Recruitment support system, recruitment support method, and recruitment support program

The recruitment support system addresses inefficiencies by matching job seekers with recruiters using AI-controlled units and databases, ensuring targeted interviews with promising candidates, thus improving the hiring process efficiency.

JP2026182639APending Publication Date: 2026-11-19PERSOL CAREER CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2025079026
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-05-09
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Existing hiring processes burden job seekers with selecting from numerous employers and employers with evaluating multiple unpromising candidates, leading to inefficiencies in the recruitment process.

Method used

A recruitment support system utilizing job seeker and recruiter terminals, databases, and AI-controlled units to match job seekers with recruiters based on preferences and evaluation criteria, facilitating targeted interviews.

Benefits of technology

Reduces the burden on both job seekers and employers by providing accurate recruitment proposals and interviews with promising candidates, enhancing the efficiency of the hiring process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026182639000001_ABST
    Figure 2026182639000001_ABST
Patent Text Reader

Abstract

We provide recruitment support systems, recruitment support methods, and recruitment support programs that extract appropriate information from job seekers and lead to highly effective recruitment interviews. [Solution] The recruitment support system 10a comprises a job seeker terminal 2, a recruiter terminal 3 used by recruiters, a communication-enabled system comprising a job seeker database DB1, a recruiter database DB2 storing basic information and evaluation criteria of recruiters, a proposal unit 14 that suggests recruiters who match the job seeker's preferences, a selection preference receiving unit that receives the job seeker's preference for selection by the proposed recruiters from the job seeker terminal 2, an evaluation unit 15 that evaluates whether the job seeker meets the selection criteria of the recruiter they wish to be selected by, an evaluation transmission unit that transmits the evaluation of the job seeker to the job seeker terminal 2, and a communication processing unit 16 that receives the job seeker's intention to interview with a recruiter who has been determined to meet the selection criteria from the job seeker terminal 2 and transmits the job seeker's information along with the job seeker's request for an interview to the recruiter terminal 3.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to an employment support system, an employment support method, and an employment support program.

Background Art

[0002] In hiring human resources, it is common practice for interviews to be conducted between the hiring company and job seekers. In this regard, there is a demand for a technology that can reduce the burden on both the hiring company and job seekers, improve the accuracy of hiring through interviews, and streamline human resource hiring activities.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Patent Document 1 proposes an interview apparatus comprising: an input unit that inputs, as the input text, first text representing a respondent's answer and an instruction to ask a question regarding the answer, to a text generation model that generates an output text according to an input text; and an output unit that outputs, to the respondent, the question as the output text generated by the text generation model.

[0005] However, even if the efficiency of the interview itself is improved, from the job seeker's perspective, the burden of having to select an employer from a huge number of hiring entities remains, and for employers, the burden of having to consider multiple candidates with no promising potential is not insignificant.

[0006] An object of the present invention is to provide an employment support system that can elicit appropriate information from job seekers and lead to highly accurate employment interviews.

Means for Solving the Problems

[0007] To achieve the above objective, a recruitment support system according to one aspect of the present invention comprises: a job seeker terminal used by a job seeker; a recruiter terminal used by a recruiter; a job seeker information storage unit configured to communicate and storing information about the job seeker; a basic information storage unit storing basic information about the recruiter; an evaluation criteria storage unit storing evaluation criteria for the recruiter; a proposal unit that refers to the basic information storage unit and proposes a recruiter who matches the job seeker's preferences; a selection preference receiving unit that receives the job seeker's preference for selection by the proposed recruiter from the job seeker terminal; an evaluation unit that evaluates whether the job seeker meets the selection criteria of the recruiter they wish to be selected by; an evaluation transmission unit that transmits the evaluation of the job seeker to the job seeker terminal; an interview preference receiving unit that receives the job seeker's intention to interview with the recruiter who has been determined to meet the selection criteria from the job seeker terminal; and an interview preference transmission unit that transmits an interview request with the job seeker to the recruiter terminal along with the job seeker's information indicating their intention to interview.

[0008] The proposal unit is comprised of a proposal device that conducts interviews with the job seeker and outputs companies to propose to the job seeker, wherein the content of the interview includes at least multiple answers from the job seeker to questions, and the proposal device comprises an AI control unit that transmits instructions to an artificial intelligence unit, and a communication processing unit that acquires information on recruiters to propose to the job seeker from the artificial intelligence unit, wherein the AI ​​control unit transmits a proposal instruction to the artificial intelligence unit that, based on the multiple answers, outputs recruiters whose basic information matches the job seeker's preferences as recruiters to propose to the job seeker.

[0009] The evaluation unit comprises an evaluation device that performs a session with the job seeker and outputs an evaluation of the job seeker, the contents of which include at least multiple answers from the job seeker to questions, the evaluation device comprises an AI control unit that transmits instructions to an artificial intelligence unit, and a communication processing unit that obtains the evaluation results of the job seeker from the artificial intelligence unit, and the AI ​​control unit may transmit an evaluation instruction to the artificial intelligence unit that outputs an evaluation of the job seeker based on the multiple answers and the selection criteria.

[0010] A recruitment support method according to another aspect of the present invention may include a job seeker terminal used by a job seeker, a recruiter terminal used by a recruiter, and a computer having a job seeker information storage unit configured to communicate and storing information about the job seeker, a basic information storage unit storing basic information about the recruiter, and an evaluation criteria storage unit storing evaluation criteria for the recruiter, which may perform the following steps by referring to the basic information storage unit and proposing a recruiter who matches the job seeker's preferences; a selection preference receiving step by receiving the job seeker terminal's preference for selection by the proposed recruiter; an evaluation step by evaluating whether the job seeker meets the selection criteria of the recruiter they wish to be selected by; an evaluation transmission step by transmitting the evaluation of the job seeker to the job seeker terminal; an interview preference receiving step by receiving the job seeker terminal's preference for an interview with the recruiter who has been determined to meet the selection criteria; and an interview preference transmission step by transmitting the job seeker information of the job seeker who has indicated their preference for an interview, along with an interview preference request to the recruiter terminal.

[0011] A recruitment support program according to another aspect of the present invention causes a computer having a job seeker terminal used by a job seeker, a recruiter terminal used by a recruiter, a job seeker information storage unit configured to communicate and storing information about the job seeker, a basic information storage unit storing basic information about the recruiter, and an evaluation criteria storage unit storing evaluation criteria for the recruiter, to execute a proposal command that refers to the basic information storage unit and proposes a recruiter that matches the job seeker's preferences, a selection preference reception command that receives a preference from the job seeker terminal for selection by the proposed recruiter, an evaluation command that evaluates whether the job seeker meets the selection criteria of the recruiter they wish to be selected by, an evaluation transmission command that transmits the evaluation of the job seeker to the job seeker terminal, an interview preference reception command that receives an interview preference from the job seeker terminal with the recruiter who has been determined to meet the selection criteria, and an interview preference transmission command that transmits an interview preference request to the recruiter terminal along with the job seeker information of the job seeker who has indicated an interview preference.

