Communication evaluation system, communication evaluation method, and program

Abstract

A communication evaluation system according to the present disclosure comprises: an acquisition unit that, in bidirectional communication performed between a plurality of users belonging to a project team, acquires reaction data, which is data including content of a reaction performed by another user with respect to an action of a user; an evaluation unit that evaluates a degree of fulfillment of the communication between the plurality of users on the basis of at least the reaction data; and an output unit that outputs the evaluation of the degree of fulfillment by the evaluation unit.

Description

Communication Evaluation System, Communication Evaluation Method, and Program

[0001] The present disclosure relates to a communication evaluation system, a communication evaluation method, and a program.

[0002] Various techniques for evaluating outputs by users have been proposed. For example, Patent Document 1 discloses a technique for evaluating a person's speech and presenting the evaluation result. The information processing apparatus according to Patent Document 1 evaluates each person based on the speech uttered by each person in a meeting consisting of multiple people. Further, the information processing apparatus displays the result of the evaluation in a graph divided by each person.

[0003] Japanese Unexamined Patent Application Publication No. 2019-174768

[0004] The information processing apparatus according to Patent Document 1 evaluates each user participating in the meeting individually. Specifically, in Patent Document 1, each user is evaluated individually from the user's speaking time and the keywords uttered by the user. Therefore, with the technique according to Patent Document 1, it is not possible to evaluate the communication situation among users.

[0005] The present disclosure has been made to solve such problems, and an object thereof is to provide a communication evaluation system, a communication evaluation method, and a program capable of grasping the communication situation among a plurality of users.

[0006] The communication evaluation system according to the present disclosure includes an acquisition unit that acquires reaction data, which is data including the content of reactions made by other users to the actions of a user, in a two-way communication performed among a plurality of users belonging to a project team; an evaluation unit that evaluates the degree of enrichment of the communication among the plurality of users based at least on the reaction data; and an output unit that outputs the evaluation of the degree of enrichment by the evaluation unit.

[0007] The communication evaluation method relating to this disclosure involves a computer acquiring reaction data, which is data containing the content of reactions made by other users to user actions in two-way communication between multiple users belonging to a project team, evaluating the richness of the communication between the multiple users based on at least the reaction data, and outputting the evaluation of the richness.

[0008] The program relating to this disclosure causes a computer to perform the following steps in two-way communication between multiple users belonging to a project team: acquiring reaction data, which is data containing the content of reactions that other users make to an action taken by a user; evaluating the quality of the communication between the multiple users based on at least the reaction data; and outputting the evaluation of the quality.

[0009] This disclosure provides a communication evaluation system, a communication evaluation method, and a program that can grasp the communication status of multiple users.

[0010] Figure 1 is a block diagram showing the configuration of the communication evaluation system related to this disclosure. Figure 2 is a flowchart showing an example of the flow of the evaluation method by the communication evaluation system. Figure 3 is a block diagram showing the configuration of the communication evaluation system related to this disclosure. Figure 4 is a block diagram showing the configuration of the classification unit. Figure 5 is a block diagram showing the configuration of the evaluation unit. Figure 6 is a block diagram showing the configuration of the quantitative analysis unit. Figure 7 is a table showing an example of the analysis results output by the output unit. Figure 8 is a block diagram showing the configuration of the quality analysis unit. Figure 9 is a table showing an example of the integrated information generated by the integration unit. Figure 10 is a table showing an example of the evaluation results generated by the evaluation result generation unit. Figure 11 is a flowchart showing an example of the flow of acquiring various data by the acquisition unit. Figure 12 is a flowchart showing an example of the flow of category classification by the classification unit. Figure 13 is a flowchart showing an example of the flow of evaluation of the richness of communication by the evaluation unit. Figure 14 is a diagram showing an example of the hardware configuration of the communication evaluation system.

[0011] (Embodiment 1) Hereinafter, Embodiment 1 of the present disclosure will be described with reference to the drawings. Figure 1 is a block diagram showing the configuration of the communication evaluation system 1 according to the present disclosure. The communication evaluation system 1 is typically one or more computer devices that operate by a processor executing a program stored in memory. In other words, the communication evaluation system 1 may be composed of one computer device or two or more computer devices. For example, the communication evaluation system 1 is one or more server devices.

[0012] Communication Evaluation System 1 is a system for evaluating two-way communication between multiple users belonging to a project team. A "project" is an undertaking or operation undertaken to achieve a predetermined objective. That is, a "project team" is a group of multiple users formed to achieve a common objective. A project team is typically composed of multiple users belonging to the same company or organization, but it may also be composed of users belonging to different companies or organizations. Furthermore, a project team may be composed of multiple users belonging to the same department within a company or organization. In other words, a project team may be any single department, section, or unit within a company or organization.

[0013] Furthermore, "two-way communication" refers to communication in which one user takes a communication action from another user, and the other user reacts to that action. In other words, communication includes the words that a user uses to communicate with another user and the reactions of the other user to those words. While communication is typically verbal, it does not have to be verbal if it is possible to convey some opinion or feeling to another user. For example, in a system that allows communication via chat, reactions using reaction functions such as "like" are included in communication.

[0014] The communication evaluation system 1 comprises an acquisition unit 11, an evaluation unit 12, and an output unit 13. The acquisition unit 11, evaluation unit 12, and output unit 13 are typically software or modules whose processing is performed by a processor executing a program stored in memory. The acquisition unit 11, evaluation unit 12, and output unit 13 may also be hardware such as circuits or chips. That is, the acquisition unit 11, evaluation unit 12, and output unit 13 may each be composed of different computer devices.

[0015] The acquisition unit 11 and evaluation unit 12, and the evaluation unit 12 and output unit 13 are connected in a data communication manner. This data communication may be performed via a wired connection or a wireless connection. The data communication may be performed within the same computer device, via an internet connection, or using short-range wireless communication technology. The type of communication protocol used for data communication is not limited.

[0016] The acquisition unit 11 acquires reaction data, which is data containing the content of reactions made by other users to user actions in two-way communication between multiple users belonging to a project team. Here, "reaction data" is data that associates a reaction with the user who made that reaction. Reaction data is typically text data expressed in a tabular format. Reaction data may include reactions made using a reaction function implemented in a system with a chat function, for example. Reaction data may also include emojis. Reactions and emojis made using the reaction function may be converted into predetermined text data.

[0017] The evaluation unit 12 evaluates the level of communication among multiple users in the project team based on reaction data acquired by at least the acquisition unit 11. The level of communication may be the sum of predetermined items calculated based on the reaction data, or it may be a score when evaluating the level of communication in stages. In other words, the level of communication is typically expressed numerically.

