JP2023025793A - Human resource training support system, linkage support system, method, and computer program - Google Patents

Abstract

To provide a human resource training support system which can provide role-playing environment to increase opportunities for improving skill of a user.SOLUTION: A human resource training support system operates in conjunction with an online conference system configured to issue access destination information to a virtual meeting room depending on a request. The human resource training support system comprises: a dialogue processing unit which accesses the virtual meeting room by using the access destination information acquired directly or indirectly from the online conference system, and acquires utterance data on a dialogue performed between a trainee and a trainer who participate with the virtual meeting room; a voice analysis unit configured to perform voice analysis with respect to the utterance data; an evaluation generation unit configured to generate evaluation data depending on to a result of matching between the result of the voice analysis and evaluation reference information; and an evaluation database which stores the evaluation data.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to information processing technology for supporting human resource development through role-playing through a communication network.

Conventionally, as a human resource development method for acquiring the knowledge and skills necessary for work in companies, a method in which trainers and trainees (persons who receive guidance and education from trainers) divide roles and perform role-playing has been widely adopted. ing. 2. Description of the Related Art In recent years, many online communication tools such as web conferencing tools have been provided that enable voice calls and video calls via communication networks such as the Internet. Human resource development methods based on role-playing using such online communication tools are also widely adopted in companies.

There is also a computer system that automatically evaluates the skills of trainees by analyzing voices uttered by trainees during role-playing using online communication tools. One such computer system is disclosed, for example, in Japanese Patent Laying-Open No. 2016-76788 (Patent Document 1).

Patent Literature 1 discloses an evaluation server for evaluating the content of a call made between a person in charge terminal and a customer terminal via an integrated server that implements CTI (Computer Telephony Integration). This evaluation server has a function of supporting the self-training of the person in charge. That is, the evaluation server includes a preparation processing unit that plays the role of a trainer who assumes the role of a customer for role-playing, a speech recognition unit that converts the speech uttered by the person in charge during the role-playing into an uttered text, An evaluation unit that evaluates the content of the call of the person in charge from the spoken text using an evaluation module related to evaluation items such as keywords, call duration, and speed of speech (see paragraphs [0026] to [0033] of Patent Document 1).

JP 2016-76788 A

In conventional systems that enable human resource development through role-playing using online communication tools, the user may not be able to easily use the role-playing environment due to its low convenience. Therefore, there is a problem that opportunities for role-playing are suppressed or lost, and human resource development is not promoted.

In view of the above, an object of the present disclosure is to provide a human resource development support system, a cooperation support system, a method, and a computer program that can provide a role-playing environment and increase opportunities for skill improvement of users. is.

A human resource development support system according to the first aspect of the present disclosure is human resource development support that operates in cooperation with an online conference system configured to issue access destination information to a virtual conference room in response to a request by a trainer or trainee. A system comprising: a support control unit that directly or indirectly acquires the access destination information from the online conference system; and accesses the virtual conference room using the access destination information and participates in the virtual conference room. and a dialogue processing unit configured to acquire utterance data of a dialogue between the trainer and the trainee, and perform voice analysis by performing voice recognition on the utterance data to generate a utterance text. The configured speech analysis unit performs matching between the speech analysis result and pre-registered evaluation criteria information, and generates evaluation data indicating the evaluation result of the trainee according to the matching result. and an evaluation database storing the evaluation data.

A collaboration support system according to a second aspect of the present disclosure is characterized by comprising the human resource development support system according to the first aspect and an online conference system.

A method according to a third aspect of the present disclosure is performed in a human resource development support system operating in conjunction with an online conference system configured to issue access information to a virtual conference room in response to a request by a trainer or trainee. wherein the step of accessing the virtual conference room using the access destination information obtained directly or indirectly from the online conference system, and the trainer and the trainee participating in the virtual conference room a step of acquiring speech data of a dialogue made between; a step of performing speech analysis by subjecting the speech data to speech recognition to generate speech text; performing matching between information, generating evaluation data indicating the evaluation result of the trainee according to the matching result, and storing the evaluation data in an evaluation database. and

A computer program according to a fourth aspect of the present disclosure is a computer program read from a non-volatile memory and executed by one or more processors, wherein the method according to the third aspect is executed on the one or more processors. It is characterized in that it is configured to be implemented.

A human resource development support system according to a fifth aspect of the present disclosure opens a virtual conference room in response to a request from a trainee, and performs role-playing by automatic dialogue with the trainee participating in the virtual conference room. a dialogue processing unit configured to acquire utterance data of a trainee; a voice analysis unit configured to perform voice analysis by performing voice recognition on the utterance data to generate a utterance text; and the voice analysis. an evaluation generation unit configured to perform matching between the result of and pre-registered evaluation criteria information, and generate evaluation data indicating the evaluation result of the trainee according to the matching result; an evaluation database that stores data; a support control unit that is configured to receive setting information from an external communication terminal used by the trainee via a communication interface; a persona database including a plurality of persona data sets for performing the training, wherein the setting information includes information for selecting one virtual person image from among the plurality of virtual person images as a virtual trainer; The dialogue processing unit obtains from the persona database a persona data set for expressing the selected one virtual person image, and uses the obtained persona data set to communicate with the trainee The virtual trainer is configured to perform the role-playing.

A method according to a sixth aspect of the present disclosure is configured to open a virtual conference room in response to a request by a trainee, and perform role-playing through automatic interaction with the trainee participating in the virtual conference room. A method performed in a human resource development support system, wherein setting information including information for selecting one virtual person image from among a plurality of virtual person images as a virtual trainer is provided to an external device used by the trainee. Representing the selected one virtual person from a persona database including a plurality of persona data sets for representing each of the plurality of virtual person images. a step of acquiring a persona dataset for performing the role-playing as the virtual trainer with the trainee using the acquired persona dataset; Between the step of obtaining speech data of the trainee, the step of performing speech analysis by subjecting the speech data to speech recognition to generate speech text, and the result of the speech analysis and pre-registered evaluation criteria information. a step of executing matching, a step of generating evaluation data indicating the evaluation result of the trainee according to the result of the matching, and a step of storing the evaluation data in an evaluation database. be.

A computer program according to a seventh aspect of the present disclosure is a computer program read from a non-volatile memory and executed by one or more processors, wherein the method according to the sixth aspect is executed on the one or more processors. It is characterized in that it is configured to be implemented.

