires, as target information, identification information that identifies each of the participants and at least one of audio information and image information related to the participants' conversations linked to the identification information, an evaluation unit that evaluates the status of the participants with respect to predetermined indicators based on the acquired target information, a detection unit that detects, from among the evaluation results, evaluation results that deviate from a predetermined threshold, and an output unit that outputs the detected evaluation results. [Effects of the Invention]

[0016] According to the present disclosure, it is possible to provide an evaluation device and a program that are capable of visualizing a person's condition. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a schematic diagram showing an evaluation system including an evaluation device according to a first embodiment of the present invention. [Figure 2] FIG. 1 is a block diagram showing a configuration of an evaluation device according to a first embodiment. [Figure 3] 4 is a graph showing an example of the content evaluated by the evaluation unit of the evaluation device according to the first embodiment. [Figure 4] 4 is a graph showing an example of the content evaluated by the evaluation unit of the evaluation device according to the first embodiment. [Figure 5] 4 is a graph showing an example of the content evaluated by the evaluation unit of the evaluation device according to the first embodiment. [Figure 6] 4 is a table showing an example of the content evaluated by the evaluation unit of the evaluation device of the first embodiment. [Figure 7]3 is a schematic diagram showing an example of an organizational chart output by an output unit of the evaluation device according to the first embodiment. FIG. [Figure 8] FIG. 4 is a schematic diagram showing another example of an organizational chart output by the output unit of the evaluation device according to the first embodiment. [Figure 9] FIG. 10 is a schematic diagram showing yet another example of an organizational chart output by the output unit of the evaluation device according to the first embodiment. [Figure 10] 4 is a flowchart showing the operation of the evaluation device according to the first embodiment. [Figure 11] FIG. 10 is a block diagram showing the configuration of an evaluation device according to a second embodiment of the present invention. [Figure 12] 10 is a table showing an example of the content output by the output unit of the evaluation device according to the second embodiment. [Figure 13] 10 is a flowchart showing the flow of operations of the evaluation device according to the second embodiment. [Figure 14] FIG. 10 is a screen diagram showing the output contents of an evaluation device according to a modified example. [Figure 15] FIG. 10 is a screen diagram showing the output contents of an evaluation device according to a modified example. [Figure 16] FIG. 10 is a screen diagram showing the output contents of an evaluation device according to a modified example. [Figure 17] FIG. 10 is a screen diagram showing the output contents of an evaluation device according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0018] An evaluation device 1 and a program according to each embodiment of the present invention will be described below with reference to FIGS. First, an overview of the evaluation device 1 and the program according to this embodiment will be described.

[0019] The evaluation device 1 is a device that evaluates participants (hereinafter also referred to as participants) belonging to an organization (community) using, for example, records of online conversations between the participants. The evaluation device 1 evaluates at least one of the status of the participants and the relationships between the participants using, for example, records of conversations between superiors and subordinates or conversations between colleagues. The evaluation device 1 visualizes the status of the participants and the relationships between the participants by outputting the evaluation results. In this way, the evaluation device 1 can easily show the status and relationships of the participants in the organization.

[0020] [First embodiment] Next, an evaluation device 1 and a program according to a first embodiment of the present invention will be described with reference to FIGS. First, an evaluation system 100 including the evaluation device 1 will be described.

[0021] 1, the evaluation system 100 includes a plurality of user terminals 101 and an evaluation device 1. The evaluation system 100 is configured as a part of an in-house network, for example.

[0022] The user terminal 101 is, for example, an information communication terminal such as a PC or a smartphone. The user terminal 101 is connected to other user terminals 101 via a network N. In this embodiment, the user terminal 101 is connected to other user terminals 101 via the Internet, for example. A user terminal 101 is assigned to each participant, for example, and is used for carrying out conversations between the participants. The user terminal 101 stores, as target information, identification information for identifying the participant, and audio information and image information related to the conversation between the participants.

[0023] Next, the evaluation device 1 according to this embodiment will be described. The evaluation device 1 is a device that evaluates participants in a conversation. As shown in Fig. 2, the evaluation device 1 includes an evaluation result storage unit 10, a target information acquisition unit 11, a related information acquisition unit 12, an evaluation unit 13, an output unit 14, and an output content switching unit 15.

[0024] The evaluation result storage unit 10 is, for example, a recording device such as a hard disk. The evaluation result storage unit 10 stores the evaluation result for each participant. The evaluation result storage unit 10 may also store acquired target information, which will be described later.

[0025] The target information acquisition unit 11 is realized by, for example, the operation of a CPU. The target information acquisition unit 11 acquires, as target information, identification information that identifies each participant and at least one of audio information and image information related to the conversation of the participants linked to the identification information. The target information acquisition unit 11 acquires, for example, target information from each user terminal 101. The target information acquisition unit 11 acquires, for example, target information related to conversations that took place within a predetermined period. The target information acquisition unit 11 also acquires information on responses to a questionnaire provided by the participants. In this embodiment, an example will be described in which the target information acquisition unit 11 acquires both audio information and image information as target information.

[0026] The relationship information acquisition unit 12 is realized, for example, by the operation of a CPU. The relationship information acquisition unit 12 acquires relationship information indicating relationships between participants within a specified organization. The relationship information acquisition unit acquires, for example, network data defining the relationships between participants and the positions of the participants in the network data as relationship information. The relationship information acquisition unit 12 acquires, as network data, relationship information such as an organizational chart, a table showing relationships between colleagues, a table showing relationships with external parties, or a table showing the composition of a project team. In this embodiment, an example will be described in which the relationship information acquisition unit 12 acquires an organizational chart as relationship information.

[0027] The evaluation unit 13 is realized by, for example, the operation of a CPU. The evaluation unit 13 evaluates at least one of the status of the participants and the relationships between the participants based on the acquired target information. In this embodiment, the evaluation unit 13 evaluates both the status of the participants and the relationships between the participants based on at least one of audio information and image information of the participants included in the acquired target information. Note that in this embodiment, the evaluation unit 13 evaluates both the status of the participants and the relationships between the participants based on target information including both audio information and image information. As the relationships between the participants, the evaluation unit 13 evaluates the strength of the connections between the participants in terms of quality and quantity. Furthermore, the evaluation unit 13 evaluates the relationships between the participants using an evaluation value of the connections between the participants. Specifically, the evaluation unit 13 evaluates the relationships between the participants using a balance between multiple evaluation values ​​of the connections between the participants. In this embodiment, the evaluation unit 13 evaluates the relationships between the participants using an evaluation value of the connections between the participants as people and an evaluation value of the connections. Furthermore, in this embodiment, the evaluation unit 13 evaluates the relationships between the participants using a balance between an evaluation value of the connections between the participants as people and an evaluation value of the connections. The evaluation unit 13 includes a participant state evaluation unit 131, a person-related evaluation unit 132, a work-related evaluation unit 133, and a relevance evaluation unit .