[0018] The output unit 13 outputs the evaluation of completeness performed by the evaluation unit 12. The output unit 13 may output to a predetermined user or to another computer device. When outputting to a user, the output unit 13 may output the evaluation of completeness to the user who made the reaction related to the reaction data, or to a different user. For example, the output unit 13 can output the evaluation to a person who manages multiple users who have communicated. The output unit 13 may also output the evaluation of completeness after processing it in a predetermined way.

[0019] Next, the flow of the communication evaluation method by the communication evaluation system 1 will be explained. Figure 2 is a flowchart showing an example of the flow of the evaluation method by the communication evaluation system 1. First, the acquisition unit 11 acquires reaction data in two-way communication between multiple users belonging to the project team (S101). Next, the evaluation unit 12 evaluates the quality of communication between the multiple users based on at least the reaction data (S102). After that, the output unit 13 outputs the quality evaluation by the evaluation unit 12 (S103).

[0020] Thus, in this embodiment, the communication evaluation system 1 can evaluate bidirectional communication between multiple users belonging to a project team. Specifically, the communication evaluation system 1 can evaluate the quality of communication based at least on reaction data. The communication evaluation system 1 can then output an evaluation of the quality of communication. By considering what kind of reactions the users made, the communication evaluation system 1 can grasp the communication situation among multiple users.

[0021] (Embodiment 2) This second embodiment is a specific example of the first embodiment described above. Figure 3 is a block diagram showing the configuration of the communication evaluation system 2 according to this disclosure. The communication evaluation system 2 includes an acquisition unit 21, a classification unit 22, an evaluation unit 23, and an evaluation result output unit 24. The components of the communication evaluation system 2 are typically software or modules whose processing is performed by a processor executing a program stored in memory. The communication evaluation system 2 may also be hardware such as a circuit or a chip. That is, the components of the communication evaluation system 2 may each be different computer devices. For example, the communication evaluation system 2 is one or more server devices. Hereafter, for the sake of clarity, explanations that overlap with the first embodiment will be omitted as appropriate.

[0022] In the communication evaluation system 2, predetermined components are connected to each other in a manner that allows for data communication as needed. Data communication between components may be performed via wired or wireless connections. Data communication may be performed within the same computer device, via an internet connection, or using short-range wireless communication technology. The type of communication protocol used for data communication is not limited.

[0023] In the communication evaluation system 2, the project team is divided into, for example, a management team that manages the project and sub-teams that carry out the practical work of the project. In other words, the project team may be divided into multiple small teams. Here, the management team is typically one, and the sub-team is typically composed of multiple people. In this case, the communication evaluation system 2 can evaluate communication between users within the small teams of the project team and communication between the small teams. Specifically, the communication evaluation system 2 can evaluate communication within the management team or within the sub-teams, communication between the management team and each sub-team, and communication between the sub-teams.

[0024] Furthermore, if a project team refers to a group composed of users belonging to the same department, the communication evaluation system 2 may evaluate communication within that team. For example, if a team from the same department consists of a group leader and multiple sections, the communication evaluation system 2 may evaluate communication between the group leader and each section, as well as communication between each section.

[0025] The communication evaluation system 2 is not limited to evaluating the quality of communication for only one project team. In other words, the communication evaluation system 2 may evaluate the quality of communication for each of multiple project teams.

[0026] In this second embodiment, the communication targeted by the communication evaluation system 2 is, for example, a chat between multiple users who make up a project team. Furthermore, the communication targeted by the communication evaluation system 2 may also be email communication between users, or communication through meetings and discussions related to the project.

[0027] The acquisition unit 21 acquires reaction data in bidirectional communication between multiple users belonging to the project team. In other words, the acquisition unit 21 corresponds to the acquisition unit 11 in the communication evaluation system 1. Furthermore, the acquisition unit 21 can acquire action data, which includes the content of actions in communication. In other words, the acquisition unit 21 can acquire both action data and reaction data. Action data and reaction data together are referred to as communication data.

[0028] Communication data includes chat data, which is, for example, log data of chats between users. Here, chat data is data that associates a chat with the user who conducted that chat. Chat data is typically in tabular format. Chat data also includes the content of reactions from the reaction function implemented in the system that conducts the chat. Reactions from the reaction function include, for example, reactions such as "like". Furthermore, chat data may also include emojis entered by the user. Reactions from the reaction function and emojis may be converted into predetermined text data.

[0029] Furthermore, communication data may include email data from emails sent and received between users, as well as meeting data from conferences and discussions. Here, meeting data is typically data expressed in text. The text data of meeting data may be, for example, meeting minutes or meeting notes created based on a meeting or discussion, or it may be audio data of a recording of the meeting or discussion converted into text data. Email data and meeting data are typically tabular data that associates communication with the users who conducted the communication.

[0030] Communication data is further associated with the communication itself and the timing information of the communication. Here, timing information refers to, for example, information about the time the communication took place. In this case, the communication data may be data arranged in chronological order. In particular, if the communication data includes chat data, email data, and meeting data, the communication data may be tabular data in which this data is arranged in chronological order. Specifically, the communication data may be tabular data in which the content of the communication, the time the communication took place, the user who made the communication, and the recipient of the communication are related to each other and arranged in chronological order. The acquisition unit 21 may organize the chat data, email data, and meeting data into such a tabular format.

[0031] The acquisition unit 21 can further acquire user data that shows the relationships between multiple users belonging to the project team. User data includes, for example, identification information and attribute information of users who make up the project team, as well as data on the structure of the project team and the role of each user in the team. Here, if the project team is divided into multiple sub-teams, the user attribute information includes identification information of the sub-team to which the user belongs. User data may also include an overview of the project. User data is typically in tabular format.

[0032] When the communication evaluation system 2 evaluates the quality of communication for multiple project teams, the acquisition unit 21 can acquire reaction data for the multiple project teams. Similarly, the acquisition unit 21 can acquire action data and multiple user data for the same multiple project teams.

[0033] The classification unit 22 classifies the action data and the corresponding reaction data into multiple categories based on the action data acquired by the acquisition unit 21. In other words, the classification unit 22 classifies the data related to each action and its corresponding reaction into multiple categories based on the content of each action in the action data.