According to the first to fourth aspects, the human resource development support system operates in cooperation with an online conference system configured to issue access destination information to the virtual conference room in response to a request by a trainer or trainee. do. This human resource development support system accesses the virtual conference room using the access destination information obtained directly or indirectly from the online conference system, and the dialogue between trainers and trainees participating in the virtual conference room utterance data can be acquired. Furthermore, the human resource development support system performs speech analysis on the acquired speech data, executes matching between the speech analysis result and pre-registered evaluation criteria information, and evaluates data according to the matching result. can be stored in the rating database.

Therefore, according to the first to fourth aspects, the trainee can obtain an opportunity for role-playing in the virtual conference room with a very simple procedure, and accumulate evaluation data in the evaluation database. This makes it possible to increase opportunities for trainees to improve their skills without depending on time or place.

According to the fifth to seventh aspects, when the human resource development support system receives the setting information from the external communication terminal used by the trainee, the one virtual person selected by the setting information is selected from the persona database. A persona dataset can be obtained to represent the statue as a virtual trainer. The human resource development support system can use the acquired persona data set to perform role-playing with the trainee as a virtual trainer. Furthermore, the human resource development support system acquires speech data of trainees when role-playing is performed, performs speech analysis on the acquired speech data, and compares the result of the speech analysis with pre-registered evaluation criteria information. It is possible to perform matching between the data and store the evaluation data according to the result of the matching in the evaluation database.

Therefore, according to the fifth to seventh aspects, the trainee can obtain an opportunity for role-playing in the virtual conference room with a very simple procedure, and accumulate evaluation data in the evaluation database. This makes it possible to increase opportunities for trainees to improve their skills without depending on time or place.

1 is a block diagram showing a schematic configuration of a cooperation support system according to an embodiment of the present disclosure; FIG. 1 is a block diagram showing a schematic configuration of a cooperation support system according to an embodiment of the present disclosure; FIG. FIG. 10 is a diagram showing an example of matching between speech analysis results and evaluation criteria information; FIG. 10 is a diagram showing an example of matching between speech analysis results and evaluation criteria information; FIG. 10 is a diagram showing an example of matching between speech analysis results and evaluation criteria information; 2 is a block diagram schematically showing a hardware configuration example of a back system of the personnel training support system; FIG. 10 is a flow chart schematically showing an example of a processing procedure by the personnel training support system when role-playing is performed using the online conference system; FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a video call screen displayed on the communication terminal during role-playing; 4 is a flow chart schematically showing an example of a processing procedure for feeding back evaluation data to trainees; FIG. 5 is a diagram illustrating an evaluation screen represented by display data; FIG. 5 is a diagram illustrating an evaluation screen represented by display data; 10 is a flow chart schematically showing an example of processing procedures by the personnel training support system when operating as a virtual trainer. FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a setting screen prompting for input of setting information; FIG. 10 is a diagram illustrating a video call screen displayed on the communication terminal during role-playing;

Next, an embodiment according to the present disclosure and modifications thereof will be described in detail with reference to the drawings. It should be noted that constituent elements denoted by the same reference numerals throughout the drawings have the same configuration and the same function.

FIG. 1 is a block diagram showing a schematic configuration of a collaboration support system 1 according to an embodiment of the present disclosure. As shown in FIG. 1, the collaboration support system 1 includes an online conference system 10 configured to set a virtual conference room 11 in response to a request from a user and issue access destination information to this virtual conference room 11; A human resource development support system 12 that operates in cooperation with this online conference system 10 is provided. The online conference system 10 and the human resource development support system 12 are interconnected via a communication path (not shown) such as a communication network.

The virtual conference room 11 on the online conference system 10 is used for role-playing by dialogue (voice call or video call) between the trainee Te using the communication terminal T1 and the trainer Tr using the communication terminal T2. It is an electronic conference room where The human resource development support system 12 accesses the virtual conference room 11 using the access destination information, acquires the speech data of the dialogue between the trainee Te and the trainer Tr, and analyzes and evaluates the speech data. It has a function (hereinafter referred to as “dialogue analysis function”).

As shown in FIG. 1, the communication terminals T1 and T2 used by the trainee Te and the trainer Tr are configured to be capable of two-way communication with each of the online conference system 10 and the human resource development support system 12 via the communication network NW. It is configured. The communication network NW may be, for example, a local area network (LAN) or a wide area network (WAN) such as the Internet. Each of the communication terminals T1 and T2 is, for example, a PC (Personal Computer) having a communication function, or a mobile communication device such as a mobile phone terminal or a smart phone. A client C1 for using the service of the online conference system 10 and the service of the human resource development support system 12 is installed in the communication terminal T1, and the service of the online conference system 10 and the human resource development support are also installed in the communication terminal T2. A client C2 for using the services of the system 12 is installed.

In this embodiment, each of the communication terminals T1 and T2 has one client for using the service of the online conference system 10 and the service of the human resource development support system 12, but is limited to this. not something. Each of the communication terminals T1 and T2 may individually have a client for using the service of the online conference system 10 and a client for using the service of the human resource development support system 12. FIG.

In addition to the dialogue analysis function described above, the human resource development support system 12 has an automatic response system such as a chatbot that performs role-playing through automatic dialogue with the trainee Te without using the online conference system 10. I have. Hereinafter, the trainer played by the automatic response system will be referred to as a "virtual trainer".

The online conference system 10 is, for example, “Teams” (registered trademark) provided by Microsoft, “Skype” (registered trademark) provided by Microsoft, or “Google Meet” (registered trademark) provided by Google. , etc., as long as it is composed of a single server or a plurality of servers capable of implementing voice calls and video calls between a plurality of communication devices. The trainee Te or trainer Tr can access the online conference system 10 using the client C1 or C2 and request the online conference system 10 to set up the virtual conference room 11 . In response to the request, the online conference system 10 sets the virtual conference room 11 and issues access destination information for the virtual conference room 11 . The access destination information is information necessary for participating in the virtual conference room 11 . Examples of the access destination information include, but are not limited to, address information such as a URL (Uniform Resource Locator), date and time of the virtual meeting room 11, ID (identifier), and password or passcode. do not have.

As shown in FIG. 1, the human resource development support system 12 further includes a gateway server (GW server) 21, a front system 22 and a back system 23. FIG. The GW server 21 and the front system 22 constitute a communication interface 20 used by the back system 23 to transmit and receive data to and from external communication terminals (for example, communication terminals T1 and T2).

As shown in FIG. 1, the back system 23 includes a support control unit 31, a dialogue processing unit 35, a persona database (persona DB) 36, a scene database (scene DB) 37, a speech analysis unit 40, an evaluation generation unit 44, an evaluation It has a reference database (evaluation criteria DB) 45 and an evaluation database (evaluation DB) 47 . Furthermore, the voice analysis unit 40 has a voice recognition unit 41 and a natural language processing unit 42 .