[0028] The participant state evaluation unit 131 evaluates the state of the participant. For example, the participant state evaluation unit 131 evaluates the stress, motivation, psychological safety, engagement, and sense of growth of the participant. Here, engagement can be thought of as an employee's attachment to the company, their desire to contribute, a sense of belonging, or a feeling of attachment. Engagement is , others Similar to participant motivation and psychological safety, AI analysis may be performed using video and other data. The participant state evaluation unit 131, for example, evaluates whether or not a participant is stressed using the participant's facial expression. The participant state evaluation unit 131 evaluates whether or not a participant is stressed based on, for example, how often the participant smiles. The participant state evaluation unit 131 evaluates that the more frequently a participant smiles, the less stressed the participant is. Furthermore, the participant state evaluation unit 131 makes a judgment based not only on the frequency of smiles but also on the quality (shape, size of gestures, etc.).

[0029] The participant state evaluation unit 131 also evaluates the motivation of a participant using the participant's concentration level. The participant state evaluation unit 131 evaluates the motivation of a participant using the concentration level calculated from the participant's line of sight, center of gravity, etc. The participant state evaluation unit 131 calculates the concentration level, for example, by quantifying the direction of the line of sight and the position of the center of gravity. Here, the center of gravity refers to the degree of swaying of the participant's body and changes in the distance of the body from the screen.

[0030] The participant state evaluation unit 131 also evaluates psychological safety, a sense of growth, and the like by quantifying the responses to the questionnaire. The participant state evaluation unit 131 may also comprehensively assess psychological safety, a sense of growth, and the like using other information (video information, past work history, personality assessment at the time of joining the company, healthcare product information, and the like). The participant state evaluation unit 131, for example, plots the assessed stress, motivation, psychological safety, engagement, and sense of growth into a radar chart. The participant state evaluation unit 131 determines whether the participant's state is good or bad based on whether the plotted radar chart is close to a predetermined shape. The participant state evaluation unit 131 evaluates the participant's state based on, for example, whether the plotted radar chart is closer to a square. The participant state evaluation unit 131 evaluates the participant by, for example, quantifying a radar chart that is closer to a square as a good state. The participant state evaluation unit 131, for example, represents the participant's state by replacing it with a color. For example, the participant status evaluation unit 131 expresses the participant's status as blue, yellow, red, etc., in order as the participant's status progresses from good to bad. The participant status evaluation unit 131 stores the evaluation results of the participant in the evaluation result storage unit 10, linking them to the participant's identification information.

[0031] The person-related evaluation unit 132 evaluates the personal connections between the participants. For example, the person-related evaluation unit 132 evaluates the personal connections between the participants and calculates an evaluation value. As shown in FIG. 3, the person-related evaluation unit 132 calculates the degree of empathy of the participant from, for example, the facial expression and behavior of the participant. For example, the person-related evaluation unit 132 determines that the degree of empathy of the participant is high when the facial expression and behavior are cheerful.

[0032] Furthermore, the person-related evaluation unit 132 calculates a dialogue ratio from the amount of speech of the participants. For example, the person-related evaluation unit 132 determines that the dialogue ratio is high when the amount of speech of the participants is large. The person-related evaluation unit 132 stores an evaluation of the personal connection in the evaluation result storage unit 10, linking it to the identification information of each participant.

[0033] The work-related evaluation unit 133 evaluates the work-related connections between the participants. For example, the work-related evaluation unit 133 evaluates the work-related connections between the participants and calculates an evaluation value. As shown in Fig. 3, the work-related evaluation unit 133 calculates the concentration level from the line of sight and the position of the center of gravity of the participant. For example, the work-related evaluation unit 133 determines that the concentration level is high when the line of sight and the position of the center of gravity are close to the direction toward the user terminal 101.

[0034] Furthermore, the work-related evaluation unit 133 calculates the information divergence from the utterance content. For example, the work-related evaluation unit 133 calculates the information divergence (degree of information divergence) of the participant's utterance content by quantifying the degree of relevance of the utterance content (relevance for each utterance content). For example, when the degree of relevance of the utterance content is high, the work-related evaluation unit 133 evaluates the information divergence of the participant's utterance content as low. The work-related evaluation unit 133 stores the evaluation of work-related connections in the evaluation result storage unit 10, linking it to the identification information of each participant.

[0035] The relevance evaluation unit 134 evaluates the relevance of the participants. For example, the relevance evaluation unit 134 evaluates the relevance of the participants by combining personal relevance and work-related relevance. As shown in FIGS. 4 and 5, the relevance evaluation unit 134 evaluates the participants using a quadrilateral (square) table with empathy level, dialogue ratio, concentration level, and information divergence arranged on each side. As shown in FIG. 5, the relevance evaluation unit 134 determines that the relevance of the participants is good when the four indicators are close to the center of the square. On the other hand, the relevance evaluation unit 134 determines that the relevance of the participants is poor when the four indicators are off the center of the square. The relevance evaluation unit 134 quantifies the relationship between the participants, for example, by calculating the deviation between the four indicators. Furthermore, for example, with regard to the personal connection and work connection shown in FIGS. 3 to 5, the closer the indicators are to each other, the better the relationship is. As shown in FIG. 6, the relevance evaluation unit 134 evaluates the personal connection and work connection as the state of the relationship. The relevance assessment unit 134 also evaluates the relationships using colors based on the numerical values ​​of the personal connections and the work connections. The relevance assessment unit 134 stores the evaluation result obtained by combining the personal connections and the work connections in the evaluation result storage unit 10, linked to the identification information of each participant. Note that "connections" refer to the relationships between participants in a community (organization).

[0036] The output unit 14 is realized, for example, by the operation of a CPU. The output unit 14 outputs the evaluation results in association with network data. In this embodiment, the output unit 14 outputs the evaluation results by superimposing them on an organizational chart. For example, as shown in FIG. 7, the output unit 14 outputs the evaluation results by associating the color-coded status of participants and the color-coded relationships between participants on the organizational chart. The output unit 14 outputs the evaluation results of the participants by superimposing them on the positions of the participants in the organizational chart. Furthermore, the output unit 14 outputs the evaluated relationships between the participants by superimposing them on the lines connecting the participants in the organizational chart. The output unit 14 reads, for example, the evaluation results of the participants' status, personal connections, work connections, and the combined overall connections stored in the evaluation result storage unit 10. The output unit 14 uses identification information included in the read evaluation results to display the evaluation results by superimposing them on the positions and connections of the participants in the organizational chart. Note that in FIG. 8, the output unit 14 indicates that the thicker the lines, the stronger the relationships. 8, the output unit 14 indicates that the thicker the line is for the status of the participant, the better the status is. Furthermore, instead of evaluating by color, the relevance evaluation unit 134 may output the evaluation result by marking, flashing, or enclosing (with a wavy line, dotted line, or the like) the organizational chart (network diagram), adding animation to the object, or inserting a message as a pop-up object such as a speech bubble outside the object, or the like.

