JP7462103B1 - Recruitment support system, recruitment support method and program - Google Patents

Abstract

[Problem] To provide a hiring support system capable of preparing an effective job posting. [Solution] According to one aspect of the present invention, a hiring support system is provided. The hiring support system includes a processor. The processor is configured to be able to execute the following steps. In a first reception step, an input of a job posting is accepted. In a first display control step, information on a plurality of virtual job seekers corresponding to the job posting and characteristics corresponding to the job posting are displayed based on the contents of the job posting. In a second reception step, corrections to the characteristics are accepted. In a correction presentation step, corrections to the job posting based on the corrected characteristics are presented. [Selected Figure] FIG.

Description

The present invention relates to a recruitment support system, a recruitment support method, and a program.

As disclosed in Patent Document 1, a technology is known that supports document improvement by evaluating specific documents based on specific rules.

JP 2018-045690 A

The job posting, which is a document prepared by an employer when hiring a job seeker, must contain content that matches the type of person they are looking for. If there is a discrepancy between the type of person the employer is looking for and the content of the job posting, it may become difficult to hire the person.

In consideration of the above circumstances, the present invention aims to provide a recruitment support system that can prepare effective job postings.

According to one aspect of the present invention, a hiring support system is provided. The hiring support system includes a processor. The processor is configured to execute the following steps: In a first reception step, input of a job posting is received. In a first display control step, information on a plurality of virtual job seekers corresponding to the job posting and characteristics corresponding to the job posting are displayed based on the contents of the job posting. In a second reception step, corrections to the characteristics are received. In a correction presentation step, corrections to the job posting based on the corrected characteristics are presented.

According to this embodiment, the characteristics of the job seeker desired by the employer can be adjusted based on the information of the virtual job seeker who is the subject of the job posting. Therefore, an effective job posting can be created so that the employer can easily hire the job seeker they are looking for.

1 is a configuration diagram showing a recruitment support system 1. FIG. 2 is a block diagram showing the hardware configuration of the server device 10. FIG. 2 is a block diagram showing the hardware configuration of the recruiter terminal 20. A block diagram showing functions realized by the server device 10 (control unit 11) and the recruiter terminal 20 (control unit 21). FIG. 13 is a diagram showing an example of an initial condition input screen ID. A figure showing an example of an initial condition input screen ID on which a recruitment background selection area BF1 is displayed. FIG. 13 is a diagram showing an example of a job posting JP. FIG. 13 is a diagram showing an example of a job posting list screen JL. 13 is an example of a job seeker display screen JD. 13 is an example of a correction content display screen MD. This is an activity diagram showing the flow of information processing (job posting correction processing) performed by the recruitment support system 1. FIG. 11 is a block diagram showing functions realized by the server device 10 (control unit 11) in another embodiment.

The following describes embodiments of the present invention with reference to the drawings. The various features shown in the following embodiments can be combined with each other.

The program for implementing the software used in this embodiment may be provided as a non-transitory computer-readable recording medium, or may be provided so that it can be downloaded from an external server, or may be provided so that the program is started on an external computer and its functions are implemented on a client terminal (so-called cloud computing).

In this embodiment, a "unit" may also include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In addition, this embodiment handles various types of information, which may be represented, for example, by physical values of signal values representing voltage and current, high and low signal values as a binary bit collection consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculations may be performed on a circuit in the broad sense.

In the broad sense, a circuit is a circuit realized by at least appropriately combining a circuit, circuitry, a processor, and memory. In other words, it includes application specific integrated circuits (ASICs), programmable logic devices (e.g., simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs)), etc.

1. Hardware Configuration This section describes the hardware configuration.

<Recruitment Support System 1>
1 is a configuration diagram showing a hiring support system 1. The hiring support system 1 comprises a communication line 2, a server device 10, and a plurality of recruiter terminals 20. The server device 10 and the recruiter terminals 20 are configured to be able to communicate with each other via the communication line 2. The connection between the server device 10 and the recruiter terminals 20 may be wired or wireless.

The hiring support system 1 constitutes part of a recruitment and job search system used by multiple recruiters (a first recruiter U1 and a second recruiter U2). The hiring support system 1 is primarily responsible for creating and modifying job postings. In one embodiment, the hiring support system 1 is made up of one or more devices or components. These components are described below.

<Server device 10>
2 is a block diagram showing a hardware configuration of the server device 10. The server device 10 includes a control unit 11, a storage unit 12, a communication unit 13, and a communication bus 14. The control unit 11, the storage unit 12, and the communication unit 13 are electrically connected to each other inside the server device 10 via the communication bus 14.

<Control unit 11>
The control unit 11 processes and controls the overall operation related to the server device 10. The control unit 11 is, for example, a central processing unit (CPU). The control unit 11 realizes various functions related to the server device 10 by reading out a predetermined program stored in the storage unit 12. That is, information processing by software stored in the storage unit 12 can be specifically realized by the control unit 11, which is an example of hardware, and executed as each functional unit included in the control unit 11. These will be described in more detail in the next section. Note that the control unit 11 is not limited to being single, and may be implemented to have multiple control units 11 for each function. Also, a combination of these may be used.

