JP7424710B1 - Information processing system, information processing method and program - Google Patents

Abstract

An object of the present invention is to evaluate and determine whether or not a message created by a user can appropriately express positive emotions toward a destination user before sending the message. An information processing system includes a control section. The control unit applies information about the appropriate expression of positive emotions to the sender included in the message to a learning model that is machine-learned. Enter unsent messages in . The control unit receives advice information regarding the appropriate expression in the input message from the learning model, and transmits the advice information to the sender terminal for display. [Selection diagram] Figure 4

Description

The present invention relates to an information processing system, an information processing method, and a program that can support the creation of messages sent in a business communication tool.

2. Description of the Related Art Conventionally, there has been a technology within a company that, when an employee sends a message to another employee using a business communication tool, analyzes the sent message and evaluates the user.

For example, Patent Document 1 below describes a method for diagnosing the level of activity and determining the risk of resignation based on the number of times each user sent or received thank you messages within a predetermined period. Are listed.

Patent No. 6379270

However, in the conventional technology described in Patent Document 1, the content of the message is not analyzed, and depending on the content of the message sent, there is a possibility that the feelings of the sender are not properly conveyed to the recipient. There is also.

In view of the above circumstances, an object of the present invention is to enable a user to evaluate and understand whether or not a message created by a user appropriately expresses positive emotions toward the destination user before sending it. The purpose of the present invention is to provide an information processing system, an information processing method, and a program.

In order to achieve the above object, an information processing system according to one embodiment of the present invention includes a control unit. The control unit applies information about the appropriate expression of positive emotions to the sender included in the message to a learning model that is machine-learned. Enter unsent messages in . The control unit receives advice information regarding the appropriate expression in the input message from the learning model, and transmits the advice information to the sender terminal for display.

This allows the information processing system to evaluate and understand whether the message created by the user appropriately expresses positive emotions toward the destination user before sending it, and to send a high-quality message. can. The learning model is, for example, a generative AI such as ChatGPT based on a large-scale language model (LLM), but is not limited thereto.

The control unit may receive evaluation value information indicating an evaluation value regarding the expression of emotion in the input message from the learning model, and transmit the evaluation value information to the sender terminal for display.

Thereby, the information processing system allows the user to objectively understand how well emotions are expressed in the message created by the user.

The control unit calculates the evaluation value based on a predetermined index regarding expressions of gratitude or praise included in the message, and generates information regarding the index related to the calculation of the evaluation value as the advice information. The learning model may be trained.

As a result, the information processing system can generate the above-mentioned evaluation value using an objective index, and also allows the user to grasp information regarding the index (the basis on which the evaluation value was calculated), thereby improving the message. .

The information processing system further includes a storage unit that stores user information regarding the user of the business communication tool in association with advice requesting right information regarding the user's right to request the advice information from the learning model. Good too. In this case, when the control unit receives an advice request requesting the advice information from a business communication tool used by the user, the control unit determines whether or not the right exists based on the stored advice request right information. If it is determined that the user has the right, the advice information may be requested from the learning model along with inputting the message.

As a result, if the information processing system incurs usage costs, such as when using the learning model on a pay-per-use basis, if the user uses the learning model inexhaustibly, the budget of the user's company will be significantly squeezed. However, usage per user can be managed so that the amount of usage is appropriate to the budget.

The information processing system may further include a storage unit that stores behavior history information of the user accumulated using a predetermined business tool. In this case, the control unit calculates a business relationship value indicating the level of business relationship between the user and another user based on the stored behavior history information, and calculates a business relationship value indicating the level of business relationship between the user and another user, and determines that the business relationship value is high. The learning model is trained to extract other users, information on other users with high business relationship values is received from the learning model, and the evaluation value of the advice information created so far is sent to the message. calculate statistical values for each destination user, train the learning model to extract other users with low statistical values, and extract information on other users with low statistical values from the learning model. Recommendation information that recommends another user who has a high business relationship value and a low statistical value as a new message destination may be transmitted to the sender terminal.

Thereby, the information processing system can induce the user to find a starting point for sending a message to another user who has a strong business relationship with the user but has not been able to properly convey feelings of gratitude or admiration in a message. The above-mentioned statistical values are, for example, average values, maximum values, minimum values, etc., but may also be values obtained by performing other mathematical processing.

The control unit transmits, as the recommendation information, list information of other users who have a high business relationship value and a low statistical value, in response to a request from the sender terminal. The message may be displayed on the sender terminal together with a user interface for displaying a message creation screen.

This allows the information processing system to display a list of other users to whom messages should be sent as well as an interface for creating messages, allowing users to intuitively perform tasks ranging from understanding the need to send a message to creating a message. can be set.

The control unit learns the learning model so that when the user is asked about the basis for determining that the other user's business relationship value is high and the statistical value is low, the control unit responds with information that is the basis for determining that the other user's business relationship value is high and the statistical value is low. and when the question from the user is received from the sender terminal via the chatbot, the information serving as the basis is received from the learning model and displayed on the sender terminal via the chatbot. You may let them.

Thereby, the information processing system can make the user intuitively understand the objective reason for sending a message to another user via the chatbot.

The control unit causes the sender terminal to display list information of other users with a high business relationship value together with a user interface for requesting the statistical values of the other users, and to When another user is selected, information indicating the statistical value of the other user is sent to the sender terminal along with a user interface for displaying a screen for creating a message with the other user as the destination. It may be displayed.