[0037] The output content switching unit 15 is realized by, for example, the operation of a CPU. The output content switching unit 15 switches the content of output by the output unit 14. For example, as shown in FIG. 8, the output content switching unit 15 switches to an output in which the status and relationships of participants at a specific time (for example, a busy period) are superimposed on an organizational chart, and causes the output unit 14 to output this. Furthermore, for example, as shown in FIG. 9, the output content switching unit 15 switches to an output in which work connections and motivation are superimposed on an organizational chart, and causes the output unit 14 to output this. Furthermore, the output content switching unit 15 switches to an output in which human connections and stress are superimposed on an organizational chart, and causes the output unit 14 to output this.

[0038] Next, the flow of operations of the evaluation device 1 will be described with reference to the flowchart of FIG. First, the target information acquisition unit 11 acquires target information from the user terminal 101 (step S1). Next, the relationship information acquisition unit 12 acquires relationship information (step S2). Next, the participant status evaluation unit 131 evaluates the status of the participant based on the acquired target information (step S3). The participant status evaluation unit 131 stores the evaluation result in the evaluation result storage unit 10.

[0039] Next, the person-related evaluation unit 132 evaluates the personal relationships between the participants (step S4). The person-related evaluation unit 132 stores the evaluation result in the evaluation result storage unit 10.

[0040] Next, the work-related evaluation unit 133 evaluates the work-related relevance between the participants (step S5). The work-related evaluation unit 133 stores the evaluation result in the evaluation result storage unit 10.

[0041] Next, the relevance assessment unit 134 assesses the personal relevance and the work relevance in combination (step S6). The relevance assessment unit 134 stores the assessment result in the assessment result storage unit 10.

[0042] The output unit 14 outputs the evaluation results superimposed on the organizational chart (step S7). The output unit 14, for example, displays the evaluation results superimposed on the organizational chart on a display or the like.

[0043] Next, it is determined whether or not to switch the output content (step S8). If the output content is to be switched (step S8: YES), the process proceeds to step S9. On the other hand, if the output content is not to be switched, the process according to this flow ends.

[0044] In step S9, the output content switching unit 15 switches the output content and outputs it to the output unit. Next, it is determined whether or not to switch the output content again (step S10). If the output content is to be switched (step S10: YES), the process returns to step S9. On the other hand, if the output content is not to be switched (step S10: NO), the process according to this flow ends.

[0045] Next, the program will be described. Each component included in the evaluation device 1 can be realized by hardware, software, or a combination of these. Here, being realized by software means being realized by a computer reading and executing a program.

[0046] The program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The display program may also be supplied to a computer by various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The transitory computer-readable medium can supply the program to a computer via a wired communication path such as an electric wire or optical fiber, or via a wireless communication path.

[0047] As described above, the evaluation device 1 and the program according to this embodiment have the following advantages. (1) An evaluation device 1 for evaluating participants in a conversation includes a target information acquisition unit 11 that acquires, as target information, identification information for identifying each participant and at least one of audio information and image information related to the participant's conversation linked to the identification information, an evaluation unit 13 that evaluates at least one of the participant's status and the relationship between the participants based on the acquired target information, and an output unit 14 that outputs the evaluation result. This makes it possible to output the evaluation result related to the person's status, thereby making it easy to determine the person's status.

[0048] (2) The evaluation unit 13 evaluates the relationships between the participants using the evaluation values ​​of the connections between the participants. This allows the relationships between the participants to be evaluated from multiple perspectives, making it possible to determine the relationships between the participants in more detail.

[0049] (3) The evaluation unit 13 evaluates the relationships between the participants using the balance of multiple evaluation values ​​of the connections between the participants, thereby making it possible to make a judgment from the perspective of a better balance of the connections between the participants.

[0050] (4) The output unit 14 outputs the evaluation results in association with the organizational chart, thereby making it possible to easily visualize the relationships between participants and the status of each participant.

[0051] (5) The output unit 14 outputs the color-coded statuses of the participants and the relationships between the color-coded participants in association with the organizational chart, thereby making it easier to visualize the relationships between the participants and the statuses of the participants.

[0052] (6) An evaluation device 1 for evaluating the status of participants in a community includes a target information acquisition unit 11 that acquires, as target information, identification information for identifying each participant and at least one of audio information and image information related to the participants' conversations linked to the identification information, a relationship information acquisition unit 12 that acquires relationship information indicating the relationships between the participants within a predetermined organization, an evaluation unit 13 that evaluates the status of the participants and the relationships between the participants based on the participants' audio information included in the acquired target information, and an output unit 14 that outputs the evaluation result in association with the relationship information. This makes it possible to easily determine the status of the participants and the relationships between the participants.

[0053] (7) The relationship information acquisition unit 12 acquires, as relationship information, network data defining the relationships between participants and the positions of participants in the network data, and the output unit 14 outputs the relationships between participants and evaluation results in association with the network data. This makes it easier to determine the positions in the organization, the status of participants, and the relationships between participants.

[0054] (8) The evaluation device 1 further includes an output content switching unit 15 that switches the content output by the output unit, and the output unit outputs different content of the evaluation results based on the switching by the output content switching unit 15. This makes it possible to determine the status of participants and the relationships between participants from multiple perspectives.

[0055] [Second embodiment] Next, an evaluation device 1 and a program according to a second embodiment of the present invention will be described with reference to Fig. 11 to Fig. 13. In describing the second embodiment, the same components will be denoted by the same reference numerals, and their description will be omitted or simplified.

[0056] The evaluation device 1 and program according to the second embodiment are devices capable of communicating abnormalities and best practices to human resources, management, etc. The evaluation device 1 and program according to the second embodiment have a function of separately outputting evaluation results that exceed predetermined conditions. The evaluation device 1 and program according to the second embodiment are devices that enable the detection of, for example, depression tendencies, deterioration in condition, and good condition. The evaluation device 1 outputs such detections to the outside as a warning (alert).

[0057] The evaluation device 1 according to the second embodiment differs from the first embodiment in that it further includes an index acquisition unit 16, a detection unit 17, and an output method setting unit 18. The evaluation device 1 according to the second embodiment also differs from the first embodiment in that the target information acquisition unit 11 further acquires the date and time when the participants had a conversation. The evaluation device 1 according to the second embodiment also differs from the first embodiment in that the evaluation unit 13 evaluates the state of the participants regarding predetermined indices based on the acquired target information. The evaluation device 1 according to the second embodiment also differs from the first embodiment in that the evaluation unit 13 further includes a conversation interval evaluation unit 135. The evaluation device 1 according to the second embodiment also differs from the first embodiment in that the output unit separately outputs the detection results. The indices evaluated by the evaluation unit 13 may include the evaluation results according to the first embodiment.