[0012] Computer programs can be provided by storing them on various data-readable storage media, or by making them available for download via networks such as the Internet. [Effects of the Invention]

[0013] According to the present invention, it is possible to extract appropriate information from job seekers and lead to highly effective recruitment interviews. [Brief explanation of the drawing]

[0014] [Figure 1] This is a functional block diagram showing the functions of the recruitment support system according to Embodiment 1 of the present invention. [Figure 2] This is a sequence diagram showing the processing flow executed by the above-mentioned recruitment support system. [Figure 3] This is a functional block diagram showing the functions of the recruitment support system according to Embodiment 2 of the present invention. [Figure 4]This is a schematic diagram showing an overview of the behavior of the artificial intelligence unit group in the recruitment support system, illustrating (a) behavior during an interview and (b) behavior during a session. [Figure 5] This is a sequence diagram showing the flow of interview processing executed by the recruitment support system. [Figure 6] This is a sequence diagram showing the flow of session processing executed by the recruitment support system. [Mode for Carrying Out the Invention]

[0015] ●Overview Hereinafter, embodiments of the recruitment support system according to the present invention will be described with reference to the drawings. The recruitment support systems 10a and 10b shown in Figure 1 and Figure 3 propose recruiters matching the job seeker's preferences to job seekers, evaluate job seekers based on selection criteria preset by recruiters, and support recruitment by providing recruiters with information on job seekers who meet the selection criteria. Recruiters will only conduct recruitment interviews with job seekers who satisfy the selection criteria they set, and decide whether to hire the job seekers according to the interview results. [Examples]

[0016] ●Network Configuration As shown in Figure 1, the recruitment support system 10a according to Example 1 is configured to be able to communicate with a job seeker terminal 2 operated by a job seeker and a recruiter terminal 3 operated by a recruiter via a network NW, and supports the recruiter's recruitment of the job seeker by exchanging data with the job seeker terminal 2 and the recruiter terminal 3 to propose recruiters to the job seeker and provide the recruiter with the job seeker's information. This recruitment support system 10a is, for example, under the management of an agent who mediates between job seekers and recruiters to support both parties, and the agent provides a service for supporting the recruitment of job seekers via the recruitment support system 10a (hereinafter sometimes referred to as "this service").

[0017] ●Recruitment Support System 10a The recruitment support system 10a is composed of a recruitment support device 1a, a job seeker database DB1, a recruitment database DB2, and an evaluation criterion database DB3, which are configured to be communicable via an appropriate network NW.

[0018] The recruitment support device 1a may be configured by a hardware device, or part or all of its functions may be implemented by a cloud computer. Further, each component of the recruitment support device 1a may be implemented by an API (Application Programming Interface). Furthermore, the recruitment support device 1a may be configured by a plurality of hardware configurations. In this case, the plurality of hardware configurations may be connected by wire or wirelessly, and may transmit and receive information to and from each other.

[0019] The recruitment support device 1a mainly configures each functional block including a display control unit 11, an acquisition unit 12, a storage control unit 13, a proposal unit 14, an evaluation unit 15, and a communication processing unit 16, by means of a CPU (Central Processing Unit, which is an example of the processor in the claims), computer programs executed by the CPU, a RAM (Random Access Memory) and a ROM (Read Only Memory) that store the computer programs and predetermined data.

[0020] ●Job Seeker Database DB1 The job seeker database DB1 stores various types of information that are suitable for recruiters to obtain during interviews, such as name, age, educational background, and work history. This job seeker information may also be entered by the job seeker when they use this service. In addition, various types of information that arise incidentally when a job seeker receives this service may be registered as appropriate, such as the content of the hearings conducted when the proposal unit 14 (described later) proposes a recruiter, the content and evaluation of the sessions conducted when the evaluation unit 15 (described later) evaluates the job seeker, and the results of the interview between the job seeker and the recruiter. Furthermore, all of the job seeker information stored in the job seeker database DB1 may be obtained through hearings, in which case the job seeker does not need to enter the information in advance.

[0021] ● Recruitment Database DB2 The recruiter database DB2 stores, for example, basic information about recruiters and information related to selection criteria. Basic information includes various pieces of information necessary for job seekers to judge their preferences and conditions as a place of employment or career change, such as the recruiter's name, address, and business description, and recruiters can be proposed to job seekers based on this basic information. Information related to selection criteria is the minimum evaluation that recruiters require from job seekers in order to conduct an employment interview. Job seekers can request an employment interview with a recruiter if they have received an evaluation that meets the selection criteria set by each recruiter.

[0022] ●Evaluation criteria database DB3 The evaluation criteria database DB3 stores evaluation criteria that serve as standards for assessing the abilities of job applicants. These criteria consist, for example, of evaluation items related to abilities and corresponding sentences indicating the evaluation criteria for each score associated with those evaluation items. The evaluation criteria database DB3 may also store evaluation items related to the session described later. By referring to these evaluation criteria, it is possible to determine the applicant's score in the session described later and, based on the score, evaluate whether the applicant meets the score required by the employer, i.e., the selection criteria.

[0023] ●Display control unit 11 The display control unit 11, for example, performs processing to display information on the job seeker terminal 2. The display control unit 11 performs processing such as generating and sending an HTML (Hyper Text Markup Language) file, and displays a web page showing the hearing and session execution screens, which will be described later, on the job seeker terminal 2. The display control unit 11 may also perform processing such as generating and sending display data for the application used to utilize the recruitment support system 10a. This display control unit 11 may, for example, cause the job seeker terminal 2 to display an image or video of a virtual or physical agent conducting an interview or session on its screen.

[0024] ●Acquisition part 12 The acquisition unit 12 is a functional unit that acquires information via the communication processing unit 16 from terminals that are connected to the network NW in a communicable manner, such as the job seeker terminal 2 or the recruiter terminal 3.

[0025] ● Memory control unit 13 The memory control unit 13 is a functional unit that controls the appropriate storage devices and databases DB1, DB2, and DB3 provided by the adoption support device 1a, and performs data writing and reading.

[0026] ●Proposal Department 14 The proposal unit 14 refers to the recruit database DB2 and proposes recruits that match the job seeker's preferences. In the proposal process, for example, an agent conducts an interview with a job seeker to ascertain their preferences and inputs the interview content into the recruitment support device 1a. Based on the input interview content, the proposal unit 14 refers to the recruiter database DB2 to identify recruiters who match the job seeker's preferences and then transmits the information of the recruiters who match the job seeker's preferences to the job seeker terminal 2. Job seekers select candidates from those proposed as matching their preferences and then participate in a session. Alternatively, instead of job seekers selecting candidates, they may choose candidates to exclude from the proposed pool.

[0027] ●Evaluation Unit 15 The evaluation unit 15 refers to the recruit database DB2 and the evaluation criteria database DB3 to evaluate whether the job seeker meets the selection criteria of the recruit they wish to be selected for. In the evaluation process, for example, an agent conducts a session with a job seeker to gather information necessary for evaluating the job seeker, and inputs the session content into the recruitment support device 1a. Based on the input session content, the evaluation unit 15 refers to the evaluation criteria database DB3 to obtain a score, and then refers to the employer's selection criteria to evaluate whether the job seeker meets the employer's selection criteria. The job seeker's evaluation is transmitted to the job seeker terminal 2, and if the job seeker meets the employer's selection criteria, the job seeker can request an interview with the employer.

[0028] ●Communication processing unit 16 The communication processing unit 16 is a processing unit that enables the transmission and reception of data in accordance with a predetermined protocol with the job seeker terminal 2 and the recruiter terminal 3 via a network NW such as the Internet. For example, the communication processing unit 16 receives the selection preferences of the proposed recruiter and the willingness to interview with recruiters who have been determined to meet the evaluation criteria, as well as transmits the evaluation of the job seeker, between itself and the job seeker terminal 2. Also, for example, between itself and the recruiter terminal 3, it transmits the job seeker's information along with the job seeker's willingness to interview, as well as an interview request for that job seeker.