The support control unit 31 has a function of authenticating an entity (user or device) requesting use of the service of the human resource development support system 12 . The trainee Te can log in to use the services of the human resource development support system 12 using the client C1 of the communication terminal T1. At this time, the communication terminal T1 transmits a request for using the service of the human resource development support system 12 and authentication information. The support control unit 31 can determine, for example, whether to permit use of the service using an external authentication server (not shown) and the authentication information. As authentication information, for example, a combination of an ID (user ID, group ID or device ID) and a password or passcode can be used, but the information is not limited to this. The trainer Tr can also log in to use the services of the human resource development support system 12 using the client C2 of the communication terminal T2.

The support control unit 31 has a function of receiving setting information input via the communication interface 20 from each of the communication terminals T1 and T2. The support control unit 31 can display a setting screen prompting the trainee Te or the trainer Tr to input setting information on the communication terminal T1 or the communication terminal T2. The trainee Te or trainer Tr can input setting information through the setting screen displayed on the communication terminal T1 or the communication terminal T2.

When the personnel training support system 12 executes the dialogue analysis function using the online conference system 10 , the setting information includes access destination information required to access the virtual conference room 11 . The dialogue processing unit 35 can acquire speech data input to the virtual conference room 11 by accessing the virtual conference room 11 using the access destination information. When the online conference system 10 directly transmits the access destination information to the human resource development support system 12, the support control unit 31 receives the access destination information via the communication interface 20, and converts the access destination information into the setting information. can also be included.

On the other hand, when the personnel training support system 12 operates as a virtual trainer without using the online conference system 10, a virtual conference room 32 is set in the front system 22 as shown in FIG. The virtual conference room 32 is an electronic conference room used for role-playing through dialogue (voice call or video call) between the trainee Te using the communication terminal T1 and the virtual trainer. The front system 22 can include an application server for setting up virtual conference rooms 32 and conducting online meetings. The dialogue processing unit 35 has an automatic dialogue function that performs role-playing through automatic dialogue with the trainee Te participating in the virtual conference room 32 using the persona DB 36 and the scene DB 37 . The automatic dialogue function of the dialogue processing unit 35 can be realized by a rule-based automatic response system, more preferably by an automatic response system using a machine learning model such as a deep learning model.

The persona DB 36 stores a plurality of persona data sets for representing a plurality of virtual personas (personas) as virtual trainer candidates. Since the setting information includes information for selecting one persona from a plurality of personas, the interactive processing unit 35 stores the persona data set for representing the one persona selected according to the setting information in the persona DB 36. can be obtained from The dialogue processing unit 35 can perform role-playing as a virtual trainer represented by the acquired persona data set. The support control unit 31 can cause the communication terminal T1 to display a setting screen prompting the trainee Te to select a desired persona from a plurality of personas. When the trainee Te selects a desired persona, the support control unit 31 includes information for selecting the desired persona in the setting information. On the other hand, when the trainee Te does not select a desired persona, the support control unit 31 can include information for selecting a default persona in the setting information.

In addition, the automatic dialogue function of the dialogue processing unit 35 dynamically generates a 3D model image imitating the character of the virtual trainer as an avatar during role-playing with the trainee Te, and supplies the avatar to the virtual conference room 32. can do.

The scene DB 37 stores a plurality of scene data sets representing a plurality of hypothetical scenes such as learning themes, scenarios, and customer attributes. The support control unit 31 can cause the communication terminal T1 or the communication terminal T2 to display a setting screen prompting the trainee Te or the trainer Tr to select a desired assumed scene from a plurality of assumed scenes. When the trainee Te or the trainer Tr selects a desired assumed scene, the support control unit 31 includes information for selecting the desired assumed scene in the setting information. On the other hand, when the trainee Te or the trainer Tr does not select a desired assumed scene, the support control unit 31 can include information for selecting a default assumed scene in the setting information.

When the human resource development support system 12 operates as a virtual trainer without using the online conference system 10, the automatic dialogue function of the dialogue processing unit 35 generates a scene data set representing one hypothetical scene selected according to the setting information. It is possible to acquire from the scene DB 37 and perform role-playing based on an assumed scene represented by the acquired scene data set.

The dialogue processing unit 35 acquires speech data input by the trainee Te during role-playing between the trainee Te and the trainer Tr or during role-playing between the trainee Te and the virtual trainer, and converts the speech data to is supplied to the speech analysis unit 40 . The voice analysis unit 40 performs voice analysis on the speech data and outputs the voice analysis result to the dialogue processing unit 35 . More specifically, the speech recognition unit 41 performs speech recognition on the speech data input from the dialogue processing unit 35 to generate speech text. The natural language processing unit 42 can perform natural language processing (NLP) including various analyzes such as morphological analysis, syntactic analysis, and semantic analysis on the generated utterance text.

When the voice analysis unit 40 supplies the data indicating the voice analysis result, the automatic dialog function of the dialog processing unit 35 generates a response text corresponding to the voice analysis result based on the persona data set and the scene data set, The response text can be provided in the text display area of virtual conference room 32 (FIG. 2). Alternatively, the automatic dialogue function of the dialogue processing unit 35 can convert the response text into speech data using a known speech synthesis technology and supply the speech data to the virtual conference room 32 .

On the other hand, the data representing the speech analysis result generated by the speech analysis unit 40 is also supplied to the evaluation generation unit 44 via the dialogue processing unit 35 and the support control unit 31 . The evaluation generation unit 44 quantitatively performs matching between the speech analysis result and evaluation criteria information registered in advance in the evaluation criteria DB 45, and based on the result of the matching, evaluates the trainee Te as an evaluation result. data can be generated. The generated evaluation data is stored in the evaluation DB 47 by the support control unit 31 .