[0058] The index acquiring unit 16 is realized, for example, by the operation of a CPU. The index acquiring unit 16 acquires a predetermined index and a threshold value for the predetermined index. The index acquiring unit 16 acquires, for example, the index and threshold value stored in advance in the evaluation device 1.

[0059] The evaluation unit 13 evaluates the state of the participant in relation to the acquired indicators and thresholds. Specifically, the participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, and the relevance evaluation unit 134 evaluate the state of the participant in relation to the acquired indicators and thresholds. The participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, and the relevance evaluation unit 134 evaluate, for example, whether or not the participant is prone to depression, whether or not the participant's condition is good, and whether or not the relationships between participants are good, as indicators. Furthermore, the participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, and the relevance evaluation unit 134 evaluate the state of the participant by comparing each indicator with the acquired thresholds. For example, the participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, and the relevance evaluation unit 134 evaluate an indicator that exceeds the threshold on the good side as a good indicator. On the other hand, the participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, and the relevance evaluation unit 134 evaluate, for example, an index that exceeds a threshold value on the negative side as a negative index.

[0060] The conversation interval evaluation unit 135 evaluates the intervals at which participants have conversations. For example, the conversation interval evaluation unit 135 evaluates the interval from the last date and time to the next date and time for the timing of a conversation between participants. That is, the conversation interval evaluation unit 135 evaluates the interval from the last date and time to the date and time of the conversation included in the acquired target information for the conversation between participants. For example, the conversation interval evaluation unit 135 acquires the evaluation results of conversations that have already been evaluated from the evaluation result storage unit. The conversation interval evaluation unit 135 evaluates the interval between the acquired evaluation results and the conversation included in the acquired target information. For example, the conversation interval evaluation unit 135 evaluates the timing of conversations that are shorter than a threshold value as being close. On the other hand, the conversation interval evaluation unit 135 evaluates the timing of conversations that are longer than the threshold value as being diluted.

[0061] Furthermore, the conversation interval evaluation unit 135 evaluates the interval between the previous date and time and the current date and time for the timing of conversations between participants. The conversation interval evaluation unit 135, for example, acquires the date and time of the previous conversation from the evaluation results stored in the evaluation result storage unit 10. The conversation interval evaluation unit 135 evaluates the interval between the acquired previous date and time and the date and time of evaluation. For example, the conversation interval evaluation unit 135 evaluates that the timing of conversations with an interval longer than a threshold value is diluted.

[0062] The detection unit 17 is realized by, for example, the operation of a CPU. The detection unit 17 detects evaluation results (indicators) that deviate from a predetermined threshold among the evaluation results. The detection unit 17 detects evaluation results that deviate from a threshold, for example, regarding the quality of the participant's condition and the quality of the relationship between the participants. Specifically, the detection unit 17 detects a participant's tendency toward depression, a deterioration in condition, a good condition, a deterioration in relationship, and an improvement in relationship.

[0063] Furthermore, the detection unit 17 evaluates the intervals between conversations between participants and detects intervals that are greater than or less than a predetermined value. The detection unit 17 detects, for example, intervals between conversations that are greater than or less than a threshold value, thereby detecting whether the relationship between participants is weakening or whether the relationship between participants is becoming closer. The detection unit 17 also detects intervals greater than or equal to the threshold value between conversations that have not been held up until the evaluation date and time. In this way, the detection unit 17 detects whether the relationship between participants is weakening.

[0064] The output method setting unit 18 is realized, for example, by the operation of a CPU. The output method setting unit 18 sets the output method and timing of the evaluation results. For example, the output method setting unit 18 accepts settings regarding the output method and timing and sets the output method and timing by the output unit. For example, the output method setting unit 18 sets at least one of only displaying on a dashboard showing the status of the participants, notification by email, and notification by report. Furthermore, for example, the output method setting unit 18 sets at least one of immediately after evaluation and after batch processing as the timing.

[0065] The output unit 14 outputs the detected evaluation results (indexes). The output unit 14 outputs the detected evaluation results separately from the evaluation results of the first embodiment. For example, the output unit 14 outputs the detected evaluation results using the output method and timing set by the output method setting unit 18. For example, as shown in FIG. 12, the output unit 14 outputs the evaluation results in descending order of the degree of deviation from the threshold.

[0066] Next, the operation of the evaluation device 1 will be described with reference to the flowchart of FIG.

[0067] First, the target information acquisition unit 11 acquires target information (step S11). Next, the participant state evaluation unit 131, the person-related evaluation unit 132, the work-related evaluation unit 133, the relevance evaluation unit 134, and the conversation interval evaluation unit 135 acquire predetermined indices and thresholds (step S12).

[0068] Next, the participant state evaluation unit 131 evaluates the state of the participant based on a predetermined index (step S13). The participant state evaluation unit 131 also stores the evaluation result in the evaluation result storage unit 10. Next, the person-related evaluation unit 132 evaluates the personal connections between the participants based on a predetermined index (step S14). The person-related evaluation unit 132 stores the evaluation result in the evaluation result storage unit 10. Next, the work-related evaluation unit 133 evaluates the work-related connections between the participants based on a predetermined index (step S15). The person-related evaluation unit 132 stores the evaluation result in the evaluation result storage unit 10. Next, the relevance evaluation unit 134 evaluates the personal relevance and the work relevance in combination (step S16). The relevance evaluation unit 134 stores the evaluation result in the evaluation result storage unit 10. Next, the conversation interval evaluation unit 135 evaluates the conversation interval between the participants based on a predetermined index (step S17). The conversation interval evaluation unit 135 stores the evaluation result in the evaluation result storage unit 10 .

[0069] Next, the detection unit 17 determines whether or not there is an evaluation result that falls outside the threshold value among the evaluated indices (step S18). If there is an evaluation result that falls outside the threshold value, the process proceeds to step S19. On the other hand, if there is no evaluation result that falls outside the threshold value, the process according to this flow ends.

[0070] In step S19, the detection unit 17 detects an evaluation result that falls outside the threshold. Next, the output unit acquires the output method set by the output method setting unit 18 (step S20). Next, the output unit 14 outputs the detected evaluation result in accordance with the set output method (step S21). This completes the operation of this flow.

[0071] As described above, the evaluation device 1 and the program according to this embodiment have the following advantages. (9) An evaluation device 1 for evaluating the status of participants in a community includes a target information acquisition unit 11 that acquires, as target information, identification information for identifying each participant and audio information and image information related to the participant's conversations linked to the identification information; an evaluation unit 13 that evaluates the status of the participants with respect to predetermined indicators based on the acquired target information; a detection unit 17 that detects, among the evaluation results, evaluation results that deviate from a predetermined threshold; and an output unit 14 that outputs the detected evaluation results. This makes it possible to visualize the status of a person. In particular, since the device can detect and output abnormalities and best practices as evaluation results, it is possible to provide a device that contributes to early response to situations.