<Storage unit 12>
The storage unit 12 stores various information defined by the above description. This can be implemented, for example, as a storage device such as a solid state drive (SSD) that stores various programs and the like related to the server device 10 executed by the control unit 11, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the program calculation. The storage unit 12 stores various programs, variables, etc. related to the server device 10 executed by the control unit 11. The storage unit 12 may also be an external storage device connected to the server device 10.

<Communication unit 13>
The communication unit 13 is preferably a wired communication means such as USB, IEEE 1394, Thunderbolt (registered trademark), wired LAN network communication, etc., but may also include wireless LAN network communication, mobile communication such as 3G/LTE/5G, BLUETOOTH (registered trademark) communication, etc. as necessary. In other words, it is more preferable to implement it as a collection of multiple communication means. In other words, the server device 10 may communicate various information from the outside via the communication unit 13 and the network.

The server device 10 may be in an on-premise form or in a cloud form. As a cloud-based server device 10, for example, the above-mentioned functions and processes may be provided in the form of SaaS (Software as a Service) or cloud computing. In addition, the server device 10 may be a multi-computer including multiple computers, or may be a virtual machine virtually constructed by software.

<Employer terminal 20>
Fig. 3 is a block diagram showing the hardware configuration of the recruiting party terminal 20. As shown in Fig. 3, the recruiting party terminal 20 comprises a control unit 21, a memory unit 22, a communication unit 23, an input unit 24, an output unit 25, and a communication bus 26. The control unit 21, the memory unit 22, the communication unit 23, the input unit 24, and the output unit 25 are electrically connected inside the recruiting party terminal 20 via the communication bus 26. The explanation of the control unit 21, the memory unit 22, and the communication unit 23 is omitted here because it is the same as the explanation of each unit in the server device 10. Note that the recruiting party terminal 20 may be a terminal operated by a human resources agency that interacts with job seekers on behalf of the recruiter.

<Input unit 24>
The input unit 24 accepts operation inputs made by the user. The operation inputs are transferred as command signals to the control unit 21 via the communication bus 26. The control unit 21 may execute predetermined control or calculations based on the transferred command signals as necessary. The input unit 24 may be included in the housing of the recruiter terminal 20, or may be attached externally. For example, the input unit 24 may be implemented as a touch panel integrated with the output unit 25. When the input unit 24 is implemented as a touch panel, the user can input tap operations, swipe operations, and the like to the input unit 24. As the input unit 24, a switch button, a mouse, a track pad, a QWERTY keyboard, and the like can be adopted instead of a touch panel.

<Output unit 25>
The output unit 25 displays a screen of a graphical user interface (GUI) that can be operated by the user. The output unit 25 may be included in the housing of the recruiter terminal 20, or may be attached externally. Specifically, the output unit 25 may be implemented as a display device such as a CRT display, a liquid crystal display, an organic EL display, or a plasma display. It is preferable that these display devices are implemented by using different devices depending on the type of recruiter terminal 20.

2. Functional Configuration In this section, the functional configuration of the present embodiment will be described. Information processing by the software stored in the storage unit 12 is specifically realized by the control unit 11, which is an example of hardware, and can be executed as each functional unit included in the control unit 11 (a processor included in the recruitment support system 1).

Figure 4 is a block diagram showing the functions realized by the server device 10 (control unit 11) and the recruiter terminal 20 (control unit 21).

As shown in FIG. 4A, the server device 10 (control unit 11) includes a basic display control unit 111, a job posting creation unit 112, a job posting evaluation unit 113, a first reception unit 114, a first display control unit 115, a second reception unit 116, a second display control unit 117, a correction presentation unit 118, a third reception unit 119, and an artificial intelligence unit 120. As shown in FIG. 4B, the recruiter terminal 20 (control unit 21) includes a display unit 211 and an operation reception unit 212.

<Basic display control unit 111>
The basic display control unit 111 is configured to display various information on the recruiting party terminal 20. For example, the basic display control unit 111 displays, on the display unit 211 of the recruiting party terminal 20, resumes and curriculum vitae prepared by the job seeker, job advertisements and scouting documents prepared by the recruiter, corrections made by the correction presentation unit 118, and the like.

Employers include organizations such as for-profit corporations (e.g., companies), non-profit organizations (e.g., cooperatives, foundations), and public corporations (e.g., local governments). Employers also include recruitment agencies that act as agents of organizations to mediate between job seekers and organizations. Recruitment agencies are also called headhunters, agents, etc.

<Job posting creation unit 112>
The job posting creation unit 112 is configured to create a job posting based on input of initial conditions for creating a job posting from the recruiter terminal 20. The initial conditions include at least one of a persona and a recruitment background. The persona is the type of person (what kind of human resources do they want to hire) that the recruiter is looking for (the target of recruitment), and includes, for example, information on the job (position). The recruitment background is the reason for the recruitment of the human resources or position, and includes the organization's goals, issues, business situation (for example, the start of a new project), etc. The initial conditions may also be the type of job, position, skill, products handled, etc., as long as they are information that defines the type of person that the recruiter is looking for.