As a result, the information processing system allows the user to select a user to whom he or she is interested from among other users to whom messages should be sent, and to grasp the evaluation values of messages sent to that user so far. The need to send messages to other users can be more emphasized.

An information processing method according to another embodiment of the present invention includes:
Messages created by the user using business communication tools on the sender's device, but not yet sent to the recipient, are applied to a learning model that machine-learns information about the appropriate expression of positive emotions to the sender included in the message. Enter
receiving advisory information regarding the appropriate expression in the input message from the learning model;
The method includes transmitting the advice information to the sender terminal and displaying the advice information.

A program according to another aspect of the present invention is provided in an information processing device,
Messages created by the user using business communication tools on the sender's device, but not yet sent to the recipient, are applied to a learning model that machine-learns information about the appropriate expression of positive emotions to the sender included in the message. and the step of entering
receiving advisory information regarding the appropriate expression in the input message from the learning model;
A step of transmitting the advice information to the sender terminal and displaying the advice information is executed.

As described above, according to the present invention, it is possible to evaluate and determine whether or not a message created by a user appropriately expresses positive feelings toward the destination user before sending the message. However, this effect does not limit the present invention.

1 is a diagram showing the configuration of a message creation support system according to a first embodiment of the present invention. 1 is a diagram showing a hardware configuration of a mediation server according to a first embodiment of the present invention. FIG. FIG. 2 is a diagram showing the configuration of a database included in the mediation server according to the first embodiment of the present invention. FIG. 2 is a sequence diagram showing the flow of advice information transmission processing by the message creation support system according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of a user interface for advice information displayed on a user terminal by the mediation server according to the first embodiment of the present invention. FIG. 2 is a sequence diagram showing the flow of a process for extracting a message recipient by the message creation support system according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by the mediation server according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by the mediation server according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by the mediation server according to the first embodiment of the present invention. FIG. 7 is a sequence diagram showing the flow of a process of extracting a message recipient by the message creation support system according to the second embodiment of the present invention. FIG. 7 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by the mediation server according to the second embodiment of the present invention. FIG. 7 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by the mediation server according to the second embodiment of the present invention. FIG. 7 is a sequence diagram showing the flow of a process for extracting a message recipient by the message creation support system according to the third embodiment of the present invention. FIG. 7 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by a mediation server according to a third embodiment of the present invention. FIG. 7 is a diagram showing an example of a user interface of message transmission target person recommendation information displayed on a user terminal by a mediation server according to a third embodiment of the present invention. FIG. 12 is a sequence diagram showing the flow of a process for extracting a message recipient by the message creation support system according to the fourth embodiment of the present invention. FIG. 12 is a sequence diagram showing the flow of a process for extracting a message recipient by the message creation support system according to the fifth embodiment of the present invention.

Embodiments of the present invention will be described below with reference to the drawings.

<First embodiment>
First, a first embodiment of the present invention will be described.

[System configuration]
FIG. 1 is a diagram showing the configuration of a message creation support system according to this embodiment.

As shown in the figure, this system is configured by connecting two intermediary servers 100A and 100B, a generation AI 300, and a plurality of user terminals 200 of each company via a network such as the Internet 50. .

The user terminal 200 is a terminal used by a user who is an employee of a company, and is, for example, a notebook PC (personal computer), a desktop PC, a smartphone, a tablet PC, or the like.

A business communication tool 20 (hereinafter also simply referred to as business tool 20) is installed in the user terminal 200, which is used by employees of each company to communicate with each other in business.

Business communication tools 20 include mailers such as Outlook (registered trademark), groupware such as Microsoft Teams (registered trademark) and Slack (registered trademark), messaging apps, group chat apps, and business SNS (Social Networking Service). It also includes tools (hereinafter also referred to as ``emotional communication tools'') that specialize in emotional communication between employees (transmitting gratitude and praise to colleagues within the company).

The mediation server 100A is a server that mediates communication between the user terminals 200 using general business tools such as Outlook (registered trademark) and Microsoft Teams (registered trademark) among the business tools described above.

The mediation server 100B is a server that mediates communication between the user terminals 200 using the emotional communication tool among the employees among the business tools.

The generation AI 300 is a learning model based on a large-scale language model (LLM), such as ChatGPT, and is connected to the mediation servers 100A and 100B via an API.

In this embodiment, the intermediary servers 100A and 100B handle message creation when the user (employee) of the user terminal 200 sends the above-mentioned emotional communication message (thank you/praise message) to another employee. Execute processing to support.

Specifically, before sending the message, the mediation server 100B has the generation system AI 300 evaluate and advise whether or not the message created by the user appropriately expresses positive emotions toward the destination user. It is possible to generate information and provide the advice information to the user terminal 200.

Furthermore, in this embodiment, the intermediary servers 100A and 100B express gratitude to the generation AI 300 based on the user's past exchanges of gratitude and praise messages with other employees and business interactions using business tools. - It is possible to induce the user to create a message by extracting other employees to whom a praise message should be sent and providing information on the target employee to the user terminal 200.

[Hardware configuration of mediation server]
FIG. 2 is a diagram showing the hardware configuration of the mediation server 100B. As shown in the figure, the mediation server 100B includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, an input/output interface 15, and a bus 14 that connects these to each other. Equipped with.