[0072] (10) The target information acquisition unit 11 further acquires the date and time when the participants had a conversation as target information, the evaluation unit 13 evaluates the intervals between the participants' conversations, and the detection unit 17 detects intervals that are longer or shorter than a predetermined interval. This makes it possible to detect whether the relationship between the participants is becoming closer or weaker. Therefore, it is possible to detect indicators that will help maintain better relationships between the participants.

[0073] (11) The evaluation device 1 further includes an output method setting unit 18 that sets the output method and timing of the evaluation results, and the output unit 14 outputs the evaluation results using the output method and timing set by the output method setting unit 18. This allows the detected evaluation results to be output using the output method and timing set separately from general evaluation results. Therefore, it is possible to output the results in a way that makes it easy to identify the occurrence of anomalies or best practices.

[0074] [Third embodiment] Next, an evaluation device 1 and a program according to a third embodiment of the present invention will be described. In describing the third embodiment, the same components will be denoted by the same reference numerals, and the description thereof will be omitted or simplified.

[0075] The evaluation device 1 according to the third embodiment detects the occurrence of harassment behavior during a conversation between participants, thereby making it possible to prompt early action against harassment behavior.

[0076] The evaluation device 1 according to the third embodiment differs from the first and second embodiments in that the person-related evaluation unit 132 evaluates words included in the target information (audio information). The evaluation device 1 according to the third embodiment also differs from the first and second embodiments in that the detection unit 17 detects terms evaluated as inappropriate from the detected words. The evaluation device 1 according to the third embodiment differs from the first and second embodiments in that the output unit 14 outputs an evaluation result based on the detected terms.

[0077] The person-related evaluation unit 132 evaluates the words by, for example, extracting words included in the audio information of the target information. The person-related evaluation unit 132 evaluates the extracted words by, for example, calculating the similarity of the extracted words with respect to terms that have been set in advance as inappropriate. The detection unit 17 detects, for example, a phrase whose calculated similarity is evaluated to be equal to or greater than a predetermined threshold value.

[0078] As described above, the evaluation device 1 and the program according to this embodiment have the following advantages. (12) The evaluation unit 13 evaluates the words included in the target information, the detection unit 17 detects terms set as inappropriate from the detected words, and the output unit 14 outputs the evaluation result based on the detected terms. This makes it possible to easily detect emergency situations that require immediate action. In addition, the detected results can be output in a way that makes them easy to distinguish.

[0079] Although preferred embodiments of the evaluation device and program of the present invention have been described above, the present disclosure is not limited to the above-described embodiments and can be modified as appropriate.

[0080] For example, in the above embodiment, the output unit 14 may be configured to output the results of conversations held, as shown in Figures 14 and 15. Furthermore, the output unit 14 may output the status of conversations held by each participant and the transition of relationships with other participants, as shown in Figure 16. Furthermore, the output unit 14 may be configured to output the amount of speech, information dissemination, and transition of conditions between participants, as shown in Figure 17. This makes it possible to determine not only the instantaneous relationships, but also the state of the participants and the transition of the relationships between the participants.

[0081] In the above embodiment, the participants belong to a single organization, but the present invention is not limited to this. The participants may belong to any organization as long as they are able to converse using information processing terminals. For example, the participants may be a person who belongs to a single organization and a business partner. Also, for example, the participants may be a person who has received a job offer and a person from that company. Also, the participants may be a person who is joining a specific organization for the first time and a person who is already a member of the organization.

[0082] In the above embodiment, there may be multiple participants. For example, the meeting may be a group meeting in which each participant participates using one user terminal 101. As long as the target information acquisition unit 11 can acquire corresponding identification information, audio information, and image information from each user terminal 101, there is no limit to the number of participants or the content of the conversation.

[0083] In the above embodiment, the relevance assessment unit 134 quantifies the relationship between the participants by calculating the deviation between the four indices, but this is not limiting. The relevance assessment unit 134 may evaluate the relationship between the participants using the variance or distribution of four or more indices. The relevance assessment unit 134 may also perform evaluation using a polygon having a number of sides corresponding to the number of indices. The relevance assessment unit 134 may also evaluate the relationship between the participants according to the conditions of each individual indices without using a polygon.

[0084] The relevance evaluation unit 134 may determine the evaluation based on predetermined criteria corresponding to the index, without being limited to balance (deviation), variance, or distribution. For example, when the supervisor is in a coaching phase with a subordinate, the relevance evaluation unit 134 preferably evaluates the supervisor's speaking ratio based on the criterion that a higher ratio is better, rather than a balanced ratio. Furthermore, the relevance evaluation unit 134 preferably evaluates, for example, psychological safety, empathic behavior (nodding, empathizing with the other person's feelings, mirroring facial expressions or speaking speed, etc.), and approving behavior (words of approval, not saying negative words, not interrupting the other person, etc.) based on the criterion that a higher ratio is generally better. The relevance evaluation unit 134 may change the criteria depending on the state, personality, etc. of the person being evaluated.

[0085] In addition, the relevance evaluation unit 134 considers that, for example, if the concentration gauge is too high, there is a risk of a tendency toward power harassment, so a middle ground is best, whereas in interviews with a lot of ice-breaking between subordinates and superiors (such as at the beginning of assignment), it may be judged that a low concentration gauge is more relaxed and desirable, so it is preferable to evaluate based on the criterion that a lower concentration gauge is better.

[0086] Furthermore, the relevance assessment unit 134 may flexibly change the criteria depending on the attributes of the participant or data associated with the participant (past actions (history of taking leave from work, behavior in past videos, etc.), current state (healthcare product indicators, etc.), or thoughts (surveys, etc.)). For example, the relevance assessment unit 134 may change the desired value for a person who has taken leave from work in the past due to mental health problems or a person who continues to work excessive hours.

[0087] The relevance evaluation unit 134 may also change the criteria for evaluating the level of one indicator depending on the level of the other indicator. For example, if the relevance evaluation unit 134 determines from a video that a person tends to stop speaking when spoken to quickly, it may change the indicator appropriately. The relevance evaluation unit 134 may also change the indicator appropriately for a person who is determined to be in poor condition based on healthcare product and attendance data acquired using videos of the day or the most recent day, surveys, etc. The relevance evaluation unit 134 may also change the indicator appropriately based on a personality diagnosis, stress check, personnel evaluation information (including 360-degree evaluation, etc.), performance information (sales performance, etc.), frequency and quality of communication with other members, attributes (job title, educational background, department, etc.), data associated with the participant (past actions (history of leave of absence, behavior in past videos, etc.), current status (healthcare product indicators, etc.), thoughts, etc. (surveys, etc.)), connection trends seen from an organizational chart (e.g., many or few members involved in work), output results of healthcare products, etc.). The relevance assessment unit 134 may also adjust the evaluation value according to the characteristics and implicit roles of the members constituting the organization (for example, there is a concept called buff personnel, who are "members who do not make a high level of direct or visible contribution to work but who make a high level of indirect or implicit contribution to work by, for example, lightening the mood"). Note that the relevance assessment unit 134 acquires the characteristics and implicit roles of the members constituting the organization from, for example, videos, questionnaires, attributes, data associated with participants (past actions (history of leave of absence, behavior in past videos, etc.), current status (healthcare product indicators, etc.), or thoughts (surveys, etc.)), or other data, etc.