[0034] Specifically, the classification unit 22 can classify user actions into categories by analyzing the content of those actions. Categories include, for example, "casual conversation," "reporting," "questioning," and "discussion." The casual conversation category is where user actions that are casual conversations are classified. Similarly, the reporting category, questioning category, and discussion category are where user actions that are reports, questions, and discussions are classified, respectively. Here, the classification unit 22 may classify the communication data into categories not only based on the action data but also on the reaction data to that action. Hereafter, the explanation will continue assuming that the classification unit 22 classifies the action data into the four categories described above. Note that there may be categories other than these four. Also, the classification unit 22 typically classifies one action into one category, but it may classify one action into multiple categories.

[0035] The configuration of the classification unit 22 will now be explained. Figure 4 is a block diagram showing the configuration of the classification unit 22. The classification unit 22 comprises a receiving unit 221, an input unit 222, and an output unit 223. The receiving unit 221 receives communication data acquired by the acquisition unit 21. The receiving unit 221 can also receive user data acquired by the acquisition unit 21.

[0036] The input unit 222 inputs the communication data received by the reception unit 221 into a predetermined program. Here, the input unit 222 may input only action data to the predetermined program, or it may input both action data and reaction data. The predetermined program is a program that, by inputting the communication data received by the reception unit 221, can classify the communication data into one of the predetermined categories. That is, the program into which the input unit 222 inputs the communication data can output category information corresponding to the communication data. Here, the category information is, for example, data indicating the category for the action data. Specifically, the category information may be tabular data in which action data and categories are associated. Furthermore, the category information may be data to which reaction data corresponding to action data is further associated.

[0037] The predetermined program input to the input unit 222 is, for example, a trained model that outputs category information when communication data is input. For example, if the reception unit 221 receives only action data, the program is a trained model that outputs category information to which the action belongs when action data is input. The trained model is a model generated by a neural network (NN) based machine learning algorithm such as a CNN (Convolutional Neural Network). In this case, the trained model comprises an input layer, an output layer, and a hidden layer. The input layer receives communication data including action data. The output layer outputs category information for the input communication data. The hidden layer's parameters are learned using training data that takes communication data including action data as input and category information as output. This trained model takes communication data as input to the input layer, performs calculations in the hidden layer, and outputs category information from the output layer.

[0038] The input unit 222 may input only action data to the trained model, or it may input both action data and reaction data. If only action data is input, the trained model will generate the action data as training data. If both action data and reaction data are input, the trained model will generate the action data and reaction data as training data. The input unit 222 may also input user data to the trained model. In this case, the trained model will include the user data as training data. Inputting user data can improve the accuracy of classification into categories.

[0039] The output unit 223 outputs category information from a predetermined program input by the input unit 222. When the input unit 222 inputs communication data to the aforementioned trained model, the output unit 223 outputs category information from the trained model. That is, if the action is small talk, the output unit 223 outputs category information indicating that the action data belongs to the small talk category. Similarly, if the action is a report, a question, or a discussion, the output unit 223 outputs category information indicating that the action data belongs to one of the report, question, or discussion categories. As a result, the classification unit 22 can classify the action data into one of the following categories: "small talk," "report," "question," or "discussion."

[0040] Furthermore, when the communication evaluation system 2 evaluates the level of communication for multiple project teams, the classification unit 22 can classify each of the action data for the multiple project teams acquired from the acquisition unit 21 into categories.

[0041] Returning to Figure 3, let's continue the explanation of the configuration of the communication evaluation system 2. The evaluation unit 23 evaluates the quality of communication between multiple users, at least based on reaction data. In other words, the evaluation unit 23 corresponds to the evaluation unit 12 in the communication evaluation system 1. In the communication evaluation system 2, the evaluation unit 23 evaluates the quality of communication based on reaction data corresponding to action data, according to the category of action data classified by the classification unit 22. Specifically, the evaluation unit 23 evaluates the quality by analyzing the reaction data into the category to which the reaction data belongs.

[0042] Here, the configuration of the evaluation unit 23 will be explained. Figure 5 is a block diagram showing the configuration of the evaluation unit 23. The evaluation unit 23 comprises a quantitative analysis unit 231, a qualitative analysis unit 232, and an evaluation result generation unit 233. The quantitative analysis unit 231 analyzes reaction data by calculating the number of reactions belonging to each category. That is, the quantitative analysis unit 231 analyzes the quantity of reactions. Here, the quantity of reactions is one of the levels of richness of communication. That is, the number of reactions is one of the levels of richness. The number of reactions includes reactions from the reaction function. For example, the quantitative analysis unit 231 calculates the number of reactions related to reaction data classified into the casual conversation category. Similarly, the quantitative analysis unit 231 calculates the number of reactions related to reaction data classified into the reporting category, the question category, and the discussion category.

[0043] The quantitative analysis unit 231 may calculate the number of reactions related to reaction data within a small team by acquiring user data from the acquisition unit 21. The quantitative analysis unit 231 may also calculate the number of reactions between small teams. Furthermore, the quantitative analysis unit 231 may calculate the number of actions corresponding to reactions related to reaction data. In other words, the quantitative analysis unit 231 may calculate the number of both actions and reactions.

[0044] Here, the configuration of the quantity analysis unit 231 will be described. FIG. 6 is a block diagram showing the configuration of the quantity analysis unit 231. The quantity analysis unit 231 includes a reception unit 231a, an input unit 231b, and an output unit 231c. The reception unit 231a receives communication data and category information from the classification unit 22. Further, the reception unit 231a may receive user data from the acquisition unit 21 in addition to the communication data.

[0045] The input unit 231b inputs the communication data and user data received by the reception unit 231a into a predetermined program. The predetermined program is typically a rule-based program capable of calculating the number of actions and reactions. Incidentally, when the communication data and user data are input, the input unit 231b may input the communication data and user data into a learned model capable of outputting the number of actions and reactions.

[0046] The output unit 231c outputs an analysis result regarding the quantity from the predetermined program into which the input unit 231b inputs the communication data and user data. Specifically, the output unit 231c outputs the number of reactions for each category. The analysis result output by the output unit 231c is one of the evaluation results regarding the richness of communication. Here, the output unit 231c may output the number of actions in addition to the number of reactions. Further, the output unit 231c may output the number of actions and reactions for each small team within the project team. Furthermore, the output unit 231c may output the number of actions and reactions between small teams.