FIG. 5 is a diagram showing an example of an initial condition input screen ID. As shown in FIG. 5A, the initial condition input screen ID includes a persona input field PF and a first button B1. The persona input field PF accepts input of a persona (a sentence showing the content of the persona) from the user. When the user inputs all or part of the sentence showing the content of the persona, as shown in FIG. 5B, a list of personas related to the sentence showing the content of the input persona is displayed from among the personas prepared in advance. When the user selects a displayed persona, the selected persona is input into the persona input field PF. When the first button B1 is input (pressed) with a persona input into the persona input field PF, the job posting creation unit 112 accepts the persona input into the persona input field PF as the persona of the initial condition. The sentence showing the content of the persona may be input by the user selecting from a list of prepared sentence options displayed without the user directly inputting it. In addition, as options for sentences indicating the content of the persona, those related to the user's information previously registered by the user may be extracted from the sentences indicating the content of the persona prepared in advance and displayed.

Figure 6 is a diagram showing an example of an initial condition input screen ID in which a recruitment background selection area BF1 is displayed. When the first button B1 shown in Figure 5A is input on the initial condition input screen ID, the recruitment background selection area BF1, the recruitment background display area BF2, and the second button B2 shown in Figure 6 are displayed. In the recruitment background selection area BF1, a selection of recruitment backgrounds related to the persona whose input has been accepted is displayed. Note that, not limited to the persona whose input has been accepted, a selection of recruitment backgrounds related to other information input by the user may be displayed. In the recruitment background display area BF2, a recruitment background selected by the user from the selections is displayed. When at least one recruitment background is selected and the second button B2 is input, a recruitment board is created by the first job board creation model of the artificial intelligence unit 120. Note that, in response to input of all or part of the text indicating the content of the recruitment background by the user, a recruitment background related to the input text may be displayed as a selection from among the recruitment backgrounds prepared in advance.

The first job posting creation model is a learning model trained to input initial conditions and output a job posting. In other words, the first job posting creation model is a learning model trained using the initial conditions and the corresponding job posting data as training data. The first job posting creation model may also be a generative AI including a large-scale language model. In this case, the first display control unit 115 inputs a prompt into the first job posting creation model that inserts the job posting creation instruction and the initial conditions input by the employer, and causes the first job posting creation model to output the job posting. The first display control unit 115 may also input a prompt into the first job posting creation model that inserts, for example, one or more samples of the initial conditions and one or more corresponding job posting samples in addition to the job posting creation instruction and the initial conditions.

Figure 7 is a diagram showing an example of a job posting JP. The job posting JP includes multiple items such as the name of the position being advertised, the job content/working conditions, application qualifications, and selling points. The job posting JP may also include other items. The first job posting creation model generates text for these items based on the initial conditions. The job posting created by the job posting creation unit 112 is added to the database stored in the memory unit 12.

<Job posting evaluation unit 113>
The job posting evaluation unit 113 is configured to evaluate the job posting. Specifically, for example, the job posting evaluation unit 113 outputs a score for each of two or more evaluation elements for a job posting to be evaluated among the job postings included in the database. The evaluation elements may include one or more elements selected from the amount of information included in the job posting and the expression of the sentences included in the job posting. The evaluation elements may include an element calculated based on whether the job posting contains a term that represents the job content of the human resources desired by the employer.

The job posting evaluation unit 113 calculates a score for the amount of information contained in the job posting based on the amount of information contained in the items that make up the job posting. For example, a score for an item may be calculated based on a comparison between the number of characters contained in an item and a preset range of appropriate character numbers for each item, and a score for the amount of information in the entire job posting may be calculated based on the score for each item. In addition, the job posting evaluation unit 113 calculates a score for the expression of the text contained in the job posting based on the consistency between the items that make up the job posting and the text contained in these items. For example, if the "job description" item contains text other than the job description (such as content related to application qualifications), the score for the expression may be calculated to be low.

The job posting evaluation unit 113 may calculate a score for the expression of the sentences included in the job posting based on the consistency between the sentences included in each of two or more items that make up the job posting. For example, if a sentence included in the "job description" item is inconsistent with a sentence included in the "selling points" item, a low score for the expression may be calculated. Furthermore, the job posting evaluation unit 113 may calculate a score for the expertise of the personnel desired by the employer by referring to a database in which terms that represent the job description of the personnel desired by the employer are registered. Such analysis of sentences and calculation of scores may be performed by natural language processing, including morphological analysis, using artificial intelligence.