The CPU 11 accesses the RAM 13 and the like as necessary, and performs various arithmetic operations while collectively controlling each block of the mediation server 100B. The ROM 12 is a non-volatile memory in which firmware such as an OS, programs, and various parameters to be executed by the CPU 11 is permanently stored. The RAM 13 is used as a work area for the CPU 11, and temporarily stores the OS, various applications being executed, and various data being processed.

A display section 16, an operation reception section 17, a storage section 18, a communication section 19, etc. are connected to the input/output interface 15.

The display unit 16 is a display device using, for example, an LCD (Liquid Crystal Display), an OELD (Organic ElectroLuminescence Display), a CRT (Cathode Ray Tube), or the like.

The operation reception unit 17 is, for example, a pointing device such as a mouse, a keyboard, a touch panel, or other input device. When the operation receiving section 17 is a touch panel, the touch panel can be integrated with the display section 16.

The storage unit 18 is, for example, a nonvolatile memory such as an HDD (Hard Disk Drive), a flash memory (SSD; Solid State Drive), or other solid state memory. The storage unit 18 stores the OS, various applications, and various data.

In particular, in this embodiment, the storage unit 18 stores user information of the emotional communication tool, information regarding gratitude/praise messages exchanged between users using the tool, and uses these data to perform mediation. The server 100B stores applications and other programs for executing message creation support processing. As will be described later, the storage unit 18 includes a user information database, an advice request right information database, and an emotional score information database as databases containing such data.

The communication unit 19 is, for example, various modules for wireless communication such as an Ethernet NIC (Network Interface Card) or a wireless LAN, and is responsible for communication processing between the user terminal 200, the mediation server 100A, and the generation system AI 300.

Although not shown, the hardware configuration of the mediation server 100A and the user terminal 200 are also similar to the hardware configuration of the mediation server 100B.

[Mediation server database configuration]
FIG. 3 is a diagram showing the configuration of a database included in the mediation server 100B.

As shown in the figure, the mediation server 100B has a user information database 31, an advice request right information database 32, and an emotional score information database 33 in the storage unit 18.

The user information database 31 stores user information of each employee, such as user name, user ID, position, department, date of joining, age, gender, address, profile image, etc.

The advice request right information database 32 stores information regarding the user's right to request advice regarding a thank you/praise message from the generation AI 300. This is because if the generation system AI 300 is a pay-as-you-go system, if the user uses the generation system AI 300 inexhaustibly, it will greatly strain the user's company's budget. This is to manage the usage amount of the generation AI 300 per user.

Specifically, the intermediary server 100B controls the user's right to request advice from the generation AI 300 (hereinafter also referred to as "advice request right") by using one of the following four management methods (1) to (4). Or manage by a combination of multiple methods. Note that the detailed conditions below are merely examples and are not limited thereto.

(1) Managing the number of advice requests: Grant rights to each user, for example, 5 times per month, and manage the remaining number.
(2) Management by points: Grant 1000 points/month to the user, set control to consume 200 points per advice request, and manage remaining points.
(3) Managing the time from the previous advice request to the next: Set a control that requires a gap of at least two days between the previous advice request and the next one, and manage the interval between advice requests by time.
(4) Manage actions from the previous advice request to the next: Set a control in which the advice request is locked when an advice request is made, and is unlocked by the action of sending a thank you/praise message.・Manage by whether or not there is an action to send a praise message.

Then, the advice request right information database 32 stores, for each user, the remaining number of advice requests per month, remaining points per month, elapsed time since the previous advice request, etc., according to the management methods described in (1) to (4) above. It manages information such as whether advice requests are locked or not.

The emotional score information database 33 indicates an evaluation value (emotional score) generated by the generative AI 300 by evaluating expressions of gratitude or praise included in a message when the user requests advice from the generative AI 300. Information is stored per user and per message.

Although not shown, the intermediary server 100A accumulates behavior history data of each user in business tools such as Outlook (registered trademark) and Microsoft Teams (registered trademark). Behavior history information includes, for example, the number of chat messages sent, the length of online meetings, the number of shared documents, the number of mentions, the number of emails sent and received, and the sending and receiving of reactions (likes, emoji stamps), etc. These include, but are not limited to:

These databases are mutually referenced and used as necessary in message creation support processing by mediation servers 100A and 100B, which will be described later.

[System operation]
Next, the operation of the message creation support system configured as above will be explained. This operation is a collaboration between hardware such as the CPU 11 and communication unit 19 of the mediation server 100B, the CPU and communication unit of the mediation server 100A, and software stored in the storage unit 18 of the mediation server 100B and the storage unit of the mediation server 100A. It is carried out by working. In the following description, for convenience, the CPUs of the mediation servers 100A and 100B are assumed to be the main operating bodies.

[Advice information provision processing]
First, the process of providing advice information in response to the gratitude/praise message by the message creation support system will be described.

FIG. 4 is a sequence diagram showing the flow of advice information transmission processing by the system. Further, FIG. 5 is a diagram showing an example of a user interface for advice information displayed on the user terminal 200 by the mediation server 100B. Figure 5 (A) is a display example when the user has the right to request advice and the message satisfies the advice permission conditions, and Figure 5 (B) shows a display example when the user does not have the right to request advice or the message does not satisfy the advice permission conditions. This is a display example.