[0047] The analysis results output by the output unit 231c are tabular data in which categories are associated with the number of reactions in each category. When calculating the number of times for each small team and the number of times between small teams, the output unit 231c may further associate this information. Here, an example of the analysis results output by the output unit 231c will be described. FIG. 7 is a table showing an example of the analysis results output by the output unit 231c. In FIG. 7, the quantitative analysis unit 231 calculates the number of times for both actions and reactions. In FIG. 7, the project team is divided into a management group and three sub-teams. Each sub-team is referred to as sub-team A, sub-team B, and sub-team C. The output unit 231c outputs the calculated number of communication times within the management group, within the sub-team, and between the management group and the sub-team for each of the categories of chatting, reporting, questioning, and discussion. For example, in FIG. 7, the number of times of chatting, reporting, questioning, and discussion within the management group (Management G) are 50 times, 50 times, 100 times, and 10 times, respectively. Similarly, the number of times within sub-team A (Sub A) are 20 times, 70 times, 100 times, and 20 times. Also, the number of times between the management group and sub-team A (Management - A) are 70 times, 5 times, 125 times, and 5 times.

[0048] Returning to FIG. 5, the description of the configuration of the evaluation unit 23 will be continued. The qualitative analysis unit 232 analyzes the content of actions and reactions related to action data and reaction data belonging to a category. Based on this, the qualitative analysis unit 232 analyzes the quality of the reaction to an action. The quality of the reaction is one aspect of the richness of communication. The analysis of the content of the reaction by the qualitative analysis unit 232 differs for each category. In other words, the qualitative analysis unit 232 can change the method of analyzing the content of the reaction for each category. Incidentally, the qualitative analysis unit 232 may analyze the content of the reaction by the same analysis method across all categories.

[0049] An example of analysis methods for each category is described below. When reaction data is classified into the casual conversation category, the quality analysis unit 232 analyzes the degree of familiarity of the reactions to the casual conversation actions. The quality analysis unit 232 can quantify the degree of familiarity using a score calculated based on a predetermined index, for example. The degree of familiarity of a reaction is one aspect of the richness of communication. The quality analysis unit 232 can analyze this degree of familiarity using a pre-trained model that outputs the degree of familiarity of a reaction to reaction data when communication data is input. The pre-trained model can be generated using, for example, an NN-based machine learning algorithm. In this case, the pre-trained model comprises an input layer, an output layer, and a hidden layer. The input layer receives communication data as input. The output layer outputs the degree of familiarity of the reaction for the input communication data. The parameters of the hidden layer are learned using training data that takes communication data as input and the degree of familiarity of the reaction as output. This pre-trained model takes communication data as input to the input layer, performs calculations in the hidden layer, and outputs the degree of familiarity of the reaction from the output layer.

[0050] If the reaction data is classified into a reporting category, the quality analysis unit 232 analyzes the responsiveness of the reactions. Responsiveness is the degree to which other users have responded to a user's reporting action. Responsiveness is one aspect of the richness of communication. Responsiveness may include messages such as "understood" or reactions using reaction functions such as "like". For each reporting category, the quality analysis unit 232 analyzes the responsiveness by, for example, analyzing the number of reactions. The quality analysis unit 232 can quantify the responsiveness using a score calculated based on a predetermined standard for the number of reactions to a single reporting action. For example, the quality analysis unit 232 can analyze the responsiveness by using a rule-based program that can calculate the number of reactions to an action.

[0051] If the reaction data is classified into question categories, the quality analysis unit 232 analyzes the response rate of the reaction. Response rate refers to the degree to which the response is appropriate to the content of the question and the speed of the response. Response rate is one aspect of the quality of communication. Regarding the appropriateness of the response, the quality analysis unit 232 can analyze this appropriateness using a trained model that, for example, outputs the appropriateness of the response related to the reaction data when communication data is input. When using an NN-based machine learning algorithm, the input layer receives communication data as input. The output layer outputs the appropriateness of the response to the input communication data. The parameters of the hidden layer are learned using training data that takes communication data as input and outputs the appropriateness of the response.

[0052] Regarding the speed of responses, the quality analysis unit 232 can analyze the speed of responses using, for example, timing information of the action performed in the question and timing information of the response to that action. In this case, the quality analysis unit 232 can analyze the speed of responses by, for example, using a rule-based program that calculates the time from the timing information related to the action and the response to the response. For each question category, the quality analysis unit 232 can quantify the degree of response by assigning a score based on predetermined indicators for the appropriateness of the response and the speed of the response.

[0053] If the reaction data is classified into a discussion category, the quality analysis unit 232 analyzes the degree of discussion coherence. Discussion coherence is the degree to which the points of discussion in the action and the reaction match. Coherence is one aspect of the quality of communication. The quality analysis unit 232 can quantify the discussion coherence using a score calculated based on a predetermined indicator, for example. Here, the quality analysis unit 232 can analyze the degree of discussion coherence using a trained model that outputs the discussion coherence when communication data is input. When using an NN-based machine learning algorithm, the input layer receives communication data as input. The output layer outputs the appropriateness of the response to the input communication data. The parameters of the hidden layer are learned using training data that takes communication data as input and outputs the appropriateness of the response.

[0054] The configuration of the quality analysis unit 232 will now be described. Figure 8 is a block diagram showing the configuration of the quality analysis unit 232. The quality analysis unit 232 comprises a reception unit 232a, a determination unit 232b, an input unit 232c, an output unit 232d, and an integration unit 232e. The input unit 232c comprises a first input unit 232c1, a second input unit 232c2, a third input unit 232c3, and a fourth input unit 232c4. The output unit 232d comprises a first output unit 232d1, a second output unit 232d2, a third output unit 232d3, and a fourth output unit 232d4.

[0055] The reception unit 232a receives communication data and category information from the classification unit 22. The reception unit 232a may also receive user data from the acquisition unit 21. The determination unit 232b determines which category the action and reaction belong to. The determination unit 232b then determines which input unit 232c to input the action data and reaction data into. Specifically, if the action belongs to the casual conversation category, the determination unit 232b inputs the data into the first input unit 232c1. Similarly, if the action belongs to the report category, question category, or discussion category, the determination unit 232b inputs the data into either the second input unit 232c2, the third input unit 232c3, or the fourth input unit 232c4.

[0056] The input unit 232c inputs the communication data and user data received by the reception unit 232a into a predetermined program. Here, the program into which the input unit 232c inputs data may differ depending on the category to which the action belongs. If the action is entered into the casual conversation category, the first input unit 232c1 inputs the communication data into the corresponding program. Similarly, if the action belongs to the reporting category, question category, or discussion category, the second input unit 232c2, the third input unit 232c3, or the fourth input unit 232c4 inputs the communication data into the corresponding program. Here, the program may be a pre-trained model as described above, or it may be a rule-based program.