[0029] ● Job seeker terminal 2 The job seeker terminal 2 is a terminal operated by the job seeker, and is, for example, a smartphone, tablet, or personal computer. The job seeker terminal 2 mainly consists of a display unit 21, an operation reception unit 22, and a communication processing unit 23, comprising a CPU (Central Processing Unit), a computer program executed by the CPU, and RAM (Random Access Memory) and ROM (Read Only Memory) that store the computer program and predetermined data.

[0030] The display unit 21 is implemented by a display or the like for outputting data. For example, the display unit 21 displays the evaluation of job applicants and information about the candidates that job applicants wish to be considered for.

[0031] The operation reception unit 22 is implemented using a touch panel, keyboard, mouse, or microphone for data input. For example, during interviews or sessions, the operation reception unit 22 receives answers to questions from job seekers in text or voice.

[0032] The communication processing unit 23 is a processing unit that enables the transmission and reception of data with the adoption support device 1a via a network NW such as the Internet, in accordance with a predetermined protocol, and is implemented by an application or a web browser.

[0033] ●Employee terminal 3 The recruiter terminal 3 is a terminal operated by recruiters of the recruiting company, and is, for example, a smartphone, tablet, or personal computer. The job seeker terminal 2 mainly consists of a display unit 31, an operation reception unit 32, and a communication processing unit 33, comprising a CPU (Central Processing Unit), a computer program executed by the CPU, and RAM (Random Access Memory) and ROM (Read Only Memory) that store the computer program and predetermined data.

[0034] The display unit 31 is implemented by a display or the like for outputting data. For example, the display unit 31 displays evaluations and information of job applicants who are requesting an interview.

[0035] The operation reception unit 32 is implemented by a touch panel, keyboard, mouse, or microphone for inputting data.

[0036] The communication processing unit 33 is a processing unit that enables the transmission and reception of data with the adoption support device 1a in accordance with a predetermined protocol via a network NW such as the Internet, and is implemented by an application or a web browser.

[0037] ●Processing flow As shown in Figure 2, the flow of this service is shown along with the flow of processing performed by the recruitment support system 10a. Furthermore, at least the employer must register the necessary employer information in advance to receive this service. However, job seekers may also register the necessary job seeker information in advance to receive this service.

[0038] The agent interviews the job seeker to understand their preferences, such as their intention to find employment or change jobs, and the conditions they desire in a potential employer (S101). When the agent inputs the results of the interview into the recruitment support system 10a, the proposal unit 14, based on the inputted interview content, refers to the recruiter database DB2 to identify recruiters who match the job seeker's preferences, and then sends the information of the recruiters who match the job seeker's preferences to the job seeker terminal 2 as a recruiter proposal (S102). In addition, the recruiter proposal may suggest one or more recruiters, and if multiple recruiters are suggested, they may be presented as a list.

[0039] The job seeker selects any recruiter they wish to be considered based on the recruiter's proposal and sends their selection preference to the recruitment support system 10a via the job seeker terminal 2 (S103). The agent then conducts a session with the job seeker (S104). The session is a dialogue between the agent and the job seeker conducted to find a match between the job seeker and the recruiter the job seeker wishes to be considered. Through this session, the agent aims to gather sufficient information to evaluate whether the job seeker meets the recruiter's selection criteria. Alternatively, instead of having job seekers select candidates they wish to interview from a list of proposed candidates, they may choose candidates to exclude from the list. By selecting or excluding candidates from the list, it is possible to eliminate candidates who have already undergone interviews or who do not match the job seeker's preferences, thereby narrowing down the candidates who will proceed to the next stage of selection.

[0040] When the agent inputs the job seeker information collected through the session into the recruitment support system 10a, the evaluation unit 15, based on the input information, refers to the recruiter database DB2 and the evaluation criteria database DB3 to output an evaluation of whether the job seeker meets the recruiter's selection criteria. The output evaluation is transmitted from the recruitment support system 10a to the job seeker terminal 2 (S105).

[0041] If the evaluation results in a job seeker meeting the employer's selection criteria, the job seeker sends an interview request to the recruitment support system 10a via the job seeker terminal 2 (S106). In response to the job seeker's interview request, the recruitment support system 10a sends the job seeker's interview request, along with the job seeker information stored in the job seeker database DB1, to the employer terminal 3 (S107). The recruiter conducts an interview with the job seeker based on the job seeker information, in response to the interview intention received from the recruitment support system 10a, and decides whether or not to hire the job seeker (S108).

[0042] This allows job seekers to receive proposals for candidates that match their preferences, rather than having to choose from a huge number of applicants, and to focus their interviews on candidates with a high probability of being hired. On the other hand, employers can significantly reduce the burden of recruitment activities by only having to go through the selection process with promising job seekers. [Examples]

[0043] Figure 3 shows the recruitment support system 10b according to Example 2. The recruitment support system 10b according to this embodiment executes the hearings and sessions conducted in the recruitment support system 10a according to Embodiment 1 described above using AI (Artificial Intelligence) functions. In the figures, functional components denoted by the same reference numerals as in Example 1 perform the same functions as in Example 1 unless otherwise specified.

[0044] The recruitment support device 1b is configured to communicate with the artificial intelligence unit group 60. The artificial intelligence unit group 60 includes multiple artificial intelligence units that behave differently from each other, namely the first artificial intelligence unit 61, the second artificial intelligence unit 62, the third artificial intelligence unit 63, and the fourth artificial intelligence unit 64. Each of the artificial intelligence units 61 to 64 included in the artificial intelligence unit group 60 has an AI (Artificial Intelligence) function. The artificial intelligence unit group 60 can also search appropriate databases, which will be described in detail later, and obtain information.

[0045] Here, we will explain the concept of the behavior of the artificial intelligence group 60. The first artificial intelligence unit 61 acts as a "controller AI" that primarily receives instructions from the recruitment support device 1b and controls the second artificial intelligence unit 62 and the third artificial intelligence unit 63 by generating instructions to send to them. The second artificial intelligence unit 62 acts as an "interviewer AI" that generates questions to actually ask job seekers in response to instructions from the first artificial intelligence unit 61. The third artificial intelligence unit 63 acts as a "proposal AI" that proposes suitable candidates to job seekers based on their responses in response to instructions from the first artificial intelligence unit 61. The fourth artificial intelligence unit 64 acts as an "evaluation AI" that evaluates the abilities and personality of job seekers based on their responses in response to instructions from the first artificial intelligence unit 61.

[0046] Figure 4 shows the specific behavior of the artificial intelligence unit group 60. Figure 4(a) shows the hearing process, and Figure 4(b) shows the session process. First, regarding the interview process, as shown in Figure 4(a), when the first artificial intelligence unit 61 receives appropriate instructions from the recruitment support system 10b, it refers to the interview agenda stored in the agenda database DB4 and determines the question outline. The question outline includes, for example, job seeker information to be elicited from the job seeker. The first artificial intelligence unit 61 also instructs the second artificial intelligence unit 62 to generate specific question sentences to be sent to the job seeker terminal 2 based on the question outline.

[0047] The second artificial intelligence unit 62, in response to a question generation instruction received from the first artificial intelligence unit 61, appropriately refers to the question database DB5, generates a question for the recruiter's proposal, and transmits it to the job seeker terminal 2. The job seeker recognizes the question via the job seeker terminal 2 and inputs their answer. The input answer is transmitted from the recruitment support system 10b to the first artificial intelligence unit 61. The response may be in the form of video data, audio data, or text data, or it may be text data obtained by appropriate transcription processing from video data or audio data. The first artificial intelligence unit 61 determines the gist of the next questions based on the job seeker's response and the interview agenda.

