Device, method and program for conducting online questions and answers to allow company to recruit applicant

Abstract

To enhance the precision of obtaining useful information for allowing a company to evaluate an applicant in questions and answers conducted online for recruitment.SOLUTION: First, a device 100 transmits a first message to an applicant terminal 110 (S201), and receives a first reply from the terminal 110 (S202). The device 100 requests a first AI model to generate a second message in response to the first reply (S203). The first AI model receives the request and generates the second message (S204). The device 100 may also request a second AI model to generate a nod message based on the first reply (S205 to S208). When the device 100 receives the second message (S209), it transmits the second message to the terminal 110 (S210), and receives a second reply to the second message from the terminal 110 (S211).SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to an apparatus, a method, and a program therefor for a company to conduct a question-and-answer session for hiring applicants online.

Background Art

[0002] When a company hires personnel, in many cases, an interview is conducted after document screening. Through the interview, it is possible to grasp the qualifications, character, etc. of the applicant that cannot be known from the application documents.

[0003] Although the interview is useful in this way, the burden of schedule adjustment required between the company's hiring staff and the applicant is large, and attempts have been made to automate the company's interview response online using software.

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the recruitment interview conducted online has a stereotyped question-and-answer session, and it is not always sufficient for a company to select an applicant, and improvement is required.

[0005] The present invention has been made in view of such points, and the problem is to increase the accuracy of obtaining useful information for a company to select an applicant in an apparatus, a method, or a program therefor for conducting a question-and-answer session for hiring an applicant online.

Means for Solving the Problems

[0006] To achieve such an object, a first aspect of the present invention is a method for a company to conduct a question-and-answer session for the employment of applicants online, comprising: sending a first message to an applicant terminal used by the applicant; receiving a first reply to the first message from the applicant terminal; making a first request to a first AI model to generate a second message in response to the first reply, wherein the first AI model is a generation AI model; sending the second message generated by the first AI model to the applicant terminal; and receiving a second reply to the second message from the applicant terminal.

[0007] Further, a second aspect of the present invention is the method according to the first aspect, wherein when the first reply does not meet the conditions defined by the company for the first question topic to which the first message belongs, the second message is a message for asking additional questions about the first question topic.

[0008] Further, a third aspect of the present invention is the method according to the first or second aspect, wherein the first request includes setting conditions defined by the company for the first question topic to which the first message belongs.

[0009] Further, a fourth aspect of the present invention is the method according to the third aspect, wherein the setting includes a description by one or more sentences.

[0010] Further, a fifth aspect of the present invention is the method according to any one of the second to fourth aspects, wherein the first question topic is any one of a message prompting the applicant to introduce himself / herself, a message asking about the applicant's work experience, a message asking about the reason for job-hopping, and a message asking about employment conditions.

[0011] Further, a sixth aspect of the present invention is the method according to any one of the first to fifth aspects, further comprising sending a feedback message including a response to the first reply to the applicant terminal before sending the second message.

[0012] Further, a seventh aspect of the present invention is the method of the sixth aspect, wherein the reaction message is a message indicating a reaction to the first reply.

[0013] Further, an eighth aspect of the present invention is the method of the sixth or seventh aspect, wherein the step of transmitting the reaction message includes making a second request to a second AI model to request generation of the reaction message, receiving the reaction message generated by the second AI model, and transmitting the reaction message to the applicant terminal.

[0014] Further, a ninth aspect of the present invention is the method of the eighth aspect, wherein the second AI model is a generation AI model.

[0015] Further, a tenth aspect of the present invention is the method of the eighth or ninth aspect, wherein the second AI model is different from the first AI model.

[0016] Further, an eleventh aspect of the present invention is the method of any one of the eighth to tenth aspects, wherein the second AI model is provided on the same platform as the first AI model.

[0017] Further, a twelfth aspect of the present invention is the method of any one of the first to eleventh aspects, wherein the applicant terminal is a smartphone.

[0018] Further, a thirteenth aspect of the present invention is the method of any one of the first to twelfth aspects, further including storing, as a video including the applicant's face, a process until the first message is output as audio at the applicant terminal and the second reply is input as audio to the applicant terminal.