[0057] The output unit 232d outputs analysis results regarding the quality of reactions from a predetermined program. Specifically, the first output unit 232d1 outputs the degree of friendliness of the reaction to the casual conversation action. Similarly, the second output unit 232d2, the third output unit 232d3, and the fourth output unit 232d4 output the degree of responsiveness of the reaction, the degree of response of the reaction, and the degree of coherence of the discussion, respectively.

[0058] The integration unit 232e integrates the analysis results output by the output unit 232d. That is, the integration unit 232e generates integrated information based on the analysis results output by the output unit 232d. The integrated information is tabular data that associates categories with evaluation values ​​for the quality of reactions in each category. Here, the evaluation value is a score calculated for the quality of each reaction based on a predetermined index. The integrated information generated by the integration unit 232e is one of the evaluation results regarding the quality of communication. When calculating evaluation values ​​for each small team and evaluation values ​​between small teams, the integration unit 232e may further associate this information.

[0059] Here, an example of integrated information generated by the integration unit 232e will be described. Figure 9 is a table showing an example of integrated information generated by the integration unit 232e. The table in Figure 9 is based on the same assumptions as in Figure 7. In Figure 9, the evaluation values ​​are divided into four levels, with higher values ​​indicating higher quality of reaction. In Figure 9, the evaluation values ​​for the quality of reactions to casual conversation, reporting, questioning, and discussion actions within the management group are 4, 4, 4, and 1, respectively. Similarly, the evaluation values ​​within sub-team A are 2, 2, 1, and 3. Furthermore, the evaluation values ​​between the management group and sub-team A are 4, 2, 3, and 2.

[0060] Furthermore, when the communication evaluation system 2 evaluates the level of completeness for multiple project teams, the quantitative analysis unit 231 and the qualitative analysis unit 232 can analyze the quantity and quality of reactions to the communication data of the multiple project teams.

[0061] Returning to Figure 5, let's continue explaining the configuration of the evaluation unit 23. The evaluation result generation unit 233 generates an evaluation result regarding the degree of communication enrichment based on the quantitative analysis results from the quantitative analysis unit 231 and the quality analysis results from the quality analysis unit 232. In other words, the evaluation result generation unit 233 generates an evaluation result based on the analysis results output by the output unit 231c and the integrated information generated by the integration unit 232e. Here, the evaluation result is also called the diagnostic result. The evaluation result generation unit 233 can generate an evaluation result by organizing the analysis results output by the output unit 231c and the integrated information generated by the integration unit 232e.

[0062] Furthermore, the evaluation result generation unit 233 can generate improvement suggestions to enhance the quality of communication based on the quality of communication. The improvement suggestions are typically in the form of text data. The evaluation result generation unit 233 may generate improvement suggestions by, for example, inputting the quality of communication data into a trained model that outputs improvement suggestions when the quality of communication data and the quality of communication are input, and then outputting improvement suggestions. Specifically, the evaluation result generation unit 233 may generate improvement suggestions by inputting the communication data acquired by the acquisition unit 21, the analysis results output by the output unit 231c, and the integrated information generated by the integration unit 232e into a trained model that outputs improvement suggestions, and then outputting improvement suggestions.

[0063] For example, when generating a pre-trained model using an NN-based machine learning algorithm, the pre-trained model comprises an input layer, an output layer, and an intermediate layer. The input layer receives communication data, analysis results from quantitative analysis, and integrated information from qualitative analysis. The output layer outputs improvement suggestions for the input communication data, analysis results, and integrated information. The intermediate layer's parameters are learned using training data that takes communication data as input and improvement suggestions as output. This pre-trained model takes communication data as input, performs calculations in the intermediate layer, and outputs improvement suggestions from the output layer.

[0064] When the communication evaluation system 2 evaluates the quality of communication for multiple project teams, the evaluation result generation unit 233 may generate the evaluation result after comparing the quality of communication in one project team with the quality of communication in multiple other project teams. In other words, the evaluation result generation unit 233 may reflect an index (reference value) set based on the quality of communication in multiple other project teams in the evaluation result. For example, regarding the quantitative analysis result, the evaluation result generation unit 233 may calculate the average number of actions and reactions for each category in the other project teams and reflect the difference between that average and the number in the target project team in the evaluation result. Here, the average in the other project teams may be the average number of actions and reactions per person over a certain period, based on the timing information of the communication. In this case, the evaluation result generation unit 233 calculates the difference between this and the number in the target project team over the same period.

[0065] If a project team is divided into smaller teams, and these smaller teams are hierarchically divided into "management groups" and "sub-teams," the average number of occurrences among the smaller teams may be a value that takes into account the number of smaller teams at the same hierarchical level. Specifically, for example, suppose a project team consists of one management group and three sub-teams. In this case, the three sub-teams are at the same hierarchical level. Therefore, the evaluation result generation unit 233 may use the average values ​​between management groups and sub-teams, and the average values ​​between sub-teams, for other project teams relative to the target project team, as the average number of occurrences per person and per sub-team over a certain period. The evaluation result generation unit 233 may then use the average values ​​between sub-teams in those other project teams to calculate the difference between those values ​​and the number of occurrences among all sub-teams in the target project team.

[0066] Furthermore, the evaluation result generation unit 233 can similarly calculate the average evaluation value for other project teams based on the quality analysis results. The evaluation result generation unit 233 can then reflect the difference between the target project team's evaluation and the average value in the evaluation results.

[0067] Here, we will explain an example of the evaluation results generated by the evaluation result generation unit 233. Figure 10 is a table showing an example of the evaluation results generated by the evaluation result generation unit 233. Figure 10 is information that organizes the analysis results from Figure 7 and the integrated information from Figure 9. Also, Figure 10 is generated for output to the user. Here, the analysis results by the quality analysis unit 232 are converted from evaluation values ​​to symbols. Specifically, in Figure 9, "4" is converted to "◎", "3" to "○", "2" to "△", and "1" to "×".

[0068] In Figure 10, the analysis results for quantity and quality that fall below the average level of completeness in other project teams are highlighted with borders. This allows users to intuitively understand the differences in completeness compared to other project teams. Here, for items that fall below the average level, the evaluation result generation unit 233 may use shading, enlarge the font size, or use a color that distinguishes the text from other items.

[0069] Returning to Figure 3, we will continue the explanation of the configuration of the communication evaluation system 2. The evaluation result output unit 24 outputs the evaluation of the quality of communication by the evaluation unit 23. In other words, the evaluation result output unit 24 corresponds to the output unit 13 in the communication evaluation system 1. The evaluation result output unit 24 is configured to output the evaluation results generated by the evaluation result generation unit 233. The evaluation result output unit 24 may output the evaluation results generated by the evaluation result generation unit 233 to a user or to another computer device. The user to whom the evaluation result generation unit 233 outputs is, for example, the project team leader or manager. In other words, the evaluation result output unit 24 can output the evaluation results to someone who is not a user who communicates within the project team. Furthermore, the evaluation result output unit 24 can output improvement suggestions generated by the evaluation unit 23.