[0088] For example, the relevance assessment unit 134 may use attentive listening behavior and approving behavior as indicators of the superior's behavior. Furthermore, for example, the relevance assessment unit 134 may use proactiveness, openness, and concentration as indicators of the subordinate's behavior. For example, the relevance assessment unit 134 may use quantity (e.g., the number of conversations held, the length of conversations, etc.) as an indicator of the relationship. For example, the relevance assessment unit 134 may use nodding and backchanneling (mirroring of display, speech rate, and posture changes) as features indicating the level of attentive listening. For example, the relevance assessment unit 134 may use affirmative or agreeing words, negative words, and the number of interruptions as features indicating approving behavior.

[0089] For example, the relevance assessment unit 134 may use activity (positive words) and the amount of comments as features indicating proactiveness. For example, the relevance assessment unit 134 may use activity (negative words) and the amount of comments as features indicating openness. For example, the relevance assessment unit 134 may use eye movement and body sway as features indicating concentration. For example, the relevance assessment unit 134 may use frequency, amount, and dialogue ratio as features indicating amount.

[0090] To achieve the above, the evaluation device 1 may include a content identification unit (not shown) that identifies the content of at least one of the audio information and the image information included in the target information, a participant information identification unit (not shown) that identifies participant information including the status and attributes of the participants, an index determination unit (not shown) that determines indexes to be used for evaluation based on the identified content and the participant information, and an optimal value acquisition unit (not shown) that acquires predetermined optimal values ​​for the determined indexes. The relevance evaluation unit 134 may evaluate the relationships between the participants using the acquired optimal values ​​and evaluation values ​​for the determined indexes of the participants included in the target information. In this case, the evaluation unit 13 may not include the person-related evaluation unit 132 and the work-related evaluation unit 133. The index determination unit may also determine externally specified indexes as the indexes to be used for evaluation. The participant information identification unit may also identify pre-entered attributes, questionnaire results, participants' health conditions, participants' emotions, etc. as participant information. In addition, the participant information identification unit may identify past video information, healthcare product information, company attribute information, personnel data (past evaluations, etc.), frequency and quality of communication with other members, and connection trends as seen from the organizational chart (such as whether there are many or few members involved in the work), as participant information.

[0091] Here, the index determination unit may determine the indexes using at least one of the specified content related to the target information and the participant information. Furthermore, the index determination unit may determine the priority of the indexes to be used for evaluation based on the specified content and the participant information, and determine specific indexes based on the priority and a predetermined number of indexes to be used. For example, the index determination unit may determine a specified number of indexes as the indexes to be used for evaluation from the indexes whose priorities have been determined for the specified content. As an example, consider a case where the content determination unit determines that the content is a warning from a superior to a subordinate, and the participant information determination unit determines that the participant's personality is shy. In this case, the index determination unit may prioritize indicators such as whether the participant is overbearing, whether the speaking speed is fast, whether intimidating language is used, and whether the conversation is one-sided, as the indexes to be used for evaluation. Furthermore, the index determination unit may prioritize indicators such as whether the participant is timid, whether there are any breathing abnormalities, and whether the participant is aggressive, as the indexes to be used for evaluation.

[0092] As another example, consider a case where the content identification unit determines that the conversation is about work between colleagues, and the participant information identification unit identifies the participants as being in the same period and as being cheerful. In this case, the index determination unit determines, as the index to be used preferentially for evaluation, indexes such as whether the conversation is serious, whether the intervals between the participants are appropriate, whether the participants are appropriately diverging in information, and whether the conversation is lively.

[0093] Furthermore, for example, in the case of a coaching conversation, the AI ​​analysis results may reveal that a pattern in past conversations (with the same boss or a different boss) in which the subordinate talks in the first half and the boss speaks in the second half has a positive effect on each indicator of the subordinate (smile, concentration, engagement, subsequent work results, etc.). In this case, the indicator determination unit may determine the indicator for giving priority to advice to the boss about the speaking rate and its transition rate as the indicator for evaluation.

[0094] In addition, the relevance assessment unit 134 may assess the relevance by taking into account the attributes of the participant (for example, personality, year of joining the company, new graduate / mid-career, job title, educational background, department of experience, internal survey data, data obtained during recruitment activities (including videos), etc.) or data associated with the participant (past actions (history of leave of absence, words and actions in past videos, etc.), current state (healthcare product indicators, etc.), or thoughts, etc.).

[0095] Furthermore, in the above embodiment, the target information includes image information, but is not limited to this. The target information may be only audio information. In the case of only audio information, the evaluation unit 13 may use voice tremor, tone of voice, volume, and pitch instead of eye gaze, center of gravity, and facial expression. For example, psychological safety, approval behavior, etc. may be determined from the audio. Furthermore, the evaluation unit 13 may obtain the content of the utterance and perform natural language processing on the data obtained to provide AI suggestions.

[0096] Furthermore, in the above embodiment, the output unit 14 may output, in addition to the status of the participants and the relationships between the participants, the questionnaire results or AI analysis data obtained by performing a composite analysis on the questionnaire results, etc. Furthermore, as one example, the output unit 14 may output AI analysis data on the evaluated stress, motivation, psychological safety, and sense of growth together with the questionnaire.

[0097] Furthermore, in the third embodiment, the occurrence of harassment behavior is detected from audio information, but this is not limiting. The evaluation unit 13 may evaluate the behavior of the participant included in the target information. The detection unit 17 may detect behavior set as inappropriate from the detected behaviors. The output unit 14 may output an evaluation result based on the detected behavior. This makes it possible to detect behaviors such as yelling loudly or banging on a desk as harassment behavior. Therefore, it is possible to detect a wider range of situations that require more immediate action.

[0098] In the above embodiment, the output unit 14 outputs the evaluation results by superimposing them on the organizational chart, but the present invention is not limited to this. The output unit 14 may output the related evaluation results by placing a cursor on a predetermined position in the network data. The output unit 14 may also output the evaluation results as a video. The output unit 14 may also display the evaluation results on two screens or on separate screens.

[0099] In the above embodiment, the evaluation unit 13 may perform evaluation using only image information. For example, even if the audio information or the microphone is broken, the evaluation unit 13 may perform AI analysis using only image information such as facial expressions and behavior. Furthermore, the evaluation unit 13 may perform AI analysis using only image information without using audio information, for example, when a participant with a disability communicates using this service.