3, 4 and 5 are diagrams showing examples of matching between speech analysis results and evaluation criteria information, respectively. In the example in Fig. 3, the message "Yes, may I ask why Mr. XXXX joined Super Cancer Insurance?" , that is, the speech text 50A is obtained. The natural language processing unit 42 generates a natural language processing result, that is, a speech analysis result 51A by performing natural language processing on the spoken text 50A. The voice analysis result 51A includes "yes", "XXXX", "sama", "ga", "super", "cancer", "insurance", "ni", "go", "subscription", "not It consists of multiple linguistic units, each of which has a meaning, such as "ta", "motivation", "wo", "o", "inquire", "do it", "yoroshii", and "maybe?" . The evaluation criteria information 52A includes a plurality of evaluation keywords such as [cue], [reason], [concern], . . . The evaluation generation unit 44 matches the evaluation keyword included in the speech analysis result 51A with the evaluation keyword included in the speech analysis result 51A by executing matching between the evaluation keyword included in the evaluation criteria information 52A and a plurality of linguistic units included in the speech analysis result 51A. Or you can find similar linguistic units. In the example of FIG. 3, the matching result 53A has the linguistic unit "kikake". For example, the evaluation generation unit 44 can calculate an evaluation score that increases as the degree of matching or similarity between the speech analysis result 51A and the evaluation criterion information 52A increases. Specifically, for example, the evaluation generating unit 44 can evaluate that the degree of matching or the degree of similarity is higher as the number of linguistic units matching or similar to the evaluation keyword in the speech analysis result 51A increases.

The natural language processing unit 42 may vectorize each language unit. In this case, the evaluation criteria DB 45 can store vectorized evaluation keywords. The evaluation generator 44 performs matching between the vectorized linguistic units and the vectorized evaluation keywords (e.g., calculates cosine similarity or vector space distance) to obtain a match or similarity can be calculated. The natural language processing unit 42 generates vectorized linguistic units by converting each linguistic unit into a real-valued vector called a distributed representation using, for example, a neural network model. be able to. The method of distributed representation is disclosed, for example, in Non-Patent Document 1 below, but is not limited to this.

・Non-Patent Document 1: Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg Corrado and Jeffrey Dean, "Distributed Representations of Words and Phrases and their Compositionality," Advances in Neural Information Processing Systems 26, pp. 3111-3119, 2013.

In the example of FIG. 4, "Yes. Now for the information provision. Well then, from my side, I would like to talk a little about the current medical situation. Yes, are you in the hospital now? ], that is, the spoken text 50B is obtained. As in the case of FIG. 3, the natural language processing unit 42 generates a natural language processing result, that is, a speech analysis result 51B by performing natural language processing on the spoken text 50B. The evaluation generating unit 44 generates evaluation keywords ([medical circumstances], [treatment methods], [advanced medical care], ...) included in the evaluation criteria information 52B and a plurality of linguistic units included in the speech analysis result 51B. By executing matching, linguistic units that match or are similar to the evaluation keyword can be found from the speech analysis result 51B. In the example of FIG. 4, the matching result 53B has the linguistic unit "medical circumstances".

In addition, in the example of Figure 5, it is said that ``It has been a while since you signed up for Super Cancer Insurance, so I asked you to explain the details of the contract. ], that is, the spoken text 50C is obtained. As in the case of FIG. 3, the natural language processing unit 42 generates a natural language processing result, that is, a speech analysis result 51C by performing natural language processing on the spoken text 50C. The evaluation generation unit 44 generates evaluation keywords ([contract], [description], [confirmation], [description], ...) included in the evaluation criteria information 52C and a plurality of language units included in the speech analysis result 51C. By executing the matching, it is possible to find linguistic units that match or are similar to the evaluation keyword from the speech analysis result 51C. In the example of FIG. 5, the matching result 53C has linguistic units of "contract", "content", and "description".

The evaluation criteria information described above may be a Boolean operation expression that is a combination of a plurality of evaluation keywords and at least one Boolean operator (logical operator). Boolean operators include, for example, logical negation (NOT) operators, disjunctive (OR) operators, logical AND (AND) operators, and exclusive OR (XOR) operators. The evaluation generator 44 can find a linguistic unit that matches the Boolean expression by performing matching between the Boolean expression and a plurality of linguistic units included in the speech analysis result. In the example of FIG. 3, the evaluation criteria information 52A may have the following Boolean operation expression S using the AND operator "&". In this case, no linguistic unit that matches the Boolean operation formula S is found in the speech analysis result 51A.
S = [Trigger] & [Reason] & [Worry]

In response to a request from the client C1 of the communication terminal T1, the support control unit 31 acquires evaluation data specified by the request from the evaluation DB 47, and can display evaluation results on the communication terminal T1 based on the evaluation data. It has a function to generate display data for The support control unit 31 can transmit the display data to the communication terminal T1 via the communication interface 20. FIG. For example, if the client C1 has a web client (web client) function, the support control unit 31, in response to a request from the client C1, writes a web written in a markup language such as HTML (Hypertext Markup Language). The page should be generated as a display page. Thereby, the trainee Te can confirm the evaluation result displayed on the screen of the communication terminal T1.

The configuration of the human resource development support system 12 described above may be realized by a single information processing device, or may be realized by a distributed processing system comprising a plurality of information processing devices interconnected via communication paths. may In addition, all or part of the human resource development support system 12 may be realized in the form of on-premises in which hardware equipment such as physical servers and storages are managed by the company itself, or hardware It may be implemented in the form of a cloud that does not require the company to manage the device itself. Cloud refers to the use of services provided by other business operators through the Internet.

FIG. 6 is a block diagram schematically showing an information processing device (computer) 60, which is an example of the hardware configuration of the backup system 23 of the human resource development support system 12. As shown in FIG. The signal processing device 60 includes a processor 61 that executes digital signal processing, a random access memory (RAM) 62, a nonvolatile memory 63, a large capacity memory 64, an input/output interface circuit 65, and a signal path 66. It is A signal path 66 is a bus for interconnecting the processor 61, RAM 62, nonvolatile memory 63, mass memory 64 and input/output interface circuit 65. FIG. The RAM 62 is a data storage area used when the processor 61 executes digital signal processing. When the processor 61 incorporates an arithmetic unit such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), the nonvolatile memory 93 is a data storage that stores program code of software or firmware executed by the processor 91. have an area.

Next, the processing contents of the personnel training support system 12 will be described in detail.

FIG. 7 shows an example of a processing procedure by the human resource development support system 12 when role-playing is performed between the trainee Te and the trainer Tr using the virtual conference room 11 (FIG. 1) on the online conference system 10. 2 is a schematic flow chart;

After the virtual conference room 11 is set in the online conference system 10, the trainer Tr logs into the human resource development support system 12 using the client C2 of the communication terminal T2. In response to this, the support control unit 31 executes setting processing (step S11 in FIG. 7). That is, the support control unit 31 causes the communication terminal T2 to display a setting screen prompting the trainer Tr to input setting information, and in response to this, receives setting information input from the communication terminal T2 via the communication interface 20. . Since the communication terminal T2 is equipped with input devices such as a key input device and a pointing device (for example, a mouse), the trainer Tr can input setting information using the input device while viewing the setting screen displayed on the communication terminal T2. can be entered. Here, instead of the trainer Tr, the trainee Te may use the client C1 of the communication terminal T1 to log in to the human resources development support system 12 and input the setting information. After receiving the setting information, the support control unit 31 waits until the online conference starts (NO in step S12).