[0048] Furthermore, the first artificial intelligence unit 61 transmits the responses received from job seekers or appropriate information based on those responses to the third artificial intelligence unit 63, and also sends a suggestion instruction to the third artificial intelligence unit 63 to propose suitable candidates to the job seekers. The third artificial intelligence unit 63 performs appropriate processing to propose suitable candidates to the job seekers in accordance with the suggestion instruction and outputs the processing results to the first artificial intelligence unit 61. The first artificial intelligence unit 61 may, if necessary, extract further question summaries based on the processing results and instruct the second artificial intelligence unit 62 to generate specific question sentences.

[0049] In this way, the first artificial intelligence unit 61, the second artificial intelligence unit 62, and the third artificial intelligence unit 63 work together to conduct interviews with job seekers via the job seeker terminal 2. By generating instructions from one artificial intelligence unit to other artificial intelligence units, and having multiple artificial intelligence units operate accordingly, the workload for each artificial intelligence unit is reduced, and the prompt configuration can be simplified. Furthermore, since the overall prompt length can be reduced, unexpected behavior or excessive omission of behavior in each artificial intelligence unit can be suppressed, allowing for proper operation. Ultimately, this enables comprehensive questioning of job seekers during interviews and the appropriate acquisition of necessary information from them.

[0050] In this embodiment as well, the job seeker database DB1 does not necessarily have to be prepared in advance, and job seeker information may be collected through interviews. Furthermore, if a job seeker provides some job seeker information in advance and this is stored in the job seeker database DB1, the artificial intelligence group 60 may acquire job seeker information that supplements the information provided by the job seeker, or it may uniformly acquire the necessary job seeker information regardless of the job seeker information stored in the job seeker database DB1. In addition, when supplementing job seeker information, the system may refer to the job seeker database DB1 as appropriate to identify the job seeker information to be supplemented.

[0051] Next, regarding the session process, as shown in Figure 4(b), the first artificial intelligence unit 61, similar to the hearing described above, receives appropriate instructions from the recruitment support system 10b, refers to the interview agenda stored in the agenda database DB4, and determines the question outline. The question outline includes, for example, job seeker information to be extracted from the job seeker. The first artificial intelligence unit 61 also instructs the second artificial intelligence unit 62 to generate specific question sentences to be sent to the job seeker terminal 2 based on the question outline.

[0052] The second artificial intelligence unit 62, in response to a question generation instruction received from the first artificial intelligence unit 61, appropriately refers to the question database DB5, generates evaluation questions, and transmits them to the job seeker terminal 2. The job seeker recognizes the questions via the job seeker terminal 2 and inputs their answers. The input answers are transmitted from the recruitment support system 10b to the first artificial intelligence unit 61. Based on the job seeker's answers and the interview agenda, the first artificial intelligence unit 61 determines the summary of the next questions to ask.

[0053] Furthermore, the first artificial intelligence unit 61 transmits the responses received from job seekers or appropriate information based on those responses to the fourth artificial intelligence unit 64, and also sends an evaluation instruction to the fourth artificial intelligence unit 64 to evaluate the job seekers. The fourth artificial intelligence unit 64 performs appropriate processing to evaluate the job seekers in accordance with the evaluation instruction and outputs the processing results to the first artificial intelligence unit 61. The first artificial intelligence unit 61 may, if necessary, extract further question summaries based on the processing results and instruct the second artificial intelligence unit 62 to generate specific question sentences.

[0054] In this way, the first artificial intelligence unit 61, the second artificial intelligence unit 62, and the fourth artificial intelligence unit 64 work together to conduct a session with the job seeker via the job seeker terminal 2. Similar to the hearing described above, this session also generates instructions from one artificial intelligence unit to other artificial intelligence units, allowing multiple artificial intelligence units to operate independently. This ensures that all necessary questions are asked of the job seeker and that the required information is appropriately obtained from them.

[0055] The recruitment support system 10b has a proposal device 4 and an evaluation device 5 as replacements for the proposal unit 14 and evaluation unit 15 provided in the recruitment support device 1a according to the above-described embodiment 1. Specifically, the proposal device 4 conducts interviews with job seekers, refers to the basic information in the recruiter database DB2, and proposes recruiters that match the job seeker's preferences based on the content of the interviews. The evaluation device 5 conducts sessions with job seekers, refers to the recruiter database DB2 and the evaluation criteria database DB3, and evaluates whether the job seeker meets the recruiter selection criteria based on the information of the recruiters the job seeker wishes to be considered for.

[0056] The proposed device 4 is a server device mainly equipped with an AI control unit 41 and a communication processing unit 42, and similarly, the evaluation device 5 is a server device mainly equipped with an AI control unit 51 and a communication processing unit 52. However, notwithstanding this embodiment, the proposed device 4 and the evaluation device 5 may be included in the recruitment support device 1b as functional units constituting the recruitment support device 1b.

[0057] ● AI control unit 41, 51 The AI ​​control units 41 and 51 are functional units that primarily control input to the artificial intelligence unit group 60. The various instructions generated by the AI ​​control units 41 and 51 are, for example, prompts, but are not limited to any instructions that the artificial intelligence unit group 60 can interpret. When sending multiple instructions to the artificial intelligence unit group 60, the AI ​​control units 41 and 51 provide instruction blocks and send instructions for each instruction block. In addition, the AI ​​control units 41 and 51 send an ordered list that sets the execution order of each instruction along with the instructions. With this configuration, the artificial intelligence unit group 60 can execute instructions accurately.

[0058] The AI ​​control units 41 and 51 send prompts to the first artificial intelligence unit 61, the second artificial intelligence unit 62, the third artificial intelligence unit 63, and the fourth artificial intelligence unit 64, each assigning them different roles. For example, the AI ​​control unit 41 assigns the first artificial intelligence unit 61 the role of a controller that controls the second artificial intelligence unit 62 and the third artificial intelligence unit 63, the second artificial intelligence unit 62 the role of an interviewer that asks questions to job seekers via the job seeker terminal 2, and the third artificial intelligence unit 63 the role of a proposer that suggests suitable candidates to job seekers. More specifically, the AI ​​control unit 41 sends a prompt to the second artificial intelligence unit 62, for example, saying, "Your role is to ask questions to candidates based on instructions received from the controller." Furthermore, the AI ​​control unit 51 assigns the first artificial intelligence unit 61 the role of a controller for controlling the second artificial intelligence unit 62 and the fourth artificial intelligence unit 64, assigns the second artificial intelligence unit 62 the role of an interviewer for asking questions to job seekers via the job seeker terminal 2, and assigns the fourth artificial intelligence unit 64 the role of an evaluator for evaluating job seekers. By using AI control units 41 and 51 to assign different roles to individual artificial intelligence units 61 to 64, the roles of each artificial intelligence unit 61 to 64 become simple and clear, allowing the artificial intelligence unit group 60 to perform appropriate operations.

[0059] Furthermore, the AI ​​control units 41 and 51 assign different objectives to the first artificial intelligence unit 61, the second artificial intelligence unit 62, the third artificial intelligence unit 63, and the fourth artificial intelligence unit 64, respectively. For example, the AI ​​control unit 41 sends a prompt to the second artificial intelligence unit 62 such as, "The purpose is to collect information from job seekers that is necessary to propose candidates that meet the job seekers' preferences." Similarly, the AI ​​control unit 51 sends a prompt to the second artificial intelligence unit 62 such as, "The purpose is to collect information from job seekers that is necessary to evaluate them according to the selection criteria of the recruiters."