The job posting evaluation unit 113 determines whether or not the job posting needs to be revised based on the representative value of each score described above. Examples of the representative value of each score include the smallest of the scores, the sum of the scores, and the average value of the scores. The job posting evaluation unit 113 assigns a graded evaluation to the job posting, such as "no revision required," "requires consideration," or "revision required," depending on the magnitude of the representative value of each score. The job posting evaluation unit 113 may also determine whether or not the job posting needs to be revised based on a single score. Note that the job posting evaluation unit 113 may evaluate the job posting by known methods, not limited to the above-mentioned methods.

<First Reception Unit 114>
The first reception unit 114 is configured to receive an input of a job posting to be modified from the recruiter terminal 20. The job posting to be modified may be one created by the job posting creation unit 112, or one created outside the recruiting support system 1 and imported into the recruiting support system 1.

Figure 8 is a diagram showing an example of a job posting list screen JL. The list screen JL displays a number of job postings stored in the memory unit 12 in list format. The list screen JL displays the identification code JS, job posting name JN, persona name PN, job posting evaluation JV, last editor FE, creation date GD, and last edit date FD of each job posting. The identification code JS is the ID of each job posting. The job posting name JN is the name of the position in the job posting, and is a sentence indicating the content of the persona entered or selected as the initial condition in the job posting creation unit 112. The job posting evaluation JV is the evaluation result by the job posting evaluation unit 113. The user selects the job posting they wish to modify from the list screen JL.

<First display control unit 115>
The first display control unit 115 is configured to display information on a plurality of virtual job seekers corresponding to a job advertisement and job seeker characteristics corresponding to the job advertisement, based on the contents of the job advertisement selected by the employer terminal 20. Here, a "virtual job seeker" is a virtual person who may be the target of recruitment in the job advertisement. A virtual job seeker may be rephrased as a virtual person who matches the contents of the job advertisement. "Information on a virtual job seeker" includes information corresponding to job seeker information registered by an actual job seeker, such as the company the virtual job seeker works for, the department to which the virtual job seeker belongs, annual income, age, additional information, etc. Additional information includes career history, experience, skills, qualifications, etc.

The information on the virtual job seeker may be generated based on the information on a job seeker registered in advance. In other words, the information on the virtual job seeker may be generated by using a part of the information (job seeker information) registered by an actual job seeker in the recruitment support system 1. The first display control unit 115 extracts and displays the information on the virtual job seeker corresponding to the job posting from the information on the multiple virtual job seekers generated in this way. This increases the realism of the information on the virtual job seeker, so that the employer can select information on the virtual job seeker that is close to the information on the actual job seeker, making it easier to imagine the job seeker they are looking for. In addition, it is possible to avoid presenting information on a virtual job seeker that does not actually exist. Note that the "job seeker information" is registered information on the job seeker, and is, for example, information included in the job seeker's resume, career history, or other profile information. A "resume" is a document that mainly describes the job seeker's profile, current situation, educational background, employment history, desired working conditions, etc., and a "career history" is also called a resume, and is a document in which the job seeker conveys to the employer his or her career history, experience, skills, qualifications, etc. related to his or her previous work. The job seeker information may also include conditions such as the industry and job type the job seeker desires.

The information on the virtual job seeker may be generated by combining the information on multiple job seekers that have been registered in advance. This allows a wide variety of realistic virtual job seekers to be prepared, allowing employers to select information on a virtual job seeker that is closest to the person they are looking for and making it easier for them to imagine the job seeker they are looking for. In addition, the information on the virtual job seeker may be generated, for example, by vectorizing the information on multiple job seekers extracted from a database in which information on actual job seekers is registered using known methods such as natural language processing using morphological analysis or encoding of categorical variables, generating vectors indicating the information on each job seeker, and combining and combining these vectors. This ensures the reality of the virtual job seeker while reliably avoiding the information on actual job seekers from being displayed as is.

The virtual job seeker information may be generated by modifying or deleting part of the information of a pre-registered job seeker. This also makes it possible to prepare a wide variety of realistic virtual job seekers, allowing employers to select virtual job seeker information that is closest to the desired personality and making it easier to imagine the desired job seeker. This ensures the reality of the virtual job seeker while reliably avoiding the information of an actual job seeker being displayed as is. For example, the virtual job seeker information may be information generated by vector synthesis as described above, with part of the information hidden by a mask.

For example, the information of a plurality of actual job seekers is input to the virtual job seeker creation model of the artificial intelligence unit 120, and the virtual job seeker creation model is made to output the information of the virtual job seeker. The virtual job seeker output model is a learning model that is trained to input information of an actual job seeker and output a virtual job seeker. In other words, the virtual job seeker output model is a learning model that is trained using information of an actual job seeker and data of a corresponding virtual job seeker as training data. The virtual job seeker output model may also be a generative AI including a large-scale language model. In this case, a prompt that inserts an instruction to create a virtual job seeker and information of the actual job seeker is input to the virtual job seeker output model, and the virtual job seeker is output to the virtual job seeker creation model. In addition to the output instruction of the virtual job seeker and the information of the actual job seeker, for example, a prompt that inserts a sample of information of one or more job seekers and a sample of one or more virtual job seekers corresponding to the sample may be input to the virtual job seeker output model.