As shown in FIG. 4, the user terminal 200 first receives input from the user in the message input field 51 of the gratitude/praise message creation screen on the emotional communication tool, as shown in FIG. - Create a praise message (step 41).

Subsequently, the user terminal 200 accepts an advice request regarding the created thank-you/praise message via the advice request button 52 placed, for example, at the bottom right of the message input field 51 on the thank-you/praise message creation screen (step 42 ), and transmits the advice request to the mediation server 100B (step 43).

When the CPU 11 of the mediation server 100B receives the advice request (step 44), it refers to the advice requesting right information database 32 and determines whether the advice requesting user has the right to request advice (step 45).

If it is determined that the user does not have the right to request advice, the CPU 11 of the intermediary server 100B displays information to that effect in the advice display field 53 of the gratitude/praise message creation screen, as shown in FIG. 5(B). let Alternatively, in this case, the CPU 11 may control the advice request button 52 itself to be inactive (inoperable).

If it is determined that the user has the right to request advice, the CPU 11 of the mediation server 100B determines whether the created message satisfies the advice availability condition (step 46).

Specifically, the CPU 11 determines that a message that satisfies the following two conditions satisfies the advice availability condition, but the determination conditions are not limited to these.
・The number of characters must be 50 or more.
・Words expressing gratitude such as thank you, thanks, and thank you, and words expressing praise such as amazing, wonderful, wonderful, etc., must be used at least once.

If it is determined that the message does not satisfy the advice acceptance condition, the CPU 11 of the mediation server 100B displays information to that effect in the advice display column 53 of the gratitude/praise message creation screen, as shown in FIG. 5(B). Display. Alternatively, in this case, the CPU 11 may control the advice request button 52 itself to be inactive (inoperable).

If it is determined that the message satisfies the advice availability condition, the CPU 11 of the mediation server 100B transmits an advice generation request together with the message body to the generation system AI 300 (step 47).

When the generation AI 300 receives the advice generation request (step 48), it analyzes the message, calculates an emotional score based on a predetermined index regarding expressions of gratitude and praise included in the message, and points out points for improvement. By generating the information, advice information including the information is generated (step 49).

Since the generation system AI 300 executes the processing, the CPU 11 of the mediation server 100B calculates the emotional score of the message based on the following four indicators and judgment conditions (1) to (4), for example. The generation system AI 300 is trained via the API. The indicators and judgment conditions are not limited to these.

(1) Repeatedly conveying feelings of gratitude and admiration (out of 10 points): 5 points for 1 time, 10 points for 2 times.
(2) Describes the reason for gratitude or praise in detail (40 points): Expresses the details, appropriateness, etc. of what the other person did in what situation and how it contributed. Judgment is made based on the number of expressions and matching with pre-registered expressions.
(3) Praising the other person for gratitude or praise (as expected, thanks to..., etc.) (out of 40 points): Are you able to pay attention to the other person's characteristics (skills, mindset, personality, etc.) and express them appropriately? The score will be calculated based on the number of expressions expressed.
(4) Promise to give something back to the other person (10 points): If the answer is not written, 0 points will be given, and if it is written, it will be given a full score.

The CPU 11 uses sentences with high scores and sentences with low scores as teacher data in light of the indicators (1) to (4) above, and edits the sentences so that the former has a high score and the latter has a low score. The generation system AI 300 repeatedly executes a process of causing the generation system AI 300 to respond with a score in response to the input.

Regarding the weighting, the emotional score is calculated by setting (1) to 10%, (2) to 40%, (3) to 30%, and (4) to 20%, but this is not the only setting. isn't it.

Furthermore, the CPU 11 generates a comment that points out the factors that result in a low score as points for improvement in the emotional score calculation index, and causes the generation system AI 300 to learn to reply with the comment along with the emotional score. For example, if the index (2) above is insufficient in specificity, the device is trained to respond with more specific comments.

Note that depending on the learning status of the generation AI 300, a new learning process for generating the above advice information may be added or the above learning process may be deleted, and the presence or absence of a learning process may be determined as appropriate depending on the situation. may be determined.

After the generation AI 300 that has performed the above learning generates advice information, it transmits it to the mediation server 100B (step 50). When the mediation server 100B receives the advice information, it transmits it to the user terminal 200 (step 51), and displays it in the advice display column 53 on the message creation screen of the user terminal 200, as shown in FIG. 5(A). (Step 52).

The advice in Figure (A) has an emotional score of 74 points, and among the above indicators, the high evaluation points are that it repeatedly uses words of gratitude and also conveys praise to the other person. It includes a comment that the point for improvement is that the reasons for this are not specific.

Then, when the user of the user terminal 200 improves the message based on the advice and presses the send button 54 in the lower right corner of the message creation screen, the user terminal 200 sends the message to the destination user terminal 200 ( Step 53).

[Extraction process of message recipients]
Next, a description will be given of a process in which the intermediary servers 100A and 100B extract a target person to whom the user should send a message when creating the thank-you/praise message. FIG. 6 is a sequence diagram showing the flow of the process of extracting a message recipient by the message creation support system. Further, FIGS. 7 to 9 are diagrams showing examples of user interfaces for message transmission target person recommendation information displayed on the user terminal 200 by the mediation server 100B.

As shown in FIG. 6, the user terminal 200 first extracts colleagues with low emotional evaluation values from among people with high business relationship values with the user, based on the user's behavior history data accumulated in intermediary server A. The request is accepted (step 61).