8, 9, and 10 are diagrams illustrating GUI (Graphical User Interface) screens for prompting input of setting information, that is, setting screens SW01, SW02, and SW03. The setting screen SW01 of FIG. 8 includes a selectable button 71 for returning to the previously displayed screen (not shown), a selectable button 72 for displaying the next setting screen SW02, and a learning theme. , and a selection area 73 for selecting one hypothetical scene consisting of a combination of scenarios and customer attributes. In the example of FIG. 8, the selection area 73 includes option buttons (radio buttons) for selecting any one of the three learning themes of "single item edition", "combined sale edition" and "comprehensive security edition". , an option button (radio button) for selecting one of the two scenarios of "review of cancer insurance" and "review of medical insurance", and "53-year-old male" and "65-year-old male". and three buttons for selecting any of the three customer attributes "46 year old male".

Referring to FIG. 9, the setting screen SW02 includes a selectable button 75 for returning to the setting screen SW01 of FIG. 8, a selectable button 76 for displaying the next setting screen SW03, and "approach", " Selection areas 77, 78, 79, 80, 81, 82 for selecting six dialogue sections of "explanation of existing contract content", "providing information", "confirmation of needs", "presentation" and "closing". include. The support control unit 31 can display, on the setting screen SW02, a plurality of dialogue sections that constitute the dialogue of the hypothetical scene selected on the previous setting screen SW01.

Each of the selection areas 77, 78, 79, 80, 81, 82 in FIG. 9 is provided with a check box. For example, the trainer Tr (or trainee Te) can select an arbitrary dialogue section from six dialogue sections by selecting or deselecting each check box. One or more evaluation keywords can be preset for each dialogue section. Furthermore, a Boolean expression, which is a combination of multiple evaluation keywords and at least one Boolean operator (logical operator), can be set for each dialogue section. The evaluation criteria DB 45 (FIG. 1) stores one or more evaluation keywords or Boolean expressions set for each of a plurality of dialogue sections. For example, when the trainer Tr (or the trainee Te) selects the dialog section "explanation of existing contract content", the evaluation generation unit 44 generates the evaluation keywords set for the selected dialog section (for example, "opportunity", " Insurance money") and the audio analysis results can be performed to generate evaluation data for each dialogue section.

Next, referring to FIG. 10, the setting screen SW03 has a selectable button 84 for returning to the setting screen SW02 of FIG. 9, a selectable button 85 for starting the online conference, and the URL of the address destination information. and an input box 86 for input. The trainer Tr (or trainee Te) can start the online conference in the virtual conference room 11 by selecting the button 85 after entering the URL in the input box 86 . As a result, an online conference between the communication terminals T1 and T2 through the virtual conference room 11 is started. FIG. 11 is a diagram illustrating a video call screen TW displayed on communication terminal T1 or communication terminal T2 during the online conference. A captured image 88 of the trainee Te and a captured image 89 of the trainer Tr are shown on the video call screen TW.

Referring to FIG. 7, when the online conference is started, the support control unit 31 cancels the standby state (NO in step S12). Next, the dialogue processing unit 35 uses the access destination information to access the virtual conference room 11 of the online conference system 10 (step S13). Then, the dialogue processing unit 35 acquires the utterance data of the dialogue between the trainee Te and the trainer Tr participating in the virtual conference room 11 from the virtual conference room 11 (step S14). At this time, the speech analysis unit 40 performs speech analysis on the speech data supplied from the dialogue processing unit 35, and supplies the speech analysis result to the evaluation generation unit 44 via the dialogue processing unit 35 and the support control unit 31. there is

After the role-playing is finished, the evaluation generation unit 44 quantitatively matches the speech analysis result obtained from the speech analysis unit 40 with the evaluation criterion information registered in advance in the evaluation criterion DB 45 (step S15). Next, as described above, the evaluation generation unit 44 generates evaluation data indicating the evaluation result of the trainee Te according to the matching result (step S16). The generated evaluation data is stored in the evaluation DB 47 (step S17). Note that in the example of FIG. 7, steps S15 to S17 are executed after role-playing ends, but instead, steps S15 to S17 may be executed concurrently with role-playing.

Next, a processing procedure for feeding back the evaluation data stored in the evaluation DB 47 to the trainee Te will be described. FIG. 12 is a flow chart schematically showing an example of a processing procedure for that feedback.

When the support control unit 31 receives a request for evaluation results from the communication terminal T1 via the communication interface 20 (GW server 21 and front system 22), the support control unit 31 acquires the evaluation data specified by the request from the evaluation DB 47 (Fig. 12, step S21), based on the obtained evaluation data, display data that enables the evaluation data to be displayed on the communication terminal T1 is generated (step S22). After step S22, the support control unit 31 transmits the generated display data to the communication terminal T1 via the communication interface 20 (step S23). Thereby, the communication terminal T1 can display the evaluation result represented by the display data.

FIG. 13 is a diagram illustrating an evaluation screen EW1 represented by display data. The evaluation screen EW1 of FIG. 13 includes a display area 92 showing the overall evaluation, a display area 93 showing the evaluation result of the dialog section "approach", and a display area showing the evaluation result of the dialog section "explanation of existing contract details". 94, a display area 95 showing the evaluation results of the dialogue section "Information", a display area 96 showing the evaluation results of the dialogue section "Ascertaining needs", and a display showing the evaluation results of the dialogue section "Presentation". It has an area 97, a display area 98 showing the evaluation result of the dialogue section "closing", a display area 99 showing the utterance ratio of the trainee Te, and a display area 100 showing the utterance speed of the trainee Te.

As noted above, one or more evaluation keywords and Boolean expressions can be preset for each interaction section. It is also possible to set multiple check criteria for each dialog section and set one or more evaluation keywords and Boolean expressions for each check criteria. In the example of FIG. 13, three check criteria are provided for the dialogue section "approach", and the evaluation result indicates that three of these three check criteria are satisfied ("3/3 (100%)"). )It is shown. In addition, three check criteria are provided for the dialogue section "explanation of existing contract details", and the evaluation result indicates that two of these three check criteria are satisfied ("2/3 (67%)"). )It is shown. Display area 120 shows an overall rating (“13/19 (68%)”) that 13 of the 19 check criteria set for all interactive sections are met.