[0019] Further, a fourteenth aspect of the present invention is the method of the thirteenth aspect, further including making a third request to a third AI model to request analysis of the video.

[0020] Moreover, the 15th aspect of the present invention is the method according to any one of the 1st to 14th aspects, further including the step of making a fourth request to a fourth AI model to request generation of summaries of the first reply and the second reply.

[0021] Moreover, the 16th aspect of the present invention is the method according to any one of the 1st to 15th aspects, further including the step of obtaining applicant information regarding the applicant.

[0022] Moreover, the 17th aspect of the present invention is the method according to the 16th aspect, wherein the first request includes at least a part of the applicant information.

[0023] Moreover, the 18th aspect of the present invention is the method according to the 16th aspect, further including the step of making a fifth request to a fifth AI model to request determination of whether the second reply matches the applicant information.

[0024] Moreover, the 19th aspect of the present invention is a program for causing a computer to execute a method for a company to conduct an online question-and-answer session for applicant recruitment. The method includes the steps of sending a first message to an applicant terminal used by the applicant, receiving a first reply to the first message from the applicant terminal, making a first request to a first AI model to request generation of a second message according to the first reply, wherein the first AI model is a generation AI model, sending the second message generated by the first AI model to the applicant terminal, and receiving a second reply to the second message from the applicant terminal.

[0025] Further, a 20th aspect of the present invention is an apparatus for a company to conduct a question-and-answer session for applicant recruitment online, which transmits a first message to an applicant terminal used by the applicant, receives a first reply to the first message from the applicant terminal, and makes a first request to a first AI model, which is a generation AI model, to generate a second message according to the first reply, transmits the second message generated by the first AI model to the applicant terminal, and receives a second reply to the second message from the applicant terminal.

[0026] Further, a 21st aspect of the present invention is a method for a company to conduct a question-and-answer session for applicant recruitment online, which includes steps where an application executed on the applicant terminal provides a first message to the applicant terminal used by the applicant, the application obtains a first reply to the first message, the application uses a first AI model to generate a second message according to the first reply, where the first AI model is a generation AI model, the application provides the second message generated by the first AI model to the applicant terminal, and the application obtains a second reply to the second message.

[0027] Further, a 22nd aspect of the present invention is an application for causing an applicant terminal used by an applicant to execute a method for a company to conduct a question-and-answer session for applicant recruitment online, where the method includes steps of providing a first message to the applicant terminal, obtaining a first reply to the first message, using a first AI model to generate a second message according to the first reply, where the first AI model is a generation AI model, providing the second message generated by the first AI model to the applicant terminal, and receiving a second reply to the second message.

[0028] Further, a 23rd aspect of the present invention is a terminal for a company to conduct a question-and-answer session for applicant recruitment online, which obtains a first reply from the applicant to a first message provided from an application executed on the terminal, provides the first reply to the application, and obtains a second reply from the applicant to a second message generated by the application in response to the first reply using a first AI model that is a generative AI model.

Advantages of the Invention

[0029] According to one aspect of the present invention, by generating a second message by a generative AI model in response to a first reply from an applicant to a first message, the accuracy of obtaining useful information for screening the applicant is enhanced.

Brief Description of the Drawings

[0030]

Figure 1

Figure 2

Figure 3

Figure 4

Modes for Carrying Out the Invention

[0031] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0032] (First Embodiment) FIG. 1 shows a system according to a first embodiment of the present invention. The apparatus 100 communicates with an applicant terminal 110 used by an applicant via an IP network such as the Internet, and further communicates with a company terminal 120 used by a company that conducts recruitment as needed. In addition, the apparatus 100 communicates with a platform 130 that provides an AI model via an IP network such as the Internet. The AI model will be exemplarily described as being provided by a platform 130 capable of communicating with the apparatus 100, but it is also possible to execute an application for providing the AI model on the apparatus 100 so that the apparatus 100 provides the AI model.