For example, as shown in FIG. 7, the request for the extraction process is made by pressing the sending opportunity question button 71 provided on the user's My Page to ask the generation AI 300 about the sending opportunity (sending party). Accepted if detected.

When the extraction request is accepted, the user terminal 200 transmits the extraction request to the intermediary server 100A (step 62). Upon receiving the extraction request (step 63), the mediation server 100A transmits an extraction processing request to the generation AI 300 (step 64).

Upon receiving the extraction processing request, the generation AI 300 extracts colleagues who have many business connections with the user (high business relationship value) based on the user's behavior history data accumulated by the business tool (step 65). ).

In order for the generation AI 300 to execute the extraction process, the CPU of the mediation server 100A makes the generation AI 300 learn to extract other users using, for example, the following data items and judgment conditions for them. However, the data items and judgment conditions are not limited to these. Further, any of the following data items may be determined alone, or a plurality of data items among the following may be determined in combination.

・Total number of chat messages sent and received: Colleagues whose total number of sent and received chat messages is 200 or more per month are judged to have a high business relationship value.
- Total number of chat messages sent and received outside of work hours: Colleagues whose total number of chat messages sent and received outside of work hours is 100 or more per month are judged to have a high business relationship value.
・Total time spent in online meetings: Colleagues who attended the same online meetings for more than 35 hours/month are judged to have a high business relationship value.
・Total amount of time spent in online meetings held outside of work hours: Colleagues who attended the same online meetings outside of work hours for 10 hours or more per month are judged to have a high business relationship value.
・Total number of shared documents: Taking presentation documents as an example, colleagues who share more than 20 PowerPoints per month or 400 slides per month will be It is determined that the relationship value is high. The document may be a variety of documents, tables, etc., and the determination conditions can be set with reference to the numbers in the presentation document.
・Total number of mentions: 40 or more per month, and colleagues who have been mentioned are judged to have a high business relationship value.
・Total number of emails sent and received: Colleagues whose total number of sent and received chats is 150 or more per month are judged to have a high business relationship value.
・Total number of reactions (likes, emoji stamps) sent and received: Colleagues with a total number of sent and received reactions of 200 or more per month are judged to have a high business relationship value.

Furthermore, instead of making the determination based on the monthly action history data as described above, the intermediary server 100A may make the generation AI 300 learn to make the determination based on cumulative data, for example, as described below.
・Total number of chat messages sent and received: Extract colleagues with a cumulative total of 200 or more chat messages sent and received.
・Total time spent in online meetings: Extract colleagues who have attended the same online meeting for a cumulative total of 35 hours or more.
・Total number of mentions: Extract colleagues who have cumulatively given 40 or more mentions.

Further, regarding the determination of the business relationship value, instead of using the two choices of high/other, another determination method may be used as described below.
・Judge the business relationship value in three levels (high, medium, low): For example, a total of 200 or more chats sent and received is judged high, 100 or more but less than 200 is judged medium, and 100 or less is judged low. Of course, the determination may be made in further multiple stages such as A, B, C, and so on.
- Determining the business relationship value by score: Standardize the value of each sent/received chat with a value of 1, and if there are 200, the value will be 200, and the higher the number, the higher the value.

Depending on the learning status of the generative AI 300, a new learning process for extracting colleagues with high business relationship values may be added, or the above learning process may be deleted, as appropriate depending on the situation. , the presence or absence of a learning process may be determined.

When the generation AI 300 that has undergone the above learning generates information regarding colleagues with high business relationship values, it transmits it to the mediation server 100A (step 66).

When the CPU of the mediation server 100A receives information about colleagues with high business relationship values from the generation AI 300 (step 67), the CPU of the mediation server 100A sends information to the mediation server 100B about colleagues with low emotional scores among the colleagues with high business relationship values. An extraction request is sent (step 68).

Upon receiving the extraction request (step 69), the mediation server 100B transmits an extraction processing request to the generation AI 300 (step 70).

The generation AI 300, which has received the extraction process request, extracts colleagues who have been determined to have low emotional scores in messages sent by the user to the colleague from among colleagues with whom the business relationship value is high (step 71).

In order for the generation AI 300 to execute the extraction process, the CPU 11 of the mediation server 100B makes the generation AI 300 learn to extract co-workers under the following handling and determination conditions, for example. However, the conditions are not limited to these.

・Scope of emotional score: Emotional scores of messages sent in the last three months are targeted, but it is also possible to go back six months or a year.
・Mathematical processing of emotional scores: One example is to use statistical values such as the average, maximum, and minimum of the emotional scores of messages sent to each recipient, but other mathematical processing can also be used. You may use the values you have obtained.
- Judgment of emotional score (1): An average value of 70 points or more is judged as high, an average value of 40 points or more and less than 70 points is judged as medium, and an average value of less than 40 points is judged as low.
- Judgment of emotional score (2): A maximum value of 85 points or more is judged as high, a value of 55 points or more but less than 85 points is judged as medium, and a maximum value of less than 55 points is judged as low.
- Judgment of emotional score (3): A minimum value of 55 points or more is judged as high, a value of 25 points or more but less than 55 points is judged as medium, and a minimum value of less than 25 points is judged as low.
・If you have not sent any messages, the emotional score will be determined as 0 points.
- Correct the emotional score value by taking into account the number of messages sent: A coefficient may be applied depending on the number of messages sent. For example, if the number of sent items exceeds 10 per month or a total of 100, multiply by a factor of 1, and if less than that, multiply by a factor of 0.8.