"Job seeker characteristics" include at least one keyword or parameter condition related to the contents of the job posting. Job seeker characteristics include, for example, keywords that represent the career, experience, skills, qualifications, etc., described in the job posting, or keywords similar to the career, experience, skills, qualifications, etc., described in the job posting. Job seeker characteristics also include, for example, parameter conditions that are ranges of vector values quantified by vectorization or the like of items (e.g., experience, skills, etc.) described in the job posting. A parameter may be expressed as the sum of vector values of multiple items. In this way, job seeker characteristics relate to the skills or experience of a virtual job seeker. This makes it possible to create a job posting that places emphasis on skills or experience.

The first display control unit 115 may extract virtual job seekers and job seeker characteristics for which information is to be displayed based on the contents of the job advertisement. In other words, the first display control unit 115 may extract virtual job seekers having information related to the job advertisement, and may also extract job seeker characteristics from the description of the job advertisement. This makes it possible to present to the employer the virtual job seekers and job seeker characteristics derived from the job advertisement.

The first display control unit 115, for example, inputs a job posting and a database of virtual job seekers into the first virtual job seeker extraction model of the artificial intelligence unit 120, and causes the first virtual job seeker extraction model to output a virtual job seeker that matches the content of the job posting. The first virtual job seeker extraction model is a learning model that is trained to input a job posting and a database of virtual job seekers and output a virtual job seeker that matches the content of the job posting. In other words, the first virtual job seeker extraction model is a learning model that is trained using a job posting and data on a corresponding virtual job seeker as training data. Here, the virtual job seeker corresponding to the job posting may be a virtual job seeker that matches all or part of the content or conditions of the job posting, or a virtual job seeker who may be a target of the job posting. The first virtual job seeker extraction model may also be a generation AI including a large-scale language model. In this case, the first display control unit 115 inputs a prompt into the first virtual job seeker extraction model, which contains an instruction to extract a virtual job seeker matching the contents of the job posting by referring to the virtual job seeker database, and the job posting, and outputs the virtual job seeker to be recruited to the first virtual job seeker extraction model. Furthermore, the first display control unit 115 may input a prompt into the first virtual job seeker extraction model, which contains, for example, one or more sample job postings and one or more corresponding samples of virtual job seekers, in addition to the instruction to extract a virtual job seeker and the job posting.

The first display control unit 115, for example, inputs a job posting into the first feature extraction model of the artificial intelligence unit 120, and causes the unit 120 to output job seeker characteristics. The first feature extraction model is a learning model trained to input a job posting and output job seeker characteristics. In other words, the first feature extraction model is a learning model trained using a job posting and data on the corresponding job seeker characteristics as teacher data. The first feature extraction model may also be a generative AI including a large-scale language model. In this case, the first display control unit 115 inputs a prompt into which an instruction to extract job seeker characteristics and the job posting are inserted into the first feature extraction model, and causes the first feature extraction model to output the job seeker characteristics to be recruited. The first display control unit 115 may also input, in addition to an instruction to extract job seeker characteristics and the job posting, a prompt into which, for example, one or more samples of job postings and one or more samples of the corresponding job seeker characteristics are inserted into the first feature extraction model.

The first display control unit 115 may extract a virtual job seeker whose information is to be displayed based on the contents of the job advertisement, and may further extract job seeker characteristics based on the information of the extracted virtual job seeker. In other words, the first display control unit 115 may first extract a plurality of virtual job seekers who have information related to the job advertisement, and then extract keywords and characteristics common to the information of the extracted virtual job seekers as job seeker characteristics. This makes it possible to present to the employer job seeker job seeker characteristics that are highly correlated with the displayed virtual job seeker.

The first display control unit 115, for example, inputs a plurality of virtual job seekers into the second feature extraction model of the artificial intelligence unit 120, and causes the unit to output job seeker characteristics. The second feature extraction model is a learning model trained to input a plurality of virtual job seekers and output job seeker characteristics. In other words, the second feature extraction model is a learning model trained using virtual job seekers and data on the corresponding job seeker characteristics as teacher data. The second feature extraction model may also be a generative AI including a large-scale language model. In this case, the first display control unit 115 inputs a prompt into which an instruction to extract job seeker characteristics and a plurality of virtual job seekers are inserted into the second feature extraction model, and causes the second feature extraction model to output the virtual job seekers to be recruited. The first display control unit 115 may also input a prompt into which, in addition to an instruction to extract job seeker characteristics and the virtual job seekers, for example, one or more samples of virtual job seekers and one or more samples of the corresponding job seeker characteristics are inserted into the second feature extraction model.

Figure 9 is an example of a job seeker display screen JD. The job seeker display screen JD includes a job seeker characteristics display area CF and a job seeker display area JF. The job seeker characteristics display area CF includes a tag adjustment area TC, a recommended tag display area RT, and a parameter adjustment area PC. The job seeker display area JF includes an annual income histogram IH and a virtual job seeker list VL.