[0060] Furthermore, the AI ​​control units 41 and 51 may each provide instruction blocks to the first artificial intelligence unit 61, the second artificial intelligence unit 62, the third artificial intelligence unit 63, and the fourth artificial intelligence unit 64, and define the order in which instructions are executed in an ordered list. With such a configuration, even when multiple instructions are given to each of the artificial intelligence units 61 to 64, they can be made to follow the instructions more accurately.

[0061] Furthermore, the AI ​​control units 41 and 51 transmit, for example, question instructions to the first artificial intelligence unit 61 via the communication processing units 42 and 52 to generate questions for the job seeker. The AI ​​control units 41 and 51 then refer to the agenda database DB4 to determine the interview agenda for the first artificial intelligence unit 61.

[0062] Here, the agenda database DB4 is a database that stores a group of job seeker information items, or so-called interview agenda, which includes multiple items of information that should be extracted from job seekers. These job seeker information items may include items such as the job seeker's "approach to work" and "past experience." The agenda database DB4 may also store summaries of questions used to extract each job seeker's information, associated with the job seeker's information. Furthermore, the agenda database DB4 stores information from job seekers, such as summaries of their responses, in association with job seeker information fields.

[0063] Agenda database DB4 stores multiple interview agendas containing different job information items. Furthermore, agenda database DB4 stores different interview agendas for hearings and sessions. Session-specific interview agendas may be stored according to the recruiter, for example. Interview agendas may also be stored according to job attributes such as the recruiter's industry and the job being advertised. Additionally, interview agendas may be stored according to the job seeker's characteristics, for example, whether they are a new graduate or an experienced hire, or according to their educational background, etc.

[0064] If multiple interview agendas are stored in the agenda database DB4, the AI ​​control units 41 and 51 may send a selection instruction to the first artificial intelligence unit 61 to select an interview agenda to apply to the hearing or session. In this case, the first artificial intelligence unit 61 has a suitable learning model that takes process identification information indicating whether it is a hearing or a session as input and outputs identification information of the interview agenda to be applied. It may also have a suitable pre-trained model that takes the identification information of the recruiter or job attributes as input and outputs identification information of the interview agenda to be applied.

[0065] Furthermore, the process of selecting an interview agenda to be applied to a hearing or session from the agenda database DB4 may be performed by an appropriate selection processing unit of the selection support device 1b, instead of by instructing the first artificial intelligence unit 61. In this case, the selection support device 1b may, for example, maintain in advance an appropriate table that associates process identification information with the interview agenda to be applied, and select the interview agenda by referring to this table. Furthermore, regardless of this embodiment, the agenda database DB4 may be prepared separately for the hearing and for the session. In this case, the AI ​​control units 41 and 51 may each send instructions to specify the agenda database to be referenced, either for the hearing or for the session, or the recruitment support device 1b may perform a process to select the agenda database to be referenced.

[0066] The AI ​​control units 41 and 51 instruct the first artificial intelligence unit 61 to conduct the hearing or session in accordance with the interview agenda. The interview agenda may specify, for example, that the hearing or session should proceed with an "opening" and a "question time" in that order. If the agenda is "opening," the AI ​​control units 41 and 51 send instructions to the second artificial intelligence unit 62 to generate messages such as greetings, introductions to the dialogue, and self-introductions. If the agenda is "question time," the AI ​​control units 41 and 51 determine the gist of the questions and then send instructions to the second artificial intelligence unit 62 to generate a prompt requesting specific question sentences in accordance with that gist.

[0067] The instruction for generating a question prompt may include an instruction to cause the second artificial intelligence unit 62 to refer to the question database DB5 when generating the question.

[0068] Question Database DB5 is a database that stores example questions to be presented to job seekers. The questions may be stored linked to a summary of the question. Question Database DB5 may also store example follow-up questions. Follow-up questions are questions that add appropriate wording to the content of the previous answer, and have the effect of continuing the conversation and conveying interest in the other person, thereby encouraging them to talk more.

[0069] The AI ​​control units 41 and 51 may send a summary generation instruction to the first artificial intelligence unit 61 to generate a summary text that summarizes the answers obtained from the job seeker terminal 2. Alternatively, the AI ​​control units 41 and 51 may send an instruction to the first artificial intelligence unit 61 to store the generated summary text in the agenda database DB4, associating it with the question summary and the question text. Alternatively, the recruitment support device 1b may be configured to store the summary text obtained from the first artificial intelligence unit 61 in the agenda database DB4. In large-scale language models, the entire text of past dialogues is stored and referenced up to a set limit on the amount of tokens, but with a configuration that generates and stores a summary text, past dialogue information can be efficiently accumulated and referenced.

[0070] Furthermore, the AI ​​control units 41 and 51 may transmit specific instructions to the first artificial intelligence unit 61 to identify content related to the question summary of the question being generated from the answer summary, i.e., summary-related content. The summary-related content includes scenes, situations, episodes, etc., described in the summary. In addition, the AI ​​control units 41 and 51 transmit instructions to the first artificial intelligence unit 61 to include at least a portion of the summary-related content in the question in the prompt to be transmitted to the second artificial intelligence unit 62.

[0071] With this configuration, for example, if the previous answer was about the candidate's activity history in a relevant project, the next question can be phrased as "In that project..." to continue the conversation. The summary-related content is identified from the previous answer and is related to the summary of the next question, so it naturally connects the candidate's answer with the next question. Therefore, by using a configuration that identifies content related to the summary of the next question from the previous answer and includes it in the next question, the recruitment support device 1b can ask questions that build upon the content of the previous answer, enabling a more natural dialogue. Furthermore, by summarizing the previous answer and then identifying the summary-related content from the summary, a more natural dialogue can be mimicked, enabling a smoother conversation.

[0072] Furthermore, while multiple questions are being asked regarding the same question summary, the summaries of the multiple answers to those questions may be used as the extraction range for extracting summary-related content. Also, if the topic changes during the interview, for example, when generating a question related to a different question summary, the stored summaries may be refreshed and excluded from the extraction range. This configuration prevents unnatural shifts in topic and saves token usage.

[0073] Furthermore, the AI ​​control units 41 and 51 may transmit a re-questioning necessity determination instruction to the first artificial intelligence unit 61 to determine whether or not to ask further questions based on the job seeker's answers. A re-question is a process of presenting a question with a similar gist again, and this question may include, for example, a question that is phrased differently when it is suspected that the job seeker's answer does not grasp the intent of the question, and a question that delves deeper into the content when the job seeker's answer lacks specificity.

[0074] The AI ​​control units 41 and 51 instruct the first artificial intelligence unit 61 to send the generated prompt to the second artificial intelligence unit 62. In this way, the second artificial intelligence unit 62 generates a question sentence in response to the prompt generated by the first artificial intelligence unit 61, and the recruitment support device 1b acquires it. The AI ​​control unit 41 may also send prompt generation instructions to the first artificial intelligence unit 61 using UML (Unified Modeling Language). UML is a standardized language for visually representing the processing flow of a system.

[0075] During the hearing, the AI ​​control unit 41 generates a prompt in response to the job seeker's answers that suggests suitable candidates based on the job seeker's intentions and conditions, and sends an instruction to the third artificial intelligence unit 63 to send this prompt as a suggestion instruction. Furthermore, the AI ​​control unit 51, during a session, sends an instruction to the first artificial intelligence unit 61 to generate a prompt to evaluate the job seeker's abilities, personality, etc., in response to the job seeker's answers, and to send the prompt as an evaluation instruction to the fourth artificial intelligence unit 64.