When the trainee Te operates the input device to select the display area 94, the evaluation screen EW1 of FIG. 13 switches to the evaluation screen EW2 of FIG. As shown in FIG. 14, a display area 94d showing the evaluation contents for each check criterion appears for the dialog section "description of existing contract details". As shown in this display area 94d, three check criteria are provided, namely, "motivation for subscription", "explanation of outline of coverage", and "explanation of points to note about coverage", and a plurality of evaluation keywords are provided for each check criterion. I know it is set.

Also referring to FIG. 13, display area 99 shows role-playing time, talk time, listen time and talk-listen ratio. The talk time is the time that the trainee Te spoke during the role-playing time, the listen time is the time that the trainee Te was listening during the role-playing time, and the talk-listen ratio is the ratio of the talk time to the role-playing time. ratio.

As described above, the human resource development support system 12 operates in cooperation with the online conference system 10 when executing the dialogue analysis function. The online conference system 10 is configured to set a virtual conference room 11 and issue access destination information to the virtual conference room 11 in response to a request from a trainer Tr or trainee Te. The dialogue processing unit 35 accesses the virtual conference room 11 using the access destination information obtained directly or indirectly from the online conference system 10, and communicates between the trainee Te and the trainer Tr participating in the virtual conference room 11. utterance data of the dialogue can be obtained. Furthermore, the evaluation generation unit 44 of the human resource development support system 12 performs matching between the result obtained by performing voice analysis on the acquired speech data and the pre-registered evaluation criteria information, and the result of the matching is The corresponding evaluation data can be stored in the evaluation DB 47 . Therefore, the trainee Te can obtain an opportunity for role-playing through a very simple procedure, and can accumulate his own evaluation data in the evaluation DB 47 . As a result, it is possible to increase the opportunities for the trainee Te to improve his/her skills and to increase the opportunities for acquiring knowledge necessary for the work without depending on time and place.

Further, the dialogue processing unit 35 acquires a scene data set representing an assumed scene desired by the trainer Tr (or trainee Te) from the scene DB 37, and can perform role-playing using the acquired scene data set. can. This makes it possible to select an assumed scene suitable for skill improvement of the trainee Te.

Next, a case where the human resource development support system 12 operates as a virtual trainer without using the online conference system 10 will be described. At this time, the support control unit 31 causes the front system 22 to set the virtual conference room 32 (FIG. 2). FIG. 15 is a flow chart schematically showing an example of processing procedures by the human resource development support system 12 when operating as a virtual trainer.

When the trainee Te logs into the human resource development support system 12 using the client C1 of the communication terminal T1, the support control unit 31 executes setting processing (step S31 in FIG. 15). That is, the support control unit 31 causes the communication terminal T1 to display a setting screen prompting the trainee Te to input setting information, and in response to this, receives setting information input from the communication terminal T1 via the communication interface 20. . Since the communication terminal T1 is equipped with input devices such as a key input device and a pointing device (for example, a mouse), the trainee Te can input setting information using the input device while viewing the setting screen displayed on the communication terminal T1. can be entered.

16, 17 and 18 are diagrams exemplifying GUI screens prompting the input of setting information, that is, setting screens SW11, SW12, and SW13. The setting screen SW11 of FIG. 16 includes a selectable button 101 for returning to the previously displayed screen (not shown), a selectable button 102 for displaying the next setting screen SW12, and a learning theme. , and a selection area 103 for selecting one hypothetical scene consisting of a combination of scenarios and customer attributes. Similar to the selection area 73 in FIG. 8, the selection area 103 in FIG. It has three buttons for selection.

The support control unit 31 selects at least one virtual trainer candidate (virtual person image candidate) that matches the one assumed scene selected by the trainee Te, and selects for the trainee Te from among the virtual trainer candidates It has a function of displaying a setting screen prompting selection of one virtual trainer on the communication terminal T1. The setting screen SW12 in FIG. 17 is a screen to be displayed after the trainee Te selects one assumed scene. As shown in FIG. 17, the setting screen SW12 includes a button 105 that can be selected to return to the setting screen SW11 of FIG. 16, a button 106 that can be selected to display the next setting screen SW13, and three virtual screens. and option buttons 107, 108, 109 for selecting one virtual trainer from among the potential trainers. The setting screen SW12 presents virtual trainer candidates that match the hypothetical scene selected on the setting screen SW11.

Referring to FIG. 18, the setting screen SW13 includes a selectable button 111 for returning to the setting screen SW12 of FIG. and selection areas 113, 114, 115, 116, 117, 118 for selecting six dialogue sections: "Description", "Provide Information", "Identify Needs", "Presentation" and "Closing". Each of the selection areas 113, 114, 115, 116, 117, 118 is provided with a check box. The trainee Te can select an arbitrary dialogue section from among the six dialogue sections by setting each check box to either a selected state or a non-selected state. One or more evaluation keywords can be preset for each dialogue section. Furthermore, a Boolean expression, which is a combination of multiple evaluation keywords and at least one Boolean operator (logical operator), can be set for each dialogue section. The evaluation criteria DB 45 (FIG. 2) stores one or more evaluation keywords or Boolean expressions set for each of a plurality of dialogue sections. For example, when the trainee Te selects the dialog section “explanation of existing contract content”, the evaluation generation unit 44 generates the evaluation keywords (for example, “opportunity”, “insurance money”) set for the selected dialog section. and speech analysis results to generate evaluation data for each dialogue section.

After trainee Te has finished selecting the dialogue section, he can select button 112 to start the online conference.

Referring to FIG. 15, after the setting information is accepted, the support control unit 31 activates the virtual conference room 32 (FIG. 2) for the online conference (step S32). Next, the dialogue processing unit 35 acquires a persona data set for representing one persona from the persona DB 36 according to the setting information (step S33), and then selects one assumed scene from the scene DB 37 according to the setting information. A scene data set for representation is obtained (step S34).

Next, the dialogue processing unit 35 acquires speech data input by the trainee Te from the virtual conference room 32 while performing role-playing by automatic dialogue with the trainee Te participating in the virtual conference room 32 (step S35). ). At this time, the dialogue processing unit 35 performs role-playing as a virtual trainer represented by the persona data set acquired in step S33 based on the hypothetical scene represented by the scene data set acquired in step S34. Since the speech analysis unit 40 performs speech analysis on the speech data supplied from the dialogue processing unit 35 and outputs the speech analysis result to the dialogue processing unit 35, the dialogue processing unit 35 performs can perform role-playing.