The tag adjustment area TC displays tags including tag words related to job seeker characteristics. A tag word is a word written in a tag (i.e., a word corresponding to a tag), and is typically a keyword indicating career, experience, skills, qualifications, etc., included in the job seeker characteristics. The tag adjustment area TC displays tags representing the current job seeker characteristics (i.e., tags selected as job seeker characteristics), a move button BM for changing the order of tags, a delete button BD for deleting tags, and an add button BA for adding tags. The recommended tag display area RT displays recommended tags presented by the second reception unit 116.

In the parameter adjustment area PC, the parameter names and their ranges set as part of the job seeker characteristics are displayed as a slider-type UI. The initial values of the parameter conditions may be extracted from the information of the virtual job seeker displayed in the job posting or the virtual job seeker list VL.

The annual income histogram IH is a graph that displays the distribution of annual income. The annual income histogram IH may be the distribution of the annual income of virtual job seekers included in the virtual job seeker list VL, or, if a virtual job seeker matching the contents of the job advertisement is stored in the database other than the virtual job seeker included in the virtual job seeker list VL displayed on the job seeker display screen JD, the annual income histogram IH may be the distribution of the annual income of virtual job seekers including the virtual job seeker (the virtual job seeker not displayed in the virtual job seeker list VL). The annual income histogram IH may also be the distribution of the annual income of real job seekers who match the contents of the job advertisement. In this case, the recruitment support system 1 stores the job seeker information registered by the job seeker himself in the job seeker information database.

The information of the virtual job seeker displayed in the virtual job seeker list VL includes employment information I1, annual income I2, age I3, tags I4, summary text I5, etc. Employment information I1 indicates the company and department (team) to which the virtual job seeker is currently employed. The company to which the virtual job seeker is employed may be a real company name, or a company name extracted from actual job seeker information. This makes it easier for employers to grasp the personality of the virtual job seeker, such as their skills and experience. Annual income I2 indicates the current annual income of the virtual job seeker. Age I3 indicates the current age of the virtual job seeker. Age I3 may be calculated based on actual job seeker information, and may be, for example, the average age of multiple job seekers.

Tag I4 indicates a tag attached to a virtual job seeker. Tag I4 is attached according to the career, experience, skills, qualifications, etc. of the virtual job seeker. In other words, tag I4 indicates the career, experience, skills, qualifications, etc. of the virtual job seeker as keywords (tag words). The first display control unit 115 may further extract and display virtual job seekers to which a certain number of tags are attached from among the virtual job seekers extracted based on the job posting. The first display control unit 115 may also extract all virtual job seekers to which tags extracted based on the job posting are attached and display them as a virtual job seeker list VL. Note that tag I4 may be in the form of a tag in which text is surrounded by a frame, and the experience, skills, qualifications, etc. of the virtual job seeker may be displayed only in text.

Summary text I5 is a text summarizing the career, work history, experience, skills, qualifications, etc. of the virtual job seeker. Summary text I5 is created by the first display control unit 115 inputting the attributes of the virtual job seeker into the writing model of the artificial intelligence unit 120, causing the writing model to create it. "Virtual job seeker attributes" are information about the virtual job seeker, and include the organizations the virtual job seeker has worked in, the types of jobs the virtual job seeker has worked in, the skills and qualifications the virtual job seeker possesses, the annual income, and the age. This makes it easier for employers to understand the personality of the virtual job seeker. The summary text may also include the names of organizations (organizations of experience) such as companies to which the virtual job seeker has previously belonged. This allows the summary text to be a text that uniquely expresses the career, etc., which are attributes of the virtual job seeker.

The writing model is a learning model trained to input the attributes of the virtual job seeker and output a summary sentence. In other words, the writing model is a learning model trained using the attributes of the virtual job seeker and the corresponding summary sentence as teacher data. The writing model may also be a generation AI including a large-scale language model. In this case, the first display control unit 115 inputs a prompt into which an instruction to create a summary sentence and the attributes of the virtual job seeker are inserted into the writing model, and causes the writing model to output the summary sentence. The first display control unit 115 may also input a prompt into which, in addition to the instruction to create a summary sentence and the attributes of the virtual job seeker, for example, one or more samples of the attributes of the virtual job seeker and one or more samples of the corresponding summary sentence are inserted into the writing model. A writing model trained to output a summary sentence including the name of the organization where the virtual job seeker has experience may also be used. Alternatively, such a summary sentence may be output by inputting a prompt into which an instruction to create a summary sentence including the name of the organization where the virtual job seeker has experience is inserted into the generation AI including a large-scale language model.

Summary text I5 may also be extracted from the resume or work history of an actual job seeker that has attributes that overlap with those of the virtual job seeker. Summary text I5 may also be extracted from the resume or work history of an actual job seeker that was used to create the virtual job seeker.