[0076] Furthermore, the input information to the third artificial intelligence unit 63 or the fourth artificial intelligence unit 64 may be, in lieu of or in addition to the answer, information obtained by analyzing the input answer using natural language processing. Natural language processing analysis is performed by appropriate processing such as morphological analysis, syntactic analysis, or semantic analysis. As a preprocessing step, natural language processing analysis may be performed on a rule basis by an appropriate functional unit provided by the adoption support device 1b. Alternatively, the analysis using natural language processing may be performed by the artificial intelligence unit group 60.

[0077] The AI ​​control units 41 and 51 transmit an instruction to the second artificial intelligence unit 62 via the communication processing unit 16, instructing it to operate in response to a question from the first artificial intelligence unit 61.

[0078] The AI ​​control unit 41 transmits an instruction to the third artificial intelligence unit 63 via the communication processing unit 16, indicating that it will operate based on the proposed instruction from the first artificial intelligence unit 61. The proposed instruction is transmitted from the AI ​​control unit 41 to the third artificial intelligence unit 63. The AI ​​control unit 51 transmits an instruction to the fourth artificial intelligence unit 64 via the communication processing unit 16, instructing it to operate based on the evaluation instruction from the first artificial intelligence unit 61. The evaluation instruction is transmitted from the AI ​​control unit 51 to the fourth artificial intelligence unit 64. Furthermore, instructions to the third artificial intelligence unit 63 or the fourth artificial intelligence unit 64 shall include both configurations: one in which instructions are given indirectly from the AI ​​control units 41 and 51 via the first artificial intelligence unit 61, and another in which instructions are given directly from the AI ​​control units 41 and 51.

[0079] Here, we will describe the details of the proposal and evaluation instructions transmitted from the AI ​​control units 41 and 51 directly, or via the first artificial intelligence unit 61, to the third artificial intelligence unit 63 or the fourth artificial intelligence unit 64.

[0080] The third artificial intelligence unit 63 outputs data by appropriately referring to the recruiter database DB2. As described above, the recruiter database DB2 stores basic information about recruiters as information to determine whether the job seeker's preferences match those of the recruiters, and identifies recruiters who match the job seeker's preferences based on the information gathered from the interview with the job seeker.

[0081] The fourth artificial intelligence unit 64 outputs data by appropriately referring to the recruit database DB2 and the evaluation criteria database DB3. As described above, the recruit database DB2 stores the selection criteria for recruits, and the evaluation criteria database DB3 stores, for example, evaluation items related to abilities and sentences indicating the evaluation criteria corresponding to each score of the evaluation items. The evaluation criteria database DB3 may also store evaluation items related to the session performance, etc.

[0082] The AI ​​control unit 51 transmits, for example, a text generation instruction to the fourth artificial intelligence unit 64 to generate text that expresses the applicant's abilities from the perspective of each of several evaluation items, based on multiple responses. The evaluation items are, for example, items related to abilities, and are broadly divided into major categories such as situation / behavioral evaluation, social skills evaluation, and personality evaluation. In addition, multiple sub-items are set for each major category. For example, under situation / behavioral evaluation, there are sub-items such as spontaneity and proactiveness. The AI ​​control unit 51 generates text from the perspective of each sub-item, for example. Alternatively, the AI ​​control unit 41 may generate text from the perspective of each major category. The AI ​​control unit 51 evaluates the entire interview, including multiple responses, and generates text that expresses the applicant's abilities from the perspective of each evaluation item.

[0083] Furthermore, the AI ​​control unit 51 transmits a score determination instruction to the fourth artificial intelligence unit 64, which determines the score for the corresponding evaluation item based on the text describing the applicant's abilities obtained in response to the text creation instruction. The score is expressed numerically, for example, and is determined on a maximum of 5 points in increments of 0.1 points. Each score may be associated with a written evaluation criterion. For example, an evaluation criterion that results in a score of 5 points might be "possesses excellent logical thinking ability." On the other hand, an evaluation criterion that results in a score of 1 point might be "unable to analyze problems." In this way, by defining clear evaluation criteria for each evaluation item and providing them to the fourth artificial intelligence unit 64, fair evaluation becomes possible.

[0084] The AI ​​control unit 51 calculates a score by integrating the results of the evaluation for each evaluation item and sends an evaluation instruction to the fourth artificial intelligence unit 64 to evaluate whether the score meets the selection criteria required by the employer. With this configuration, an accurate evaluation is output based on a precise evaluation of each evaluation item.

[0085] The AI ​​control units 41 and 51 may instruct the first artificial intelligence unit 61 to send instructions to the second artificial intelligence unit 62, the third artificial intelligence unit 63, and the fourth artificial intelligence unit 64 in the form of mechanically structured messages. The messages are, for example, codes whose format is standardized by so-called Structured Output.

[0086] The AI ​​control units 41 and 51 may specify the agenda database DB4, the question database DB5, and the evaluation criteria database DB3 as attached documents to be retrieved by RAG (Retrieval-Augmented Generation). RAG is a technology that retrieves information from specified attached documents and generates answers and texts based on that information. In other words, the data designated as attached documents is not machine-learned by the artificial intelligence unit group 60, but is treated as an appropriate search target. With this configuration, it becomes possible to ask questions, make proposals, and evaluate in a way that is independent of the learning of the artificial intelligence unit group 60.

[0087] In this embodiment, the acquisition unit 12 acquires proposals for job seekers from the third artificial intelligence unit 63 via the communication processing unit 16, and acquires evaluations of job seekers from the fourth artificial intelligence unit 64. Furthermore, the communication processing unit 16 is a processing unit that enables the transmission and reception of data with the job seeker terminal 2 via a network NW such as the Internet, in accordance with a predetermined protocol. For example, the communication processing unit 16 transmits a question sentence generated by the first artificial intelligence unit 61 to the job seeker terminal 2. The communication processing unit 16 also receives, for example, an answer to the question sentence from the job seeker terminal 2.

[0088] ● Artificial Intelligence Group 60 The Artificial Intelligence Group 60 is an artificial intelligence (AI) equipped with language models such as transformers including BART (Bidirectional and Auto-regressive Transformer), BERT (Bidirectional Encoder Representations from Transformers), or GPT (Generative Pretrained Transformer, including GPT-1, GPT-2, GPT-3, GPT-4, etc.), and particularly learning models such as large language models (LLM). A learning model (also called a machine learning model) refers to a learning model based on a machine learning algorithm. Specific machine learning algorithms include nearest neighbors, naive Bayes, decision trees, and support vector machines. Deep learning, which uses neural networks to generate features and connection weights for learning, is another example. The Artificial Intelligence Group 60 can appropriately apply the above algorithms.

[0089] The artificial intelligence unit 60 has a pre-trained model that has been appropriately machine-learned. The training data may be provided by an administrator or the like, or it may include information collected from the internet or other sources. The artificial intelligence unit group 60 is mainly composed of a first artificial intelligence unit 61, a second artificial intelligence unit 62, a third artificial intelligence unit 63, and a fourth artificial intelligence unit 64, each assigned a different role. Each artificial intelligence unit 61 to 64 includes, for example, the following trained models.

[0090] Specifically, the first artificial intelligence unit 61 includes a first trained model that takes an interview agenda as input and outputs a summary of questions. The first artificial intelligence unit 61 also includes a second trained model that takes job seeker information or a summary of questions as input and outputs an instruction to the second artificial intelligence unit 62 to generate a question prompt. Furthermore, the first artificial intelligence unit 61 includes a third trained model that takes the job seeker's response as input and outputs a summary of that response. In addition, the first artificial intelligence unit 61 includes a fourth trained model that takes the question summary and summary as input and outputs content related to the summary. This fourth trained model is input with the question summary related to the next question and the summary of the previous response.