FIG. 19 is a diagram exemplifying a video call screen RW displayed on the communication terminal T1 during role-playing. The video call screen RW includes an image display area 120, a selectable button 121 for interrupting the video call, a selectable button 122 for redoing the role-playing, and a button 122 for ending the role-playing and obtaining evaluation data. and a button 123 that can be selected from the In the image display area 120, a 3D model image (avatar) 124 imitating the character of the virtual trainer and a captured image 125 of the trainee Te are displayed.

Referring to FIG. 15, after the role-playing is finished, the evaluation generation unit 44 quantifies the matching between the speech analysis result obtained from the speech analysis unit 40 and the evaluation criteria information registered in advance in the evaluation criteria DB 45. (step S36). Next, as described above, the evaluation generation unit 44 generates evaluation data indicating the evaluation result of the trainee Te according to the matching result (step S37). The generated evaluation data is stored in the evaluation DB 47 (step S38). In the example of FIG. 15, steps S36 to S38 are executed after role-playing ends, but instead, steps S36 to S38 may be executed concurrently with role-playing.

After the role-playing between the trainee Te and the virtual trainer is completed, the support control unit 31 connects the communication terminal T1 to the communication interface 20 as in the case where the dialogue analysis function is executed in cooperation with the online conference system 10. When a request for an evaluation result is received via the communication terminal T1, the evaluation data specified by the request is acquired from the evaluation DB 47, and based on the acquired evaluation data, the evaluation data can be displayed on the communication terminal T1. Data can be generated and sent. Thereby, the communication terminal T1 can display the evaluation results represented by the display data (for example, the evaluation screens EW1 and EW2 in FIGS. 13 and 14).

As described above, when the human resource development support system 12 operates as a virtual trainer without using the online conference system 10, when setting information is received from the external communication terminal T1 used by the trainee Te, the persona DB It is possible to acquire a persona data set for expressing one virtual human figure selected by the setting information as a virtual trainer. The human resource development support system 12 can perform role-playing as a virtual trainer with the trainee Te using the acquired persona data set. Further, the human resource development support system 12 acquires the speech data of the trainee Te when the role-playing is performed, performs voice analysis on the acquired speech data, and obtains the result of the voice analysis and evaluation criteria information registered in advance. , and the evaluation data corresponding to the matching result can be stored in the evaluation DB 47 . Therefore, the trainee Te can obtain an opportunity for role-playing through a very simple procedure, and can accumulate his own evaluation data in the evaluation DB 47 . As a result, it is possible to increase the opportunities for the trainee Te to improve his/her skills and to increase the opportunities for acquiring knowledge necessary for the work without depending on time and place.

Further, the dialogue processing unit 35 can acquire a scene data set representing a hypothetical scene desired by the trainee Te from the scene DB 37 and perform role-playing using the acquired scene data set. As a result, the trainee Te can also select an assumed scene suitable for improving his/her skill.

Various embodiments and modifications thereof according to the present disclosure have been described above, but the above-described embodiments and modifications are merely examples, and do not limit the scope of the present invention. It should also be understood that changes, additions and improvements may be made to the above embodiments as appropriate without departing from the spirit and scope of the invention. The scope of the present invention should be construed based on the claims, and should be understood to include equivalents thereof.

The human resource development support system, cooperation support system, method, and computer program according to the present disclosure can increase opportunities for trainees to improve their skills and acquire knowledge necessary for their work without depending on time and place. It is suitable for use as a means of human resource development related to business such as insurance-related business in companies.

NW: communication network, T1: communication terminal, Te: trainee, T2: communication terminal, Tr: trainer, C1, C2: client, 1: cooperation support system, 10: online conference system, 11: virtual conference room, 12: personnel Training support system, 20: communication interface, 21: gateway server (GW server), 22: front system, 23: back system, 31: support control unit, 35: dialogue processing unit, 36: persona database (persona DB), 37 : scene database (scene DB), 40: speech analysis unit, 41: speech recognition unit, 42: natural language processing unit, 44: evaluation generation unit, 45: evaluation criteria database (evaluation criteria DB), 47: evaluation database (evaluation DB), 49: authentication database (DB), 60: information processing device, 61: processor, 62: RAM, 63: nonvolatile memory, 64: large capacity memory, 65: input/output interface circuit, 66: signal path.

Claims (33)