[0033] The apparatus 100 includes a communication unit 101 such as a communication interface, a processing unit 102 such as a processor or a CPU, and a storage unit 103 including a storage device or a storage medium such as a memory or a hard disk, and can be configured by executing a program for performing each process or each operation in the processing unit 102. The apparatus 100 may include one or more apparatuses, computers or servers. Further, the program may include one or more programs, and can be recorded on a computer-readable storage medium to form a non-transitory program product. The program can be stored in a storage device or a storage medium such as a database 104 accessible from the storage unit 103 or the apparatus 100 via an IP network, and instructions included in the program can be executed by at least one processor of the processing unit 102. The data described below as being stored in the storage unit 103 may be stored in a storage device or a storage medium such as the database 104, and vice versa.

[0034] The applicant terminal 110 may be a mobile terminal such as a smartphone, a tablet, or a notebook PC, or may be other information terminals. Since the applicant can conduct a question-and-answer session without the need for scheduling with the company, if the applicant terminal 110 is a smartphone, it is preferable because the restrictions on time and place are also relaxed. The applicant terminal 110 can also be configured in the same way as the apparatus 100.

[0035] First, the device 100 sends a first message to the applicant terminal 110 (S201). The first message belongs to some question topic such as a message prompting the applicant to introduce themselves, a message asking about the applicant's work experience, a message asking about the reason for job change, a message asking about employment conditions, etc.

[0036] Next, the device 100 receives a first reply to the first message from the applicant terminal 110 (S202). For example, when the first message is "Please tell me about your work experience after your highest education.", an example of the first reply received could be "My first workplace was XX Pharmaceutical, a large pharmaceutical company. I worked there for one year in the marketing department of pharmaceuticals, responsible for tasks such as market research, product planning, and sales promotion."

[0037] The device 100 that has received the first reply requests (also referred to as the "first request") an AI model (also referred to as the "first AI model") to generate a second message according to the first reply (S203). The first AI model receives the first request and generates a second message (S204).

[0038] FIG. 3 shows an example of a part of the code included in the first request to the first AI model according to the first embodiment of the present invention. The example in FIG. 3 is a part of the code written in the programming language Python. By executing this code, the OpenAI API can be called to cause the first AI model provided on the platform 130 to generate a second message. The OpenAI API is an example, and other APIs may also be used. The code for making the first request is stored in the storage unit 103, and the device 100 can obtain this and execute the code obtained by setting values for one or more variables included in the code. Examples of variables include the first reply, conditions defined for the first question topic, etc.

[0039] The prompt shown in FIG. 3 requires that if the first reply meets the condition, a question for the next question topic is generated, and if it does not meet the condition, an additional question about the current question topic is generated. The additional question may be a question for asking for details of the first reply.

[0040] As an example of the second message, there is "So, it means all your work experience at XX Pharmaceutical. Is there any gap period after leaving XX Pharmaceutical? Or did you start job hunting immediately?" Since it cannot be said that the first reply meets the predetermined condition, this is for asking additional questions about the topic of work history.

[0041] The specification of a plurality of question topics, the order of the plurality of question topics, and further the conditions defined for at least a part of the plurality of question topics can be transmitted from the enterprise terminal 120 to the apparatus 100 and stored in the storage unit 103 of the apparatus 100 by the enterprise that conducts the recruitment. Since the conditions regarding the question topics only need to be defined for the question topics for which a detailed answer is desired, they are not necessarily required for all question topics. Also, the conditions preferably are not simply those including one or more terms, but are at least partially determined by one or more sentences. Incidentally, just in case, in this specification, a "sentence" is not something that has a subject and a predicate as essential components although it is longer than a single term.

[0042] The specification of conditions can be done by sending the input screen display information for condition input as, for example, an HTML-formatted file, and performing the operation from the input screen displayed on the display screen of the enterprise terminal 120 on the web browser, or by operating the application (hereinafter also referred to as "app") installed on the enterprise terminal 120 and performing the operation from the input screen displayed on the app. Note that the "input screen" can adopt various forms such as a web page, a modal window, a pop-up window, etc. when displayed on a web browser, and can be a single screen of the app when displayed on the app. In any case, as long as it is a screen including an area having an input field for inputting conditions, it corresponds to the input screen.

[0043] Since it takes a certain amount of time to generate the second message, in some cases, the applicant may receive an unnatural impression. Therefore, if necessary, before the device 100 transmits the second message generated by the first AI model to the applicant terminal 110, the device 100 may transmit a response message including a response to the first reply to the applicant terminal 110.