In this embodiment, the intermediary server 100B trains the generation AI 300 to return the colleague's emotional score itself extracted by the above method, but instead of this, the mediation server 100B learns the emotional score judgment result (high, low etc.) may be returned.

Regarding the timing of calculating the emotional score, the system may automatically score the message at the timing when it is sent and store it in the emotional score information database 33, so that the calculation of all messages may be completed in advance. , only the target messages may be calculated each time an emotional score extraction request is received.

Note that depending on the learning status of the generative AI 300, a new learning process for extracting colleagues with low emotional scores may be added, or the above learning process may be deleted, as appropriate depending on the situation. The presence or absence of a learning process may also be determined.

When the generative AI 300 that has learned as described above extracts a co-worker with a low emotional score from co-workers with a high business relationship value, it sends a message with information about the extracted co-worker and his/her emotional score. It is transmitted to the intermediary server 100B as recommendation information of the target person (step 72).

The mediation server 100B that received the target person recommendation information transfers the target person recommendation information to the mediation server 100A (step 73), and the mediation server 100A transmits the target person recommendation information to the user terminal 200 (step 74). ) is displayed in the destination recipient list display field 72 (FIGS. 7 and 8) on the my page of the user terminal 200 (step 75).

Although FIG. 7 shows the state before pressing the sending opportunity question button 71 (no extraction results), if the colleague extraction process has been executed in the past, the target person recommendation as the previous extraction result The information may be displayed in the recipient list display field 72.

As shown in Figure 8, the target person recommendation information includes each target person's name (handle name), department, business relationship value (high, medium, low), and emotional score (statistics). , For example, at the right end of each target person's information, a message creation button 73 for creating a message addressed to the target person is displayed.

When the user notices that a certain target has a particularly low emotional score among the targets displayed in the list and presses the message creation button 73 for that target, the message is displayed as shown in FIG. , the above thank you/praise message creation screen will be superimposed on your My Page. Thereafter, as described above, the user terminal 200 creates a thank you/praise message by accepting the user's input in the message input field 51 (step 76), and further when the advice request button 52 is pressed. , an advice request is sent, and advice information is returned from the generation system AI 300 via the mediation server 100B.

In addition to the business relationship value in the above candidate recommendation information, even if information about what specific work support was provided (e.g. help with creating materials, etc.) is displayed. Good, in addition to the emotional score, information on the specific state of lack of gratitude and praise (e.g., are you sending too few messages, are you sending sloppy messages, etc.) may be displayed.

[summary]
As explained above, according to the present embodiment, the message creation support system evaluates whether the message created by the user appropriately expresses feelings of gratitude and admiration for the destination user before sending it. This allows you to understand and send high-quality messages. Furthermore, it is possible to induce the user to find a starting point for sending a message to another user who has a strong business relationship with the user but has not been able to properly convey feelings of gratitude or admiration through a message.

<Second embodiment>
Next, a second embodiment of the present invention will be described. In this embodiment and subsequent embodiments, components common to those in the first embodiment are designated by the same reference numerals, and explanations thereof will be omitted or simplified.

FIG. 10 is a sequence diagram showing the flow of the process of extracting a message recipient by the message creation support system according to this embodiment. Further, FIGS. 11 and 12 are diagrams showing examples of user interfaces for message transmission target person recommendation information displayed on the user terminal 200 by the mediation servers 100A and 100B according to the present embodiment.

In the first embodiment described above, the user asked the generation AI 300 about the opportunity to send a message via the sending opportunity question button 71 provided on the my page, but in this embodiment, the user By asking the bot if there is an opportunity to send a message, the results of extracting colleagues will be displayed in a chat format. The user can also ask questions about the basis for extracting co-workers on the chat screen and receive a response from the generation AI 300.

As shown in FIG. 10, the processing in steps 61 to 71 is the same as in FIG. 6 above. When the generation system AI 300 extracts a colleague with a low emotional score from colleagues with a high business relationship value, in addition to the extracted transmission target recommendation information, the generation system AI 300 receives information from the user who has a high business relationship value and a low emotional score. When asked about the basis for determining that it is low, information regarding the basis is also transmitted to the mediation server 100B (step 72').

The mediation server 100B that received the transmission target person information and basis information transfers it to the mediation server 100A (step 73'), and the mediation server 100A further transmits it to the user terminal 200 (step 74'). Display it on the chat screen (step 75').

Since the generation system AI 300 executes the processing, the CPU 11 of the intermediary server 100B informs the user that when extracting co-workers with high business relationship values and low emotional scores, The generation system AI 300 is trained to respond with information (extraction logic) that is the basis for determining that the value is low via a chatbot when asked about the basis for determining that the condition is low.

As shown in FIG. 11, when the user inputs a question on the chat screen asking the generation AI 300 which colleague should send a thank you/praise message, the content of the question is displayed in a question display balloon 81a. Then, through the processing of steps 72' to 74', the message transmission target person recommendation information generated by the generation system AI 300 is displayed in the response display balloon 82a.