[0070] Next, the flow of the communication evaluation method by the communication evaluation system 2 will be explained. Figure 11 is a flowchart showing an example of the flow of acquiring various data by the acquisition unit 21. First, the acquisition unit 21 acquires communication data consisting of action data and reaction data (S201). Here, the acquisition unit 21 may rearrange the data in chronological order based on the timing information of when the communication took place. Next, the acquisition unit 21 acquires user data (S202). Steps S201 and S202 may be performed simultaneously, or step S202 may be performed before S201.

[0071] Figure 12 is a flowchart illustrating an example of the category classification process performed by the classification unit 22. First, the reception unit 221 of the classification unit 22 receives communication data and user data from the acquisition unit 21 (S203). Next, the input unit 222 inputs the communication data and user data into a predetermined program (S204). Here, the input unit 222 may also input the data to a trained model that outputs category information when communication data and user data are input. Subsequently, the output unit 223 outputs category information from the program (S205). Here, if the input unit 222 inputs to the trained model, the output unit 223 outputs category information from the trained model.

[0072] Figure 13 is a flowchart illustrating an example of the flow of the evaluation of the quality of communication by the evaluation unit 23. First, the quantitative analysis unit 231 of the evaluation unit 23 analyzes the quantity of communication by calculating the number of reactions based on the communication data and user data acquired by the acquisition unit 21 (S206). Next, the quantitative analysis unit 231 outputs the analysis results for the quantity analysis (S207). Specifically, the output unit 231c of the quantitative analysis unit 231 outputs the analysis results. After that, the quality analysis unit 232 analyzes the quality of reactions based on the communication data and user data (S208). Specifically, the quality analysis unit 232 analyzes the reaction data according to the category to which the reaction data belongs. After that, the quality analysis unit 232 outputs the analysis results for the quality analysis (S209). Specifically, the integration unit 232e of the quality analysis unit 232 generates integrated information of the analysis results output from the output unit 232d. Steps S206 and S208 may be performed simultaneously. Furthermore, steps S207 and S209 may be performed simultaneously. In addition, steps S208 and S209 may be performed before S206 and S207. Subsequently, the evaluation result generation unit 233 generates evaluation results based on the quantitative and qualitative analysis results (S210). Subsequently, the evaluation result output unit 24 outputs the evaluation results generated by the evaluation result generation unit 233 (S211).

[0073] In this way, the communication evaluation system 2 can concretely implement the communication evaluation system 1. That is, the communication evaluation system 2 can evaluate the quality of communication based at least on reaction data. Furthermore, by considering not only reaction data but also action data and user data, the communication evaluation system 2 can evaluate the quality of communication in more detail.

[0074] Project team activities can be evaluated by assessing the results of those activities. However, it can be difficult to evaluate project team activities during their progress. Therefore, the inventors have found that the results of team activities such as projects are correlated with the quantity and quality of communication within the project. In other words, the inventors believe that understanding and improving the communication situation within a project team may lead to the success of the team activity. Based on this background, the Communication Evaluation System 2 attempts to understand the communication situation of multiple users by analyzing communication data within a project team, such as chat data, email data, and meeting data.

[0075] Communication evaluation system 2 can classify action data into multiple categories based on at least the action data itself, and evaluate the quality of communication according to the category. For example, communication evaluation system 2 can classify action data into categories such as "casual conversation," "reporting," "questioning," and "discussion." By evaluating the quality according to the category, communication evaluation system 2 can perform a more detailed evaluation. As a result, users can easily understand the communication situation. For example, if the action is "casual conversation," communication evaluation system 2 can evaluate the quality based on the level of friendliness of the reaction to that action. Similarly, if the action is "reporting," "questioning," or "discussion," communication evaluation system 2 can evaluate the quality based on the level of responsiveness of the reaction to the action, the level of response to the reaction, or the level of coherence of the discussion. This allows users to concretely understand the communication situation.

[0076] Furthermore, the communication evaluation system 2 can evaluate the richness of communication using a pre-trained model that outputs the richness of communication when reaction data is input. This allows the communication evaluation system 2 to evaluate the richness of communication using text data about communication between users. In addition, although such communication data between users can be enormous, the pre-trained model can appropriately evaluate the richness of communication even when the data is enormous.

[0077] Furthermore, the communication evaluation system 2 can have a trained model perform categorization of action data and reaction data. Specifically, the communication evaluation system 2 can perform categorization by using a trained model that outputs category information to which an action belongs when action data is input. As a result, even if the communication data does not have classification information indicating the category attached to it, the communication evaluation system 2 can classify the communication data into the appropriate category from the text data.

[0078] Communication evaluation system 2 can generate improvement suggestions to enhance the level of communication based on its effectiveness. This allows project team managers and supervisors, for example, to improve team activities and consider strategies to increase the results of those activities.

[0079] Communication evaluation system 2 can generate improvement suggestions using a pre-trained model. Specifically, communication evaluation system 2 can generate improvement suggestions by using a pre-trained model that outputs improvement suggestions when communication data and quality are input. As a result, communication evaluation system 2 can generate improvement suggestions based on the specific content of the communication.

[0080] Furthermore, the communication evaluation system 2 can reflect the level of communication in other project teams in its evaluation results. In other words, the communication evaluation system 2 can reflect indicators (benchmark values) set based on the level of communication in multiple other project teams in its evaluation results. This allows for a comparison of the communication situation with other project teams.

[0081] (Modified Version of Embodiment 2) This embodiment is a modified version of Embodiment 2 described above. The communication evaluation system 2a in this embodiment evaluates the quality of communication over multiple periods. The configuration of the communication evaluation system 2a is the same as that of the communication evaluation system 2, so it is not shown in the illustration. Other explanations that overlap with Embodiment 2 are omitted as appropriate.

[0082] The acquisition unit 21 of the communication evaluation system 2a acquires reaction data over multiple periods. When the communication evaluation system 2a determines the level of satisfaction for each of the multiple periods, it can evaluate the level of satisfaction for each period by relatively comparing the multiple levels of satisfaction obtained. The reaction data acquired by the acquisition unit 21 is associated with the timing information of when the reaction occurred. The acquisition unit 21 may further acquire action data and user data.