[0044] Specifically, after the device 100 makes the first request, the device 100 can request (also referred to as the "second request") an AI model (also referred to as the "second AI model") to generate a response message based on the first reply (S205). The second AI model receives the second request and generates a response message (S206). Here, an example in which the device 100 makes the second request after making the first request has been described, but the order may be reversed.

[0045] An example of the response message is a message such as "I see. You have had a wide range of experiences at 〇△ Pharmaceutical." It can be something that shows reactions such as empathy, interest, agreement, etc. to the first reply by the applicant.

[0046] FIG. 4 shows an example of a part of the code included in the second request for the second AI model according to the first embodiment of the present invention. The example in FIG. 4 is a part of the code written in the programming language Python, similar to the example in FIG. 3. By executing the code, the OpenAI API can be called to generate a response message for the second AI model provided on the platform 130. The OpenAI API is an example, and other APIs may be used. The generation of the response message may not be by generation by the generative AI model. The code for making the second request may be stored in the storage unit 103, and the device 100 may acquire this, set values for one or more variables included in the code, and execute the resulting code. Examples of variables include the first message, the first reply, and the like.

[0047] When the device 100 receives the response message (S207), it transmits this to the applicant terminal 110 (S208). Then, when it receives the second message (S209), it transmits this to the applicant terminal 110 (S210) and receives a second reply to the second message from the applicant terminal 110 (S211).

[0048] As an example of the second reply, a message such as "I started job hunting immediately after leaving my previous company. At the same time, I emphasized self - investment and continued to study programming and data analysis. Even now, I am constantly trying to absorb new knowledge every day." can be cited. If the first AI model determines that the conditions defined for the first question topic are satisfied by this reply, a message regarding the reason for job transfer, which is the next question topic, may be generated. For example, a message such as "Then, could you please tell me the specific reasons for deciding to change jobs this time?" may be generated.

[0049] In FIG. 2, the first AI model and the second AI model are distinguished, but they may also be the same generative AI model. As used herein, the term "AI model" refers to a machine learning model that has been trained to be able to predict an output for an input, and the term "generative AI model" refers to a large language model (LLM) that has been trained using text data to be able to generate an output for an input. As the generative AI model, an LLM that particularly applies a transformer architecture is preferable, but it is assumed that the name of the architecture will change with the progress of technology. Therefore, as used herein, the term "transformer architecture" includes an architecture that uses one or more features of the transformer architecture or an improvement thereof. Whether the "generative AI models" in this specification are the same or not is determined by whether the types of generative AI models specified by the user are the same or not. For example, in the example of FIG. 3, if the value of the variable "model" is the same, it is expressed as being the same as a generative AI model. When the first AI model and the second AI model are not the same, they may be provided on the same platform 130. In the example of FIG. 2, where a second request is made after the first request, if the first AI model and the second AI model are the same generative AI model or generative AI models provided on the same platform, these requests may be made by a single call to the API.

[0050] Thus, in response to the first reply from the applicant terminal 110 to the first message, by generating the second message using the first generative AI model, the accuracy of obtaining useful information for screening the applicant is enhanced.

[0051] In the applicant terminal 110, the first message and the second message may be output as audio, and the applicant may be able to input the first reply and the second reply as audio in the applicant terminal 110. The process from when the first message is output as audio in the applicant terminal 110 until the second reply is input as audio to the applicant terminal 110 may be stored in the storage unit 103 as a video including the face of the applicant. The video may further record at least one of before the first message is output and after the second reply is input. By saving as a video, the company can quickly conduct a question-and-answer session with the applicant without scheduling adjustments and shorten the time required for the recruitment process. Furthermore, by saving as a video, it becomes easier for the company to evaluate the applicant by multiple responsible persons.

[0052] In addition, the device 100 may analyze the video and extract information other than the first and second replies regarding the applicant. The extraction may be performed by the device 100 itself, or a request (also referred to as the "third request") may be made to an AI model (also referred to as the "third AI model"). The third AI model may be the same as or provided on the same platform as at least one of the first AI model and the second AI model. Also, similar to the first request, the third request can be made by preparing the code in advance and having the device 100 set one or more necessary variables and execute it.