Furthermore, as shown in FIG. 12, when the user inputs a question asking about the basis for determining that one of the colleagues included in the message transmission target recommendation information has a high business relationship value and a low emotional score, This is displayed in the question display balloon 81b, and the answer to it is displayed in the answer display balloon 82b. Specifically, the logic that determined that the business relationship value was high was that the number of non-work chats sent and received and the number of attendance at the same online meeting were high, and the logic that was determined that the emotional score was low was that It has been pointed out that the messages do not mention the specific actions of the recipient, nor are they aware of the politeness of the recipient.

As described above, according to the present embodiment, the message creation support system allows the user to intuitively understand, via the chatbot, the objective reason for sending a message of gratitude or praise to another colleague.

<Third embodiment>
Next, a third embodiment of the present invention will be described.

FIG. 13 is a sequence diagram showing the flow of the process of extracting a message recipient by the message creation support system according to the present embodiment. Further, FIGS. 14 and 15 are diagrams showing examples of user interfaces for message transmission target person recommendation information displayed on the user terminal 200 by the mediation servers 100A and 100B according to the present embodiment.

In each of the embodiments described above, co-workers with high business relationship values are extracted, and among them co-workers with low emotional scores are extracted, and message recommendation information is finally presented to the user. . However, in this embodiment, the user himself/herself can select a co-worker whose emotional score is to be calculated from co-workers who have a high business relationship value.

As shown in FIG. 13, steps 81 to 87 are similar to steps 61 to 67 in FIG. 6 of the first embodiment described above.

When the mediation server 100A receives information about colleagues with high business relationship values from the generation AI 300 via the mediation server 100B (step 87), it transmits it to the user terminal 200 as transmission candidate information (step 88). .

When the user terminal 200 receives the transmission candidate information, it displays it in the transmission target person list display field 72 of my page (step 89). As shown in FIG. 14, in the sender list display field 72, the emotional score column of the sender recommendation information of the above-described embodiment is displayed blank, and in addition to the message creation button 73, An emotional score request button 74 is provided.

When the user detects a pressing operation of the emotional score request button 74 (selection operation of the candidate for which the emotional score is to be calculated) for any candidate, the user terminal 200 sends the selection request to the mediation server 100B. The candidate's emotional score extraction request is sent (step 90).

Upon receiving the emotional score extraction request for the selected candidate (step 91), the CPU 11 of the mediation server 100B refers to the emotional score information database 33 and extracts the statistics of the emotional score of the selected candidate. The value is extracted (step 92), and the extracted emotional score information is transmitted to the user terminal 200 (step 93).

When the user terminal 200 receives the emotional score information, it displays the emotional score in the corresponding candidate column of the recipient list display column 72 on the my page, as shown in FIG. 15 (step 94). In the example shown in the figure, "TAKE SUZUKI" is selected from among the candidates, and its emotional score (average 32 points) is displayed.

Then, when the user presses the message creation button 73 regarding the colleague whose emotional score has been extracted, the gratitude/praise message creation screen is superimposed on the my page as shown in the figure. Thereafter, as described above, the user terminal 200 creates a thank you/praise message by accepting the user's input in the message input field 51 (step 95), and further when the advice request button 52 is pressed. , an advice request is sent, and advice information is returned from the generation system AI 300 via the mediation server 100B.

In this way, in this embodiment, the message creation support system allows the user to select one of the candidates (whom the user is interested in) from among the candidates to whom the message should be sent (colleagues with whom the business relationship value is high), By making the user understand the emotional scores of messages sent to the candidate so far, it is possible to impress on the user the necessity of sending messages to the candidate and induce them to create a message.

<Fourth embodiment>
Next, a fourth embodiment of the present invention will be described.

In each of the embodiments described above, the mediation server 100A and the mediation server 100B exist as separate servers, and the mediation server 100A executes the process of extracting colleagues with high business relationship values, and the mediation server 100B performs the process of extracting colleagues with high emotional scores. An example of performing a low colleague extraction process was shown. However, the processes of both mediation servers may be executed by one server.

FIG. 16 is a sequence diagram showing the flow of the process of extracting a message recipient by the message creation support system according to the present embodiment. A mediation server 100C shown in the figure is a server that has a business tool management function like the mediation server 100A, and also has an emotional communication tool management function like the mediation server 100B.

As shown in the figure, in this embodiment, a mediation server 100C is responsible for each process of the mediation server 100A and the mediation server 100B shown in FIG. 6 of the first embodiment. In this way, even a single server can execute the extraction process in the first embodiment (and second embodiment) described above.

<Fifth embodiment>
Next, a fifth embodiment of the present invention will be described.

FIG. 17 is a sequence diagram showing the flow of the process of extracting a message recipient by the message creation support system according to the present embodiment. The mediation server 100C shown in the figure is a server that has a management function of a business communication tool like the mediation server 100A described above, and also has a management function of an emotional communication tool like the mediation server 100B.

As shown in the figure, in this embodiment, a mediation server 100C is responsible for each process of the mediation server 100A and the mediation server 100B shown in FIG. 13 of the third embodiment. In this way, even a single server can execute the extraction process in the third embodiment described above.

<Modified example>
The present invention is not limited to the above-described embodiments, and may be modified in various ways without departing from the gist of the present invention.

In each of the embodiments described above, emotional scores are calculated for expressions of gratitude and praise included in messages, and advice information is generated, but emotions are not limited to gratitude and praise. For example, emotional scores may be calculated and advice information may be generated regarding other positive emotions toward other users, such as approval (mainly from superiors to subordinates) and empathy.