[0083] Here, "period" refers to any period that the user can define. The period could be, for example, one month, six months, or one year. Furthermore, multiple periods may be continuous or not. In other words, multiple periods may be discrete. To put it another way, multiple periods may include periods that are not included. Also, the period may be divided according to the stages of the project that the project team is targeting. Project stages include, for example, "kick-off," "requirements definition," "testing," "acceptance," "release," and "maintenance and operation."

[0084] The evaluation unit 23 evaluates the level of completeness of communication for each of several periods by comparing the multiple levels of completeness obtained relatively. Here, the evaluation unit 23 may evaluate the level of completeness according to different criteria for each period. Specifically, the quality analysis unit 232 of the evaluation unit 23 may determine the evaluation value of the quality of reactions (e.g., the degree of familiarity with casual conversation) according to different criteria for each period. For example, the degree of familiarity with reactions to casual conversation in the initial stages of a project team may be set to a higher standard than the evaluation value for the same reaction in subsequent periods. Furthermore, the evaluation result generation unit 233 may set different indicators (standard values) for the quantity and quality of reactions for each period and reflect the difference from these indicators in the evaluation result. For example, the standard value for the amount of casual conversation in the "kick-off" stage of the project process may be different from the standard value for the amount of casual conversation in the "test" stage.

[0085] Thus, the communication evaluation system 2a can evaluate and compare the level of communication over multiple periods. By evaluating and comparing the level of communication over multiple periods, users can understand the communication situation for each period and compare the communication situations for each period. Furthermore, by dividing the evaluation period for each project stage, for example, it is possible to understand the differences in the communication situation for each stage. Moreover, by setting different indicators for each project stage and evaluating the level of communication, project team managers and supervisors can take appropriate actions for each stage.

[0086] (Other Embodiments) The communication evaluation system described above may evaluate the user's work status based on information about the timing of the communication. That is, if communication takes place at an inappropriate time (for example, on a holiday or at night), the communication evaluation system may evaluate one or more users who are making such communication as users who should be careful about their working hours. In this case, if the communication evaluation system recognizes the existence of such users, it may output this information to the project team manager or person in charge.

[0087] (Hardware Configuration Example) Figure 14 shows a hardware configuration example of the communication evaluation system 3. In Figure 14, the communication evaluation system 3 includes a processor 31 and a memory 32. The processor 31 may be, for example, a microprocessor, an MPU (Micro Processing Unit), or a CPU (Central Processing Unit). The processor 31 may include multiple processors. The memory 32 is composed of a combination of volatile memory and non-volatile memory. The memory 32 may include storage located away from the processor 31. In this case, the processor 31 may access the memory 32 via an I / O (Input / Output) interface, which is not shown.

[0088] In the embodiments described above, the disclosure was explained as a hardware configuration, but the disclosure is not limited thereto. The communication evaluation method described in the flowchart of Figure 2 and the flowcharts of Figures 11 to 13 can also be implemented by having a CPU execute a computer program.

[0089] In the above example, the program can be stored and provided to the computer using various types of non-transitory computer-readable medium. Non-transitory computer-readable medium includes various types of tangible storage medium. Examples of non-transitory computer-readable medium include magnetic storage media (e.g., magneto-optical disks), CD-ROMs, CD-Rs, CD-R / Ws, and semiconductor memory (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, RAMs). Alternatively, the program may be provided to the computer using various types of transient computer-readable medium. Examples of transient computer-readable medium include electrical signals, optical signals, and electromagnetic waves. Transitory computer-readable medium can supply the program to the computer via wired communication channels such as electric wires and optical fibers, or via wireless communication channels. Computers include various information processing devices such as PCs, servers, CPUs, MPUs, FPGAs (Field Programmable Gate Arrays), and ASICs (Application Specific Integrated Circuits).

[0090] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure can be made as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0091] Each drawing is merely illustrative to illustrate one or more embodiments. Each drawing may be associated with one or more other embodiments, rather than being associated with only one specific embodiment. As those skilled in the art will understand, various features or steps described with reference to any one drawing can be combined with features or steps shown in one or more other drawings, for example, to create embodiments not explicitly shown or described. Not all features or steps shown in any one drawing to illustrate an exemplary embodiment are necessarily required, and some features or steps may be omitted. The order of steps described in any of the drawings may be changed as appropriate.

[0092] Some or all of the above embodiments may also be described as follows, but are not limited to the following: (Note 1) A communication evaluation system comprising: an acquisition unit that acquires reaction data which is data including the content of reactions made by other users to actions of a user in two-way communication between multiple users belonging to a project team; an evaluation unit that evaluates the richness of the communication between the multiple users based on at least the reaction data; and an output unit that outputs the evaluation of the richness by the evaluation unit. (Note 2) The communication evaluation system according to Note 1, further comprising an classification unit that classifies action data into a plurality of categories based on action data which includes the content of actions taken by the users in the communication, wherein the evaluation unit evaluates the richness from the reaction data corresponding to the action data according to the category. (Note 3) The communication evaluation system according to Note 2, wherein the evaluation unit evaluates the richness by calculating the number of at least the number of reactions belonging to the category. (Note 4) The communication evaluation system according to any one of Notes 1 to 3, wherein the evaluation unit evaluates by inputting the reaction data acquired by the acquisition unit into a trained model that outputs the level of completeness when the reaction data is input, and outputting the level of completeness. (Note 5) The communication evaluation system according to any one of Notes 2 to 4, wherein the classification unit classifies by inputting the action data acquired by the acquisition unit into a trained model that outputs category information to which the action belongs when the action data is input, and outputting the category information. (Note 6) The communication evaluation system according to any one of Notes 1 or 5, wherein the evaluation unit generates improvement proposals to improve the level of completeness of communication between multiple users based on the level of completeness, and the output unit outputs the improvement proposals.(Note 7) The communication evaluation system according to any one of Notes 1 to 6, wherein the acquisition unit acquires the reaction data for a plurality of project teams, and the evaluation unit evaluates the level of completeness using an index of completeness determined based on the reaction data for the plurality of project teams. (Note 8) The communication evaluation system according to any one of Notes 1 to 7, wherein the evaluation unit evaluates the multiple levels of completeness obtained by relatively comparing them when determining the level of completeness for each of a plurality of periods. (Note 9) The communication evaluation system according to any one of Notes 1 to 8, wherein the acquisition unit further acquires user data indicating the relationships between a plurality of users, and the evaluation unit evaluates the level of completeness based on the user data. (Note 10) The communication evaluation system according to any one of Notes 6 to 9, wherein the evaluation unit generates the improvement proposal by inputting the level of completeness evaluated by the evaluation unit into a trained model that outputs the improvement proposal when action data, reaction data, and level of completeness are input, and outputting the improvement proposal. (Note 11) The communication evaluation system according to any one of Notes 2 to 10, wherein the classification unit classifies the action data into the casual conversation category when the action is casual conversation, and the evaluation unit evaluates the level of completeness for the casual conversation category based on the level of familiarity of the reaction. (Note 12) The communication evaluation system according to any one of Notes 2 to 11, wherein the classification unit classifies the action data into the reporting category when the action is reporting, and the evaluation unit evaluates the level of completeness for the reporting category based on the level of responsiveness of the reaction. (Note 13) The communication evaluation system according to any one of Notes 2 to 12, wherein the classification unit classifies the action data into the question category when the action is a question, and the evaluation unit evaluates the level of completeness for the question category based on the level of response of the reaction.(Note 14) The communication evaluation system according to any one of Notes 2 to 13, wherein the classification unit classifies the action data into a discussion category when the action is a discussion, and the evaluation unit evaluates the level of completeness for the discussion category based on the degree of coherence of the discussions. (Note 15) A communication evaluation method wherein a computer acquires reaction data, which is data containing the content of reactions made by other users to an action of a user in two-way communication between multiple users belonging to a project team, evaluates the level of completeness of the communication between the multiple users based on at least the reaction data, and outputs the evaluation of the level of completeness. (Note 16) A program that causes a computer to perform the steps of: acquiring reaction data, which is data containing the content of reactions made by other users to an action of a user in two-way communication between multiple users belonging to a project team; evaluating the level of completeness of the communication between the multiple users based on at least the reaction data; and outputting the evaluation of the level of completeness.