[0053] In addition, the apparatus 100 may receive a resume, a work history document, and other documents from an applicant in a paper medium or an electronic medium, obtain applicant information regarding the applicant, and store it in the storage unit 103. If such applicant information already exists, it is possible to include at least a part of the applicant information in the first request and make the second message reflect the applicant information. Alternatively, by determining whether the second reply to the second message matches the applicant information, it is possible to evaluate the reliability of the applicant's reply or evaluate the sufficiency of the conditions defined for the question topic to which the second message belongs. Such determination may be made by the apparatus 100 itself or by making a request (also referred to as the "fourth request") to an AI model (also referred to as the "fourth AI model"). The fourth AI model may be the same as or provided on the same platform as at least one of the first AI model to the third AI model. Also, similar to the first request, the fourth request can be made by preparing code in advance and having the apparatus 100 set one or more necessary variables and execute it.

[0054] In addition, it is possible to request an AI model (also referred to as the "fifth AI model") to generate a summary of the first reply and the second reply by the applicant so that the information required by the person in charge of the company can be grasped in a short time. The fifth AI model may be the same as or provided on the same platform as at least one of the first AI model to the fourth AI model. Also, similar to the first request, the fifth request can be made by preparing code in advance and having the apparatus 100 set one or more necessary variables and execute it.

[0055] (Second Embodiment) In the first embodiment, the AI model provided by the platform 130 was used. However, in the second embodiment, the AI model is provided by executing an application or program on the applicant terminal 110. In this case, the first message described as being sent to the applicant terminal 110 may be provided to the applicant terminal 110 by the application, and the same applies to the second message. Also, the first reply described as being received from the applicant terminal 110 may be obtained by the application on the applicant terminal 110, and the same applies to the second reply.

[0056] When viewed from the perspective of the applicant terminal 110, in the second embodiment, the applicant terminal 110 obtains a first reply by the applicant using the applicant terminal 110 to the first message provided by the application executed on the applicant terminal 110, and provides the first reply to the application. The application uses a first AI model that is a generative AI model to generate a second message in response to the first reply, and the applicant terminal 110 receives the provision thereof. Then, the applicant terminal 110 obtains a second reply by the applicant to the second message.

[0057] Note that in the above-described embodiments, if there is no description of "only" such as "based only on", "in response only to", "in only the case of", "referring only to", it is assumed that additional information may be considered in this specification. Also, as an example, note that the description "do b in the case of a" does not necessarily mean "always do b in the case of a" or "do b immediately after a" except when explicitly stated. Also, the description "each a that constitutes A" does not necessarily mean that A is constituted by a plurality of components, and includes the case where the component is singular.

[0058] Also, as a precaution, even if there is an aspect of performing operations different from the operations described in this specification in any method, program, terminal, device, server, or system (hereinafter referred to as "method, etc."), each aspect of the present invention is directed to the same operations as any of the operations described in this specification, and it is noted that the existence of operations different from the operations described in this specification does not exclude the method, etc. from the scope of each aspect of the present invention.

[0059] Also, in the above description, when referring to a plurality of AI models, for example, a second request for generating a response message may be made before the first request for generating a second message. In such a case, from the viewpoint of readability, the second AI model may be referred to as "AI model k", and the first AI model may be referred to as "AI model l". And when making a request to the second AI model after making a request to the first AI model, the first AI model may be referred to as "AI model k" and the second AI model may be referred to as "AI model m". In other cases, appropriate substitutions may be made as needed.

Explanation of Reference Numerals

[0060] 100 Device 101 Communication Unit 102 Processing Unit 103 Storage Unit 104 Database 110 Applicant Terminal 120 Company Terminal 130 Platform

Claims

1. A method for a company to conduct online question and answer sessions for recruiting applicants, comprising: sending a first message to an applicant terminal used by the applicant; receiving a first reply to the first message from the applicant terminal; making a first request to a first AI model to generate a second message in response to the first reply, the first AI model being a generative AI model; Sending the second message generated by the first AI model to the applicant terminal; receiving a second reply to the second message from the applicant terminal; Includes.