[0100] In the above embodiment, the evaluation unit 13 may analyze the target information in chronological order and output it in chronological order (images, documents, videos, etc.). The output unit 14 may output externally input questionnaire results or AI analysis data obtained by performing a combined analysis of the questionnaire results, etc., along with the evaluation results. In the above embodiment, the evaluation unit 13 may define and model the condition, organizational culture, and characteristics for each participant (company or group). In this case, the evaluation unit 13 may use the settings made by the participant, predetermined settings, or settings determined by AI from the target information accumulated up to that point as correct answer data. In addition, the external data is not limited to questionnaires, but may also be attributes of participants (for example, personality, year of joining the company, new graduate / mid-career, job title, educational background, department of experience, survey data, data obtained during recruitment activities (including videos), stress checks / personnel evaluation information (including 360-degree evaluations, etc.) / performance information (sales results, etc.) / frequency and quality of communication with other members, connection trends seen from the organizational chart (many or few members involved in work, etc.), output results of healthcare products, work history, etc.), or data associated with participants (past actions (history of leave of absence, behavior in past videos, etc.), current state (healthcare product indicators, etc.), or thoughts, etc. (surveys, etc.)).

[0101] Furthermore, the evaluation unit 13 may not only categorize the participants but also reflect the results in personnel-related policies for the organizations to which the participants belong and for the participants themselves. For example, the evaluation unit 13 may reflect the results in terms of internal communication by clarifying the culture through categorization, or as a reference when considering personnel transfers (including assignment of new employees), personnel evaluations, and other personnel policies. Furthermore, the evaluation unit 13 may not only categorize the participants but also refine the evaluation criteria of the AI. For example, the evaluation unit 13 may be used to change the criteria for AI judgment regarding the target information of the department to which the participants belong.

[0102] In the above embodiment, the evaluation device 1 can also be used as one of the employment interview systems. For example, the evaluation device 1 can link an AI for employment interviews with an AI for workplace evaluation. The evaluation device 1 can, for example, improve a series of situations from the employment interview to the workplace environment based on the judgment results of both AIs. For example, the evaluation device 1 can improve the workplace based on data at the time of recruitment. For example, the evaluation device 1 can improve the recruitment process based on workplace data. For example, the evaluation device 1 can comprehensively evaluate a series of processes from recruitment to the workplace environment, determine whether they are efficient or optimal, and provide advice. The evaluation device 1 can also optimize assignments taking into account factors such as assignment (department, team, project, etc.), personnel transfers, job content, and compatibility with employees who will be co-performing the job.

[0103] For example, when enhancing the workplace based on hiring data, AI analysis of job applicants' selection videos, interview videos, and other ancillary information (e.g., educational background, work history, aptitude tests, academic achievement tests, age, etc.) could be used to determine initial assignment after joining (e.g., which department or team is best, which job is appropriate, which supervisors or colleagues they get along with, etc.), subsequent personnel transfers, performance evaluations, and other personnel purposes. Furthermore, when enhancing the hiring process based on workplace data, information such as whether people assigned to a certain department or team are likely to thrive, grow, quit, take leave, or are a poor cultural fit could be used as reference information for hiring decisions (including other information determined at the time of hiring, such as initial assignment and career development). Furthermore, the content of the AI ​​analysis could be modified based on the job applicant's status (e.g., whether they are assigned to a specific department, attributes, and other information) and all ancillary data associated with the job applicant. Furthermore, for example, when comprehensively evaluating a series of processes from recruitment to the work environment, determining whether they are efficient and optimal, and providing advice, the above two pieces of data may be taken into consideration comprehensively to improve the accuracy of AI analysis and recommendations.Furthermore, for example, the entire flow of people from recruitment to the work environment may be evaluated as an integrated human resources system, and recommendations may be made to human resources, etc.

[0104] Furthermore, in the "comprehensive evaluation," the evaluation device 1 may treat recruitment and the work environment as the same personnel system rather than treating them as separate personnel systems, and may evaluate them while taking into consideration overall optimization. This is effective in cases where, for example, the achievement of KPIs (Key Performance Indicators) by the personnel department's recruitment team is hindering the achievement of KPIs by the personnel department's workplace team (or vice versa). For example, this is effective in cases where the recruitment team wants to increase the average score of accepted applicants in "assertiveness" in aptitude tests or the score (evaluation value) of "passion" by the evaluation unit 13, but pursuing this too much could significantly decrease scores (evaluation values) such as "psychological safety," which the workplace team places importance on. In this case, the evaluation department 13 can give the hiring team advice that is optimal for the whole team, including other departments and groups such as the workplace team (for example, other groups including the human resources department), such as, "If you only pursue the assertiveness / enthusiasm score, there is a concern that the psychological safety score of the workplace team will drop sharply. Therefore, by taking an approach such as 'keep the assertiveness / enthusiasm score at this number' or 'only accept job applicants who have a high assertiveness / enthusiasm score and also a high other score (e.g., cooperativeness)', the psychological safety score pursued by the workplace team will not drop too much, and you will be able to pursue an improvement in the assertiveness / enthusiasm score, which is the KPI of the hiring team, in a balanced manner."

[0105] Furthermore, for example, when a series of processes from recruitment to the workplace environment are comprehensively evaluated, and whether they are efficient or optimal is determined and advice is provided, the evaluation may be performed not only on the human resources department but also on the job seeker. For example, the evaluation results may be output to a job seeker who provides audio or image information including a self-introduction, without being limited to target information. In this case, the content of advice for the job seeker may be enhanced based on the evaluation results of the workplace environment and the AI ​​analysis results of the job seeker. This is particularly effective for job seekers whose assignment has been decided in advance. Furthermore, the evaluation device 1 may output an external signal if it detects a situation corresponding to a warning such as those described in the second and third embodiments during a conversation during recruitment, such as a job interview. Furthermore, inappropriate remarks may be detected in communication other than recruitment, such as in sales. For example, while conventional recruitment interviews have included stressful interviews, such stressful interviews may potentially damage the candidate experience (i.e., the job seeker's feelings of "I want to work for this company," "This company is attractive," and "It was worth the interview (I learned something, etc.)"). In traditional analog interviews, the interview process was a black box, and there was no way for a third party such as the human resources department to understand the nature of the communication from the outside. Therefore, it would be ideal if the candidate experience could be improved by detecting and alerting such inappropriate behavior.