The first display control unit 115 may receive a selection of a virtual job seeker displayed in the virtual job seeker list VL from the employer terminal 20 and display detailed information of the selected virtual job seeker (information not displayed in the virtual job seeker list VL).

<Second Reception Unit 116>
The second reception unit 116 is configured to receive modifications to the job seeker characteristics. Modifications to the job seeker characteristics include, for example, adding or deleting keywords included in the job seeker characteristics, modifying the range of parameter conditions, etc.

Specifically, the second reception unit 116 accepts the modification of the job seeker characteristics by adding, deleting, or moving tags corresponding to the characteristics prepared in advance. This allows the employer to intuitively and easily modify the job seeker characteristics in line with the image of the person he or she desires. In addition, while viewing and checking the information of the virtual job seeker, the employer can add tags of job seeker characteristics that match the image of the person the employer wants to hire, delete tags of job seeker characteristics that do not match the image of the person the employer wants to hire, and move the tags in an order that matches the image, so that the employer can clarify the image of the person he or she desires, and can select and move tags accordingly. The tags for modifying the job seeker characteristics correspond to keywords that represent the career, experience, skills, qualifications, etc. described in the job posting. These tags may be generated by extracting keywords that represent experience, skills, qualifications, etc. as tag words from the job posting or the information of the virtual job seeker, or may be generated by preparing a tag list in which multiple tags are described in advance and extracting tags that correspond to the information of the job posting or the virtual job seeker from the tag list. Here, a tag list is a list containing one or more tags, and a tag word is a word that is written in a tag (i.e., that corresponds to a tag). Thus, one or more tag words are written in a tag list. The tags (which are part of the information about the virtual job seeker) that are assigned to the virtual job seeker are also selected from the tag list. Here, the job seeker characteristics do not necessarily need to be displayed in the form of tags, but may be displayed in a form (object) that allows the user to perform operations such as selection and movement, such as buttons or labels. For example, the job seeker characteristics may be in the form of blocks of text that display words representing career, experience, skills, qualifications, etc., and that allow selection and movement.

Tags related to job seeker characteristics are not limited to exact matches between words contained in the job posting or virtual job seeker information and tag words, but may be partial matches, or words contained in the job posting or virtual job seeker information may be synonyms, similar words, or variations of the tag words.

9, tags related to job seeker characteristics are displayed in the tag adjustment area TC of the job seeker display screen JD. When the add button BA is input, the second reception unit 116 displays a search field for tags to be added or a list of selectable tags. The searched or selected tags are added to the tag adjustment area TC. When the delete button BD attached to each tag is input, the tag is deleted from the tag adjustment area TC. Furthermore, the order of tags is changed by dragging the tag with the move button BM attached to each tag selected and moving it to any position. In the job seeker characteristics, the higher the tag is positioned, the higher the weight is given in the search for virtual job seekers and the presentation of correction contents of the job posting, which will be described later. In other words, the higher the tag is positioned, the closer the job seeker characteristics are to the image of the person desired by the employer. The memory unit 12 also stores the correction history of the job seeker characteristics in the second reception unit 116 (specifically, the history of adding, deleting, and moving tags by the user).

The second reception unit 116 presents, as recommended tags, tags that include tag words that are superordinate or subordinate conceptualizations of the tag words included in the selected tag displayed in the tag adjustment area TC. This allows adjustments to be made to broaden or narrow the range of virtual job seekers that correspond to the job seeker characteristics. The recommended tags are displayed in the recommended tag display area RT of the job seeker display screen JD. When a user selects a recommended tag, the selected recommended tag is added to the tag adjustment area TC. The relationship between tags is described, for example, in a tree structure. In the example of FIG. 9, for the tag "Java (registered trademark) engineer", "software engineer" is presented as a superordinate conceptualization tag, and "financial systems engineer using Java" is presented as a subordinate conceptualization tag. In addition, a tag that adds a different perspective to the content of the selected tag may be displayed as a recommended tag. For example, the recommended tag "engineering management" may be presented for the tag "Java engineer".

The second reception unit 116 may also present, as recommended tags, tags that are tag words that have a high degree of co-occurrence with the tag words of the selected tag. Here, the degree of co-occurrence refers to the frequency with which tags are used simultaneously in searches for job information or job seeker information, or the frequency with which tags are selected simultaneously by users in the recruitment support system 1. In other words, tags that are used frequently in searches for job information or job seeker information, or tags that are selected simultaneously by users, are presented as recommended tags.

Furthermore, the second reception unit 116 may present, in a form other than tags, words that are superordinate or subordinate conceptualizations of words included in the job seeker characteristics, words that specify words included in the job seeker characteristics, or words that add a different perspective, as recommended words to add to the job seeker characteristics.

The job posting, which is the data to be corrected, and the job seeker information, which is the source data for correcting the job posting, are different types of textual data. In this way, by defining the requirements for the personnel you want to recruit through the job posting with tags while looking at the virtual job seeker information, you can set the conditions for correcting the job posting using tags, thereby realizing corrections based on different types of textual data. Furthermore, by defining the conditions for the job posting by replacing them with tags, it is possible to make appropriate corrections to the job posting even if there is only a small amount of source job seeker information.