2. 2. The method of claim 1 , If the first reply does not satisfy the conditions set by the company for the first question topic to which the first message belongs, the second message is a message that asks a follow-up question about the first question topic.

3. 3. The method according to claim 1 or 2, The first request includes a set of conditions that the business defines for a first question topic to which the first message belongs.

4. 4. The method of claim 3, The configuration includes a description in one or more sentences.

5. 5. A method according to any one of claims 2 to 4, comprising: The first question topic is any one of a message encouraging the applicant to introduce himself / herself, a message asking about the applicant's career history, a message asking about the reason for changing jobs, and a message asking about working conditions.

6. 6. A method according to any one of claims 1 to 5, comprising: The method further includes a step of transmitting, before transmitting the second message, to the applicant terminal, a backchannel message including a backchannel in response to the first reply.

7. 7. The method of claim 6, The interjection message is a message indicating a reaction to the first reply.

8. 8. The method according to claim 6 or 7, The step of transmitting the backchannel message includes: making a second request to a second AI model to generate the backchannel message; receiving the backchannel message generated by the second AI model; sending the response message to the applicant terminal; Includes.

9. 9. The method of claim 8, The second AI model is a generative AI model.

10. 10. The method according to claim 8 or 9, The second AI model is different from the first AI model.

11. 11. A method according to any one of claims 8 to 10, comprising: The second AI model is provided on the same platform as the first AI model.

12. 12. A method according to any one of claims 1 to 11, comprising: The applicant terminal is a smartphone.

13. 13. A method according to any one of claims 1 to 12, comprising: The method further includes a step of storing, as a video including the applicant's face, the process from when the first message is output by voice at the applicant terminal to when the second reply is input by voice to the applicant terminal.

14. 14. The method of claim 13, The method further includes making a third request to a third AI model to analyze the video.

15. 15. A method according to any one of claims 1 to 14, comprising: The method further includes making a fourth request to a fourth AI model to generate a summary of the first reply and the second reply.

16. 16. A method according to any one of claims 1 to 15, comprising: The method further includes obtaining applicant information regarding the applicant.

17. 17. The method of claim 16, The first request includes at least a portion of the applicant information.

18. 17. The method of claim 16, The method further includes making a fifth request to a fifth AI model to determine whether the second reply matches the applicant information.

19. A program for causing a computer to execute a method for a company to conduct an online question and answer session for recruiting applicants, the method comprising: sending a first message to an applicant terminal used by the applicant; receiving a first reply to the first message from the applicant terminal; making a first request to a first AI model to generate a second message in response to the first reply, the first AI model being a generative AI model; Sending the second message generated by the first AI model to the applicant terminal; receiving a second reply to the second message from the applicant terminal; Includes.

20. A device for a company to conduct online question and answer sessions for hiring applicants, sending a first message to an applicant terminal used by the applicant, and receiving a first reply to the first message from the applicant terminal; making a first request to a first AI model, which is a generative AI model, to generate a second message in response to the first reply; The second message generated by the first AI model is sent to the applicant terminal, and a second reply to the second message is received from the applicant terminal.

21. A method for a company to conduct online question and answer sessions for recruiting applicants, comprising: providing a first message to an applicant terminal used by the applicant by an application executed on the applicant terminal; the application obtaining a first reply to the first message; the application generating a second message in response to the first reply using a first AI model, the first AI model being a generative AI model; The application provides the applicant terminal with the second message generated by the first AI model; the application obtaining a second reply to the second message; Includes.

22. An application for causing an applicant terminal used by an applicant to execute a method for a company to conduct online question and answer sessions for the employment of an applicant, the method comprising: providing a first message to the applicant terminal; obtaining a first reply to the first message; generating a second message in response to the first reply using a first AI model, the first AI model being a generative AI model; Providing the second message generated by the first AI model to the applicant terminal; receiving a second reply to the second message; Includes.

23. A terminal for a company to conduct online question and answer sessions for hiring applicants, obtaining a first reply by the applicant to a first message provided by an application executing on the terminal; providing the first reply to the application; The application obtains a second reply from the applicant to a second message generated in response to the first reply using a first AI model, which is a generative AI model.

Images