[0093] Some or all of the elements (e.g., configuration and function) described in Appendices 2 to 14 that are dependent on Appendice 1 may also be dependent on Appendices 15 and 16 in the same way as in Appendices 2 to 14. Some or all of the elements described in any appendice may be applied to various hardware, software, recording means, systems, and methods for recording software.

[0094] This application claims priority based on Japanese Patent Application No. 2025-026596, filed on 21 February 2025, and incorporates all of its disclosures herein.

[0095] 1 Communication Evaluation System 2 Communication Evaluation System 2a Communication Evaluation System 3 Communication Evaluation System 11 Acquisition Unit 12 Evaluation Unit 13 Output Unit 21 Acquisition Unit 22 Classification Unit 23 Evaluation Unit 24 Evaluation Result Output Unit 31 Processor 32 Memory 221 Reception Unit 222 Input Unit 223 Output Unit 231 Quantitative Analysis Unit 231a Reception Unit 231b Input Unit 231c Output Unit 232 Qualitative Analysis Unit 232a Reception Unit 232b Judgment Unit 232c Input Unit 232d Output Unit 232e Integration Unit 233 Evaluation Result Generation Unit

Claims

1. A communication evaluation system comprising: an acquisition unit that acquires reaction data, which is data containing the content of reactions made by other users to user actions in two-way communication between multiple users belonging to a project team; an evaluation unit that evaluates the richness of the communication between the multiple users based on at least the reaction data; and an output unit that outputs the richness evaluation by the evaluation unit.

2. The communication evaluation system according to claim 1, further comprising a classification unit that classifies action data, which includes the content of actions taken by the user in the communication, into a plurality of categories, and an evaluation unit that evaluates the level of completeness from the reaction data corresponding to the action data according to the category.

3. The communication evaluation system according to claim 2, wherein the evaluation unit evaluates the level of fulfillment by calculating the number of at least the number of reactions belonging to the category.

4. The communication evaluation system according to any one of claims 1 to 3, wherein the evaluation unit evaluates by inputting the reaction data acquired by the acquisition unit into a trained model that outputs the level of completeness when the reaction data is input, and outputting the level of completeness.

5. The communication evaluation system according to claim 2, wherein the classification unit inputs the action data acquired by the acquisition unit into a trained model that outputs category information to which the action belongs when the action data is input, and outputs the category information.

6. The communication evaluation system according to claim 1, wherein the evaluation unit generates improvement proposals to improve the quality of communication among the multiple users based on the quality level, and the output unit outputs the improvement proposals.

7. The communication evaluation system according to claim 1, wherein the acquisition unit acquires the reaction data for a plurality of project teams, and the evaluation unit evaluates the level of completeness using the level of completeness index determined based on the reaction data for the plurality of project teams.

8. The communication evaluation system according to claim 1, wherein the evaluation unit evaluates the multiple levels of completeness obtained by relatively comparing them when determining the level of completeness for each of the multiple periods.

9. The communication evaluation system according to claim 1, wherein the acquisition unit further acquires user data indicating the relationships between a plurality of users, and the evaluation unit evaluates the level of completeness based on the user data.

10. The communication evaluation system according to claim 6, wherein the evaluation unit generates the improvement proposal by inputting the level of completeness evaluated by the evaluation unit into a trained model that outputs the improvement proposal when action data including the content of the actions taken by the user in the communication, reaction data, and the level of completeness are input, and outputting the improvement proposal.

11. The communication evaluation system according to claim 2, wherein the classification unit classifies the action data into a casual conversation category when the action is casual conversation, and the evaluation unit evaluates the level of fulfillment for the casual conversation category based on the level of friendliness of the reaction.

12. The communication evaluation system according to claim 2, wherein the classification unit classifies the action data into a report category when the action is a report, and the evaluation unit evaluates the level of completeness for the report category based on the degree of response of the reaction.

13. The communication evaluation system according to claim 2, wherein the classification unit classifies the action data into question categories when the action is a question, and the evaluation unit evaluates the level of completeness for each question category based on the degree of response of the reaction.

14. The communication evaluation system according to claim 2, wherein the classification unit classifies the action data into a discussion category when the action is a discussion, and the evaluation unit evaluates the level of completeness for the discussion category based on the degree of coherence of the discussions.

15. A communication evaluation method comprising: a computer acquiring reaction data, which is data containing the content of reactions made by other users to user actions in two-way communication between multiple users belonging to a project team; evaluating the richness of the communication between the multiple users based at least on the reaction data; and outputting the evaluation of the richness.

16. A program that causes a computer to perform the following steps: acquire reaction data, which is data containing the content of reactions made by other users to actions taken by a user in two-way communication between multiple users belonging to a project team; evaluate the quality of the communication between the multiple users based on at least the reaction data; and output the evaluation of the quality.

Images