[0091] Furthermore, the first artificial intelligence unit 61 includes a fifth trained model that takes an interview agenda, in which answers or summaries of answers are stored in association with job seeker information items, as input, and outputs suggestion instructions to the third artificial intelligence unit 63. Furthermore, the first artificial intelligence unit 61 includes a sixth trained model that takes an interview agenda in which answers or summaries of answers are associated with job seeker information items as input, and outputs evaluation instructions to the fourth artificial intelligence unit 64.

[0092] The second artificial intelligence unit 62 has, for example, a seventh trained model that takes job seeker information or a summary of questions as input and outputs question text to be presented to the job seeker. The second artificial intelligence unit 62 may also have an appropriate trained model that outputs an image or video of an avatar displayed during interviews or sessions with the job seeker, or identification information of the image or video. The seventh trained model may be configured to output the avatar image or video or its identification information along with the question text. When the seventh trained model outputs avatar identification information, for example, the recruitment support device 1b stores the avatar image or video in advance, and the recruitment support device 1b identifies the image or video to be displayed on the job seeker terminal 2 based on the outputted identification information and transmits it to the job seeker terminal 2. Furthermore, the second artificial intelligence unit 62 includes an appropriate eighth trained model that takes at least the question summary and summary-related content identified from the summary text as input and outputs the question text.

[0093] The third artificial intelligence unit 63, for example, has a ninth trained model that takes the job seeker's responses as input and outputs candidate candidates that match the job seeker's preferences.

[0094] The fourth artificial intelligence unit 64 has a tenth trained model that takes evaluation items related to the job seeker's responses and session performance as input, and outputs an evaluation indicating whether the scores in said evaluation items meet the selection criteria required by the employer. Furthermore, the tenth trained model may output the basis information for the scores and the answers that serve as the basis for the scores. Each pre-trained model has already obtained appropriate training data, and additional training can be performed as needed.

[0095] The first artificial intelligence unit 61, in response to instructions from the AI ​​control unit 41, outputs a summary of questions based on the interview agenda using the first trained model. The first artificial intelligence unit 61 also, in response to instructions from the AI ​​control unit 41, outputs an instruction to the second artificial intelligence unit 62 to generate a question prompt based on the job seeker information or the summary of questions using the second trained model.

[0096] The first artificial intelligence unit 61 outputs a summary of the job seeker's response based on that response. Furthermore, the first artificial intelligence unit 61 refers to the interview agenda in which the response or the summary of the response is associated with job seeker information items and outputs a suggestion instruction to the third artificial intelligence unit 63.

[0097] The first artificial intelligence unit 61 outputs a summary of the job seeker's response based on that response. Furthermore, the first artificial intelligence unit 61 refers to the interview agenda in which the response or the summary of the response is associated with job seeker information items and outputs an evaluation instruction to the fourth artificial intelligence unit 64.

[0098] ●Processing flow (interview) As shown in Figure 5, first, the job seeker terminal 2 displays, for example, an appropriate interview start screen and accepts input for the interview start instruction (step S201), and transmits this to the recruitment support system 10b. Prior to the start of the interview, the job seeker's personal information may be accepted. Also, prior to the start of the interview, the job seeker may be accepted to select companies to apply to, or to select companies to exclude from their application.

[0099] The recruitment support system 10b (proposed device 4) transmits a first role instruction to the first artificial intelligence unit 61, stating that the first artificial intelligence unit 61 will act as a controller for the second artificial intelligence unit 62 and the third artificial intelligence unit 63 (step S202).

[0100] The recruitment support system 10b (proposed device 4) sends a second role instruction to the second artificial intelligence unit 62, stating that the second artificial intelligence unit 62 will act as an interviewer, exchanging messages with the job seeker via the job seeker terminal 2 using voice or text (step S203).

[0101] The recruitment support system 10b (proposal device 4) transmits a third role instruction to the third artificial intelligence unit 63, stating that the third artificial intelligence unit 63 is in the role of a proposer, suggesting candidates to the job seeker based on the job seeker's responses (step S204). In this way, the recruitment support system 10b (proposed device 4) assigns different roles to the artificial intelligence units 61, 62, and 63. Steps S202 to S204 can be performed in any order and may be performed simultaneously.

[0102] Next, the recruitment support system 10b sends an instruction to the first artificial intelligence unit 61 to begin interviewing job seekers (step S205).

[0103] The first artificial intelligence unit 61 refers to the agenda database DB4, extracts the existence and type of job seeker information to be retrieved (step S206), and sends an instruction to the second artificial intelligence unit 62 to generate a question based on the summary of questions regarding the job seeker information (step S207).

[0104] The second artificial intelligence unit 62 generates a question sentence by referring to the question database DB5 in response to instructions from the first artificial intelligence unit 61, and outputs the generated question sentence to the employment support system 10b (step S208).

[0105] The recruitment support system 10b transmits the question text obtained from the second artificial intelligence unit 62 to the job seeker terminal 2 (step S209).

[0106] The job seeker terminal 2 receives the response from the job seeker and transmits it to the recruitment support system 10b (step S210).

[0107] The recruitment support system 10b stores the question summary, question text, and answer in the agenda database DB4 (step S211).

[0108] Next, the recruitment support system 10b sends a summarization instruction to the first artificial intelligence unit 61, along with the response obtained from the job seeker terminal 2 (step S212).

[0109] The first artificial intelligence unit 61 obtains the answer from the recruitment support system 10b and summarizes the answer (step S213). Next, the first artificial intelligence unit 61 stores the generated summary in the agenda database DB4, linking it to the question summary or question text (step S214).

[0110] Next, the first artificial intelligence unit 61 sends a suggestion instruction to the third artificial intelligence unit 63 to propose a candidate based on the answer or a summary of the answer (step S215).

[0111] The third artificial intelligence unit 63 refers to the recruiter database DB2 to identify recruiters who match the job seeker's preferences and outputs a proposal to the first artificial intelligence unit 61 (step S216).

[0112] ● Processing flow (session) As shown in Figure 6, first, the job seeker terminal 2 displays, for example, an appropriate session start screen and accepts the input of a session start instruction (step S301), and transmits this to the recruitment support system 10b. Prior to the start of the session, the system receives information from the job seeker about the candidates they wish to consider for selection. The job seeker's preferences can be, for example, from the suggestions mentioned above. The system may also accept the selection of candidates to exclude from the job seeker's suggestions.

[0113] The recruitment support system 10b (evaluation device 5) transmits a first role instruction to the first artificial intelligence unit 61, stating that the first artificial intelligence unit 61 will act as a controller for the second artificial intelligence unit 62 and the fourth artificial intelligence unit 64 (step S302).

[0114] The recruitment support system 10b (evaluation device 5) sends a second role instruction to the second artificial intelligence unit 62, stating that the second artificial intelligence unit 62 will act as an interviewer, exchanging messages with the job seeker via the job seeker terminal 2 using voice or text (step S303).

[0115] The recruitment support system 10b (evaluation device 5) transmits a third role instruction to the fourth artificial intelligence unit 64, stating that the fourth artificial intelligence unit 64 is in the role of an evaluator who evaluates job seekers based on their responses (step S304). In this way, the recruitment support system 10b (evaluation device 5) assigns different roles to the artificial intelligence units 61, 62, and 64. Steps S302 to S304 can be performed in any order and may be performed simultaneously.

[0116] Next, the recruitment support system 10b sends an instruction to the first artificial intelligence unit 61 to start a session with the job seeker (step S305).