A human resource development support system that operates in cooperation with an online conference system configured to issue access destination information to a virtual conference room in response to a request by a trainer or trainee,
a support control unit that directly or indirectly acquires the access destination information from the online conference system;
a dialogue processing unit configured to access the virtual conference room using the access destination information and acquire utterance data of a dialogue between the trainer and the trainee participating in the virtual conference room;
a speech analysis unit configured to perform speech analysis by performing speech recognition on the speech data to generate speech text;
an evaluation generation unit configured to perform matching between the voice analysis result and pre-registered evaluation criteria information, and generate evaluation data indicating an evaluation result of the trainee according to the matching result; ,
and an evaluation database that stores the evaluation data. The human resource development support system according to claim 1,
further comprising a communication interface configured to communicate with an external communication terminal used by at least one of the trainer and the trainee via a communication network;
The support control unit has a function of receiving setting information from the external communication terminal via the communication interface,
The human resource development support system, wherein the setting information includes the access destination information. 3. The human resource development support system according to claim 1, wherein said evaluation generation unit executes said matching between one or more evaluation keywords included in said evaluation criteria information and the results of said speech analysis. A human resource development support system configured as follows. 4. A human resource development support system according to claim 3, wherein said evaluation criteria information has a combination of said plurality of evaluation keywords and at least one Boolean operator. 5. The human resource development support system according to any one of claims 1 to 4, wherein the speech analysis unit is configured to perform the speech analysis by applying natural language processing to the spoken text. human resource development support system. 7. The human resource development support system according to any one of claims 1 to 6, wherein the support control unit acquires evaluation data specified by a request from an external communication terminal from the evaluation database, A human resource development support system, further comprising a function of generating display data that enables evaluation results to be displayed based on the acquired evaluation data, and transmitting the display data. A cooperation support system comprising the human resource development support system according to any one of claims 1 to 6 and the online conference system. A method performed in a human resources development support system that operates in cooperation with an online conference system configured to issue access destination information to a virtual conference room in response to a request by a trainer or trainee,
accessing the virtual conference room using the access destination information obtained directly or indirectly from the online conference system;
obtaining speech data of an interaction between the trainer and the trainee participating in the virtual conference room;
performing speech analysis by subjecting the speech data to speech recognition to generate speech text;
performing matching between the result of the speech analysis and pre-registered evaluation criteria information;
generating evaluation data indicating evaluation results of the trainee according to the matching result;
and storing said rating data in a rating database. 9. The method of claim 8, wherein
further comprising a step of receiving setting information input via a communication interface from an external communication terminal used by at least one of the trainer and the trainee;
The method, wherein the setting information includes access destination information. 10. A method according to claim 8 or 9, wherein said matching is performed between one or more evaluation keywords contained in said evaluation criteria information and results of said speech analysis. 11. The human resource development support system according to claim 10, wherein said evaluation criteria information has a combination of said plurality of evaluation keywords and at least one Boolean operator. 12. The method of any one of claims 8-11, wherein the speech analysis is performed by subjecting the spoken text to natural language processing. 13. A method according to any one of claims 8 to 12,
a step of acquiring from the evaluation database the evaluation data specified by a request from an external communication terminal;
a step of generating display data that enables display of evaluation results based on the acquired evaluation data;
and transmitting said data for display. A computer program, read from a non-volatile memory and executed by one or more processors, for causing the one or more processors to perform the method according to any one of claims 8 to 13. A computer program characterized by comprising: Dialogue processing configured to open a virtual conference room in response to a request from a trainee, and acquire utterance data of the trainee while performing role-playing through automatic dialogue with the trainee participating in the virtual conference room. Department and
a speech analysis unit configured to perform speech analysis by performing speech recognition on the speech data to generate speech text;
an evaluation generation unit configured to perform matching between the voice analysis result and pre-registered evaluation criteria information, and generate evaluation data indicating an evaluation result of the trainee according to the matching result; ,
an evaluation database that stores the evaluation data;
a support control unit configured to receive setting information from an external communication terminal used by the trainee via a communication interface;
a persona database including a plurality of persona data sets for representing a plurality of virtual person images as virtual trainers,
The setting information includes information for selecting one virtual human figure from among the plurality of virtual human figures as a virtual trainer,
The dialogue processing unit obtains from the persona database a persona data set for expressing the selected one virtual person image, and uses the obtained persona data set to communicate with the trainee. A human resource development support system configured to perform the role-playing as the virtual trainer in. 16. The human resource development support system according to claim 15, wherein the dialogue processing unit dynamically generates a model image imitating the character of the virtual trainer and supplies the model image to the virtual conference room. Human resource development support system. The human resource development support system according to claim 15 or 16,
further comprising a scene database containing a plurality of scene datasets for representing a plurality of hypothetical scenes;
the setting information includes information for selecting one assumed scene from the plurality of assumed scenes;
The dialogue processing unit acquires a scene data set representing the selected assumed scene from the scene database, and uses the acquired scene data set to perform the A personnel training support system configured to perform role-playing. 18. The human resource development support system according to claim 17, wherein the support control unit selects, from among the plurality of virtual human images, candidate virtual human images that match the selected one assumed scene. is selected, and a setting screen prompting selection of one virtual person image from the candidates is displayed on the external communication terminal. 19. The human resource development support system according to any one of claims 15 to 18, wherein said evaluation generating unit generates a combination of one or more evaluation keywords included in said evaluation criteria information and said speech analysis result. A human resource development support system configured to perform the matching between. The human resource development support system according to claim 19,
The evaluation criteria information has one or more evaluation keywords set for each of the plurality of dialogue sections,
The human resource development support system, wherein the evaluation generation unit is configured to perform the matching to generate the evaluation data for each of the dialogue sections. 21. A human resource development support system according to claim 19 or 20, wherein said evaluation criteria information has a combination of said plurality of evaluation keywords and at least one Boolean operator. 22. The human resource development support system according to any one of claims 15 to 21, wherein said support control unit acquires from said evaluation database evaluation data specified by a request from an external communication terminal, A human resource development support system, further comprising a function of generating display data that enables evaluation results to be displayed based on the acquired evaluation data, and transmitting the display data. 23. The human resource development support system according to any one of claims 15 to 22, wherein the speech analysis unit is configured to perform the speech analysis by applying natural language processing to the spoken text. human resource development support system. A method executed in a human resource development support system configured to open a virtual conference room in response to a request from a trainee and perform role-playing through automatic dialogue with the trainee participating in the virtual conference room. hand,
a step of receiving setting information including information for selecting one virtual person image from among a plurality of virtual person images as a virtual trainer from an external communication terminal used by the trainee via a communication interface;
acquiring a persona data set for representing the selected one virtual persona from a persona database containing a plurality of persona data sets for respectively representing the plurality of virtual persona;
performing the role-playing as the virtual trainer with the trainee using the acquired persona dataset;
obtaining speech data of the trainee as the role-playing is performed;
performing speech analysis by subjecting the speech data to speech recognition to generate speech text;
performing matching between the result of the speech analysis and pre-registered evaluation criteria information;
generating evaluation data indicating evaluation results of the trainee according to the matching result;
and storing said rating data in a rating database. 25. The method of claim 24, further comprising dynamically generating a model image simulating the virtual trainer character and providing the model image to the virtual conference room. 26. The method of claim 24 or 25, wherein
further comprising the step of acquiring a scene data set for representing the one assumed scene selected by the setting information from a scene database containing a plurality of scene data sets for representing a plurality of assumed scenes;
The method, wherein the role-playing is performed based on the selected one hypothetical scene using the acquired scene data set. 27. The method of claim 26, wherein
selecting at least one candidate virtual human figure that matches the selected assumed scene from among the plurality of virtual human figures;
and causing the external communication terminal to display a setting screen prompting selection of one virtual person image from the candidates. 28. The method according to any one of claims 24 to 27, wherein said matching is performed between one or more evaluation keywords included in said evaluation criteria information and results of said speech analysis. ,Method. 29. The method of claim 28, wherein
further comprising generating the evaluation data for each interaction section of a plurality of interaction sections;
The method, wherein the evaluation criteria information has one or more evaluation keywords set for each of the plurality of dialogue sections. 30. The method of claim 28 or 29, wherein said evaluation criteria information comprises a combination of said plurality of evaluation keywords and at least one Boolean operator. 31. The method of any one of claims 24-30,
a step of acquiring from the evaluation database the evaluation data specified by a request from an external communication terminal;
a step of generating display data that enables display of evaluation results based on the acquired evaluation data;
and transmitting said data for display. 32. The method of any one of claims 24-31, wherein the speech analysis is performed by subjecting the spoken text to natural language processing. A computer program, read from a non-volatile memory and executed by one or more processors, for causing the processor or processors to perform the method of any one of claims 24 to 32. A computer program characterized by comprising:

Images