In each of the above-described embodiments, the recipient of the thank-you/praise message is identified through the extraction process, and after the user creates the message, the message is An example was shown in which advice information is displayed on the user terminal 200 in response to an advice request from a user in an unsent state, and the user transmits a message upon receiving the advice. However, after creating the message, the user terminal 200 may transmit the message to the destination without going through the above advice request. In this case, after sending the message, the processes of steps 44 to 52 in FIG. 4 are executed after receiving the user's evaluation request or automatically, and the advice information (evaluation information) including the emotional score is may be presented to the user.

In each of the above-described embodiments, after a message addressed to the recipient of the thank-you/praise message is created by the extraction process of the recipient of the thank-you/praise message, the message is determined as to whether or not the expression of gratitude/praise contained in the message is appropriate. Advice information was generated and presented to the user. However, the generation and presentation of this advice information is not essential, and the process may be completed only by extracting the recipient of the thank-you/praise message (and presenting information about the extracted recipient to the user). . That is, the processes shown in FIGS. 4 and 5 may be omitted, and only the processes shown in FIGS. 6 to 17 may be executed. Further, after a colleague with a high business relationship value and a low emotional score is extracted and presented as a transmission target, it is not essential for the user to actually create a message addressed to the transmission target. Conversely, only the advice information generation/presentation processing shown in FIGS. 4 and 5 may be executed, and the message transmission target person extraction/presentation processing shown in FIGS. 6 to 17 may be omitted. do not have.

In each of the embodiments described above, a generative AI is used as a machine learning model, but the present invention can also be implemented using other machine learning models.

Among the inventions described in the claims of this application, the invention described as an "information processing method" is one in which each step is automatically performed by at least one device such as a computer through information processing using software. Yes, it is not performed by humans using devices such as computers. In other words, the "information processing method" is an information processing method using computer software, and is not a method in which humans operate a calculation tool called a computer.

11...CPU
18...Storage unit 19...Communication unit 31...User information database 32...Advice request right information database 33...Emotional score information database 51...Message creation field 52...Advice request button 53...Advice display field 54...Send button 71...Sending opportunity Question button 72... Sending target list display field 73... Message creation button 74... Emotional score request button 81... Question display balloon 82... Response display balloon 100A... Mediation server 100B... Mediation server 100C... Mediation server 200... User terminal 300... Generation system AI

Claims (7)

a storage unit that stores user action history information accumulated by a predetermined business tool of the sender terminal;
Based on the stored action history information, a business relationship value indicating the level of business relationship between the user and another user is calculated, and other users with a high business relationship value are extracted. machine learning the learning model,
receiving information of other users with whom the user has a high business relationship value from the learning model;
transmitting recommendation information to the sender terminal that recommends another user with a high business relationship value as a destination for a new message created by a business communication tool of the sender terminal;
An information processing system comprising a control unit and. The information processing system according to claim 1,
The control unit includes:
training the learning model to calculate an evaluation value regarding the expression of emotion in the created message based on a predetermined index regarding the expression of gratitude or praise included in the message;
receiving evaluation value information indicating the evaluation value calculated for the created message from the learning model, transmitting the evaluation value information to the sender terminal and displaying it;
The learning model is trained to calculate a statistical value of the evaluation values of the messages created so far for each user to whom the message is sent, and to extract other users whose statistical value is low. Receive information on other users with low statistical values from the learning model,
An information processing system that transmits, to the sender terminal, the recommendation information that recommends another user who has a high business relationship value and a low statistical value as a new message destination. The information processing system according to claim 2,
The control unit transmits, as the recommendation information, list information of other users who have a high business relationship value and a low statistical value, in response to a request from the sender terminal. An information processing system that displays on the sender terminal together with a user interface for displaying a message creation screen. The information processing system according to claim 2 or 3,
The control unit includes:
When the user is asked about the basis for determining that the business relationship value of the other user is high and the statistical value is low, the learning model is trained to respond with information that is the basis,
When the question from the user is received from the sender terminal via a chatbot, the information serving as the basis is received from the learning model and displayed on the sender terminal via the chatbot. Information processing system. The information processing system according to claim 2,
The control unit causes the sender terminal to display list information of other users with a high business relationship value together with a user interface for requesting the statistical values of the other users, and to When another user is selected, information indicating the statistical value of the other user is sent to the sender terminal along with a user interface for displaying a screen for creating a message with the other user as the destination. Information processing system to display. Memorizes the user's behavior history information accumulated by a predetermined business tool on the sender's terminal,
Based on the stored action history information, a business relationship value indicating the level of business relationship between the user and another user is calculated, and other users with a high business relationship value are extracted. machine learning the learning model,
receiving information of other users with whom the user has a high business relationship value from the learning model;
transmitting recommendation information to the sender terminal that recommends another user with a high business relationship value as a destination for a new message created by a business communication tool of the sender terminal;
Information processing method. In the information processing device,
a step of storing user action history information accumulated by a predetermined business tool of the sender terminal;
Based on the stored action history information, a business relationship value indicating the level of business relationship between the user and another user is calculated, and other users with a high business relationship value are extracted. a step of performing machine learning on the learning model;
receiving information about other users with whom the user has a high business relationship value from the learning model;
transmitting to the sender terminal recommendation information that recommends another user with a high business relationship value as a destination for a new message created by a business communication tool of the sender terminal;
A program to run.

Images