[0117] The first artificial intelligence unit 61 refers to the agenda database DB4, extracts the existence and type of job seeker information to be retrieved (step S306), and sends an instruction to the second artificial intelligence unit 62 to generate a question based on the summary of questions regarding the job seeker information (step S307).

[0118] The second artificial intelligence unit 62 generates a question sentence by referring to the question database DB5 in response to instructions from the first artificial intelligence unit 61, and outputs the generated question sentence to the employment support system 10b (step S308).

[0119] The recruitment support system 10b transmits the question text obtained from the second artificial intelligence unit 62 to the job seeker terminal 2 (step S309).

[0120] The job seeker terminal 2 receives the response from the job seeker and transmits it to the recruitment support system 10b (step S310).

[0121] Next, the recruitment support system 10b stores the question summary, question text, and answer in the agenda database DB4 (step S311).

[0122] Next, the recruitment support system 10b sends a summarization instruction to the first artificial intelligence unit 61, along with the response obtained from the job seeker terminal 2, to summarize the response (step S312).

[0123] The first artificial intelligence unit 61 obtains the answer from the recruitment support system 10b and summarizes the answer (step S313). Next, the first artificial intelligence unit 61 stores the generated summary in the agenda database DB4, linked to the question summary or question text (step S314).

[0124] Next, the first artificial intelligence unit 61 sends an evaluation instruction to the fourth artificial intelligence unit 64 to evaluate the job seeker based on their response or a summary of their response (step S315).

[0125] The fourth artificial intelligence unit 64 evaluates job seekers by referring to the recruit database DB2 and the evaluation criteria database DB3, and outputs the evaluation results to the first artificial intelligence unit 61 (step S316).

[0126] After the evaluation of the job seeker is output, the evaluation results are sent to the job seeker terminal 2, as in the first example described above. If the job seeker meets the employer's selection criteria, the job seeker sends an interview request to the recruitment support system 10b via the job seeker terminal 2. In response to the job seeker's interview request, the job seeker information and the job seeker's interview request are sent to the employer terminal 3, and the employer conducts a recruitment interview with the job seeker accordingly.

[0127] This embodiment significantly reduces the burden on both job seekers and employers during the job search or recruitment process. Furthermore, because interviews and sessions are conducted by AI, the psychological burden on job seekers is reduced, and the constraints on time and location for conducting interviews and sessions are also relaxed.

[0128] In addition, in the embodiment described above, the job interview between the job seeker and the employer may also be conducted using AI functions.

[0129] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of its gist. [Explanation of symbols]

[0130] 10a: Recruitment support system 1a:Recruitment support device 10b: Recruitment support system 1b:Recruitment support device 11: Display Control Unit 12: Acquisition part 13: Memory Control Unit 14: Proposal Department 15: Evaluation Department 16: Communication Processing Unit 2: Job seeker terminal 3: Employee terminal 4: Proposed device 5: Evaluation device 41:AI control section 42: Communication Processing Unit 51:AI control section 52: Communication Processing Unit 60: Artificial Intelligence Department 61: First Artificial Intelligence Department 62: Second Artificial Intelligence Department 63: Third Artificial Intelligence Department 64: 4th Artificial Intelligence Department DB1: Job seeker database DB2: Recruitment Database DB3: Evaluation Criteria Database DB4: Agenda Database DB5: Question Database NW: Network

Claims

1. The job seeker's terminal, used by the job seeker, and the recruiter's terminal, are configured to communicate with each other. A job seeker information storage unit that stores the information of the aforementioned job seeker, A basic information storage unit that stores the basic information of the aforementioned recruit, An evaluation criteria storage unit that stores the evaluation criteria for the aforementioned hires, A proposal unit that, by referring to the basic information storage unit, proposes candidates who match the job seeker's preferences, A selection request receiving unit receives the selection request from the job seeker's terminal, and An evaluation unit that evaluates whether the job seeker meets the selection criteria of the employee who wishes to be selected, An evaluation transmission unit that transmits the evaluation of the job seeker to the job seeker terminal, The interview intention receiving unit receives an interview intention from the job seeker's terminal with the person who has been determined to meet the selection criteria. An interview request transmission unit transmits an interview request to the aforementioned recruiter terminal, along with the applicant information of the applicant who has indicated their willingness to be interviewed. Having, Recruitment support system.

2. The proposal unit consists of a proposal device that conducts interviews with the job seeker and outputs companies to propose to the job seeker. The contents of the aforementioned hearing include at least multiple answers from the job seeker to the questions, The proposed device is, An AI control unit that sends instructions to the artificial intelligence unit, A communication processing unit that acquires information on recruits to propose to the job seeker from the artificial intelligence unit, Equipped with, The AI ​​control unit provides the artificial intelligence unit with respect to the AI ​​control unit. Based on the multiple aforementioned responses, the system outputs a proposal instruction that identifies the candidate whose basic information matches the job seeker's preferences as the candidate to be proposed to the job seeker, Send The recruitment support system according to claim 1.

3. The evaluation unit is comprised of an evaluation device that performs a session with the job seeker and outputs an evaluation of the job seeker. The content of the aforementioned session includes at least several answers from the job seeker to the questions, The evaluation device is An AI control unit that sends instructions to the artificial intelligence unit, A communication processing unit that acquires the evaluation results of the job applicant from the artificial intelligence unit, Equipped with, The AI ​​control unit provides the artificial intelligence unit with respect to the AI ​​control unit. Based on the multiple aforementioned responses and the aforementioned selection criteria, an evaluation instruction is provided to output an evaluation of the job applicant, Send The recruitment support system according to claim 1.

4. The job seeker's terminal, used by the job seeker, and the recruiter's terminal, are configured to communicate with each other. A job seeker information storage unit that stores the information of the aforementioned job seeker, A basic information storage unit that stores the basic information of the aforementioned recruit, An evaluation criteria storage unit that stores the evaluation criteria for the aforementioned hires, A computer having, A proposal step in which the basic information storage unit is used to propose candidates who match the job seeker's preferences, A selection request reception step in which the applicant's terminal receives the applicant's request to be considered for selection by the proposed employer, An evaluation step in which the aforementioned job seeker evaluates whether or not they meet the selection criteria of the aforementioned employee they wish to be selected for, An evaluation transmission step of transmitting the evaluation of the job seeker to the job seeker terminal, The interview intention receiving step includes receiving an interview intention from the job seeker's terminal with the person who has been determined to meet the selection criteria, The interview request transmission step involves sending an interview request to the aforementioned recruiter terminal, along with the applicant's information, who has indicated an interest in an interview. Execute Recruitment support methods.

5. The job seeker's terminal, used by the job seeker, and the recruiter's terminal, are configured to communicate with each other. A job seeker information storage unit that stores the information of the aforementioned job seeker, A basic information storage unit that stores the basic information of the aforementioned recruit, An evaluation criteria storage unit that stores the evaluation criteria for the aforementioned hires, For a computer having, A proposal command is issued that, by referring to the basic information storage unit, proposes the candidate who matches the job seeker's preferences, A selection request reception command is received from the job seeker's terminal to receive a request for selection by the proposed recruiter, An evaluation command is given to the applicant to evaluate whether they meet the selection criteria of the person they wish to hire, An evaluation transmission command to transmit the evaluation of the job seeker to the job seeker terminal, An interview intention reception command is received from the job seeker's terminal to receive an interview intention from the person who has been determined to meet the selection criteria, A command to transmit an interview request to the recruiter's terminal, along with the applicant information of the applicant who has indicated their willingness to be interviewed, and a request for an interview with the applicant, To execute Recruitment support program.

Citation Information

Patent Citations

  • Interview device, interview method, and program

    JP2025015182A