[0106] In this case, the performance of the alert function may be tuned based on information about the job seeker held by the company (for new graduate recruitment, educational background, age, gender, aptitude test results, application form contents, and other information sent by the job seeker to the company regarding the recruitment process).In addition, the detection ability of the alert function may be improved or the performance of the alert function may be tuned based on the results of a questionnaire obtained from participants (including not only job seekers but also interviewers) after the communication or the values ​​of the healthcare product. In addition, the performance of the alert function can be tuned based on information related to the interviewer (history and trends of past communication successes and failures, participant attributes (e.g., personality, year of joining, new graduate / mid-career, position, educational background, department of experience), survey data, data obtained during recruitment activities (including videos), stress checks, personnel evaluation information (including 360-degree evaluations), performance information (sales results, etc.), frequency and quality of communication with other members, connection trends seen from the organizational chart (e.g., many or few members involved in work), healthcare product output results, work history, etc.), or data related to the participant (past actions (history of leave of absence, behavior in past videos, etc.), current state (healthcare product indicators, etc.), or thoughts, etc. (surveys, etc.)). Furthermore, the types of behavior considered inappropriate may vary depending on the organization. Therefore, the performance of the alert function may be tuned to suit the characteristics of the organization. In this case, tuning may be performed by AI, or the organization or service vendor may manually perform tuning.

[0107] In the above embodiment, the evaluation device 1 may compare data obtained at the time of hiring with data obtained at the time of employment, analyze changes in the communication style of the target job seeker (new employee), and profile the job seeker (new employee). This may, for example, determine the progress of the job seeker's (new employee's) competency information (e.g., environmental adaptability, growth rate, stress tolerance, and changes in motivation) from the time of hiring to the time of the latest AI analysis, and use this information for various personnel policies (e.g., performance evaluations, personnel transfers, identifying compatible and incompatible employees, etc.), thereby enhancing recruitment and environmental activities. In terms of UI and UX, if there are two types of AI, a hiring AI and a workplace AI, and each has its own dashboard, the analysis results of one of the two may be displayed on the other's dashboard (including reports and emails) (or, if not displayed directly, displayed in a different form).

[0108] Furthermore, for example, the evaluation device 1 can be employed as part of a system for selecting personnel who are suitable for an organization in a recruitment interview, using the relationships in workplace data obtained from the evaluation device 1, the personalities and conditions of participants, etc. For example, the evaluation device 1 can be included as part of a system that outputs advice such as, "This workplace has a tendency for job seekers to quit soon after being hired at this workplace," or, based on data at the time of hiring, "This workplace has a lot of aggressive types, so you should not hire weak people," or "This is a flexible organization, so hiring people who speak directly or who use harsh language will destroy the culture."

[0109] Furthermore, in the above embodiment, the participant state assessment unit 131 uses the frequency of smiles to determine stress, but this is not limiting. The participant state assessment unit 131 does not have to use the frequency of smiles to determine stress. [Explanation of symbols]

[0110] 1 Evaluation device 11 Target information acquisition unit 12 Relevance Acquisition Unit 13 Evaluation Section 14 Output section 15 Output content switching section 17 Detector

Claims

1. An evaluation device for evaluating the status of participants in a community, comprising: a target information acquisition unit that acquires, as target information, identification information that identifies each of the participants and audio information and image information related to the conversation of the participants that are linked to the identification information; a participant state evaluation unit that calculates a stress level based on the frequency of smiles from the facial expressions of the participant based on the acquired target information, and calculates a motivation level based on the concentration level from the line of sight and center of gravity of the participant, compares the calculated patterns of the stress level and motivation level with predetermined patterns, and evaluates the state of the participant based on the comparison result; a person-related evaluation unit that calculates a degree of empathy from the facial expressions and actions of the participants based on the target information and calculates a dialogue ratio from the amount of speech of the participants; a work-related evaluation unit that calculates a concentration level from the eye lines and center of gravity positions of the participants based on the target information, and calculates information divergence by quantifying the degree of relevance of the speech content of the participants; a relevance evaluation unit that compares the empathy level and the dialogue ratio calculated by the person-related evaluation unit, and the concentration level and the information divergence balance state calculated by the work-related evaluation unit with a predetermined balance state, and evaluates the relationship between the participants based on the comparison result; a detection unit that detects evaluation results that deviate from a predetermined threshold value among the evaluation results of the participant state evaluation unit and the relevance evaluation unit; an output unit that outputs the detected evaluation result; An evaluation device comprising:

2. The target information acquisition unit further acquires, as the target information, a date and time when the conversation between the participants was held, A conversation interval evaluation unit is further provided to evaluate the interval at which the conversation between the participants is conducted, The evaluation device according to claim 1 , wherein the detection unit detects an interval that is equal to or greater than a predetermined value or that is less than a predetermined value.

3. the person-related evaluation unit evaluates a word included in the target information, the detection unit detects terms set as inappropriate from the detected text; The evaluation device according to claim 1 , wherein the output unit outputs an evaluation result based on the detected term.

4. Any of the participant status evaluation unit, the person-related evaluation unit, the work-related evaluation unit, and the relevance evaluation unit evaluates the actions of the participant included in the target information, the detection unit detects an act set as inappropriate from the detected acts; The evaluation device according to claim 1 , wherein the output unit outputs an evaluation result based on the detected behavior.

5. further comprising an output method setting unit that sets an output method and timing of the evaluation result, 5. The evaluation device according to claim 1, wherein the output unit outputs the evaluation result in the output method and at the timing set by the output method setting unit.

6. An evaluation device described in any of claims 1 to 5, wherein any one of the participant status evaluation unit, the person-related evaluation unit, the work-related evaluation unit, and the relevance evaluation unit evaluates the changes in the status of the participant at predetermined intervals.

7. The evaluation device according to claim 1 , wherein the output unit outputs the evaluation results by superimposing them on an organizational chart.

8. A program that causes a computer to operate as an evaluation device that evaluates the status of participants in a community, The computer a target information acquisition unit that acquires, as target information, identification information that identifies each of the participants and audio information and image information related to the conversation of the participants that are linked to the identification information; a participant state evaluation unit that calculates a stress level based on the frequency of smiling from the facial expressions of the participant based on the acquired target information, and calculates a motivation level based on the concentration level from the line of sight and center of gravity of the participant, compares the calculated patterns of the stress level and motivation level with predetermined patterns, and evaluates the state of the participant based on the comparison result; a person-related evaluation unit that calculates a degree of empathy from the facial expressions and actions of the participants based on the target information and calculates a dialogue ratio from the amount of speech of the participants; a work-related evaluation unit that calculates a concentration level from the eye lines and center of gravity positions of the participants based on the target information, and calculates information divergence by quantifying the degree of relevance of the speech content of the participants; a relevance evaluation unit that compares the empathy level and the dialogue ratio calculated by the person-related evaluation unit, and the concentration level and the information divergence balance state calculated by the work-related evaluation unit with a predetermined balance state, and evaluates the relationship between the participants based on the comparison result; a detection unit that detects evaluation results that deviate from a predetermined threshold value among the evaluation results of the participant state evaluation unit and the relevance evaluation unit; an output unit that outputs the detected evaluation result; A program that functions as a

Images