The second reception unit 116 accepts modifications to the job seeker characteristics by setting parameter conditions related to the information of the virtual job seeker. This makes it possible to quantitatively adjust the range of the virtual job seeker. Examples of parameter conditions include the amount of experience, the level of expertise, and annual income. Specifically, the parameter conditions are set by adjusting the positions of two knobs that set the lower and upper limits, respectively, arranged on the slider. This allows the adjustment of the parameter conditions to be performed visually, intuitively, and easily. The slider for setting the parameter conditions is displayed in the parameter adjustment area PC of the job seeker display screen JD. This makes it possible to set the level of the job seeker characteristics selected by the tag using the slider. For example, after setting the tag "Java engineer," it is possible to set whether the person has a lot of experience as a Java engineer (veteran or junior) and whether they are highly specialized (specialist or generalist) by using the slider to set the amount of experience and level of expertise.

The second reception unit 116 may also receive modifications to job seeker characteristics based on parameter conditions set by the annual income histogram IH. Specifically, by selecting one or more columns of each interval included in the annual income histogram IH, the parameter conditions related to annual income are set as the selected interval. This allows job seeker characteristics that are difficult to express using tags to be appropriately set as conditions.

<Second display control unit 117>
The second display control unit 117 is configured to display information on a virtual job seeker corresponding to the modified job seeker characteristics. This allows the employer to adjust the job seeker characteristics while checking the changes in the virtual job seeker due to the modification of the job seeker characteristics. Specifically, the second display control unit 117 searches for virtual job seekers registered in the database based on the tags (tags included in the job seeker characteristics) displayed in the tag adjustment area TC of the job seeker display screen JD and the parameter conditions (parameter conditions included in the job seeker characteristics) displayed in the parameter adjustment area PC.

The second display control unit 117 extracts virtual job seekers who are tagged with at least one tag that matches or is similar to the tag included in the job seeker characteristics and who satisfy the parameter conditions included in the job seeker characteristics. The second display control unit 117 may also extract virtual job seekers who are tagged with at least one tag that matches or is similar to the tag included in the job seeker characteristics but do not satisfy the parameter conditions included in the job seeker characteristics, and it is not necessary to extract only virtual job seekers who satisfy all conditions. Here, "similarity of tags" includes cases where tag words partially match, and cases where one tag word is a synonym, similar word, or word with a spelling variation of the other tag word. In addition, "similarity of tags" also includes cases where the vectors in natural language processing between tag words are close. The second display control unit 117 deletes the virtual job seekers that were originally displayed, and displays the extracted virtual job seekers in the virtual job seeker list VL of the job seeker display screen JD.

The second display control unit 117 determines the display order of the virtual job seekers in the virtual job seeker list VL based on the weighting scores (ranking in the tag adjustment area TC) of the tags included in the job seeker characteristics. Here, the weighting score assigned to each tag may be set to change depending on the arrangement of the tags in the tag adjustment area TC, and specifically, the weighting score may be set to be larger for tags arranged higher in the tag adjustment area TC. For example, the second display control unit 117 assigns the weighting score of the tag included in the job seeker characteristics as a priority display score to a virtual job seeker to which a tag corresponding to a tag included in the job seeker characteristics is attached. For a virtual job seeker to which multiple tags corresponding to each of the multiple tags included in the job seeker characteristics are attached, the sum of the weighting scores of the multiple tags is set as the priority display score. The second display control unit 117 displays multiple virtual job seekers so that virtual job seekers with high priority display scores are at the top (at the top of the virtual job seeker list VL).

The second display control unit 117 may adjust the priority display score of a virtual job seeker based on a tag, based on the revision history of the job seeker characteristics (the history of tag addition, deletion, and movement by the user). For example, the second display control unit 117 may deduct points from the priority display score of a virtual job seeker with a tag that the user deleted from the tag adjustment area TC. The second display control unit 117 may also add points to the priority display score of a virtual job seeker with a tag that the user added to the tag adjustment area TC, or a tag that the user moved up in rank.

<Correction suggestion unit 118>
The correction presentation unit 118 is configured to present the correction content of the job posting based on the corrected job seeker characteristics. That is, the correction presentation unit 118 corrects the job posting so that it includes a description related to the tag (keyword) and parameter condition included in the corrected job seeker characteristics, and presents the correction content to the employer terminal 20. In other words, the correction presentation unit 118 generates a job posting corresponding to a person who matches the corrected job seeker characteristics. That is, the correction presentation unit 118 corrects the job posting so that a virtual job seeker corresponding to the current job seeker characteristics (that is, a virtual job seeker displayed in the virtual job seeker list VL by the second display control unit 117) becomes a recruitment target.

For example, the correction presentation unit 118 corrects the job posting by correcting the initial conditions of the job posting. That is, the correction presentation unit 118 inputs the corrected job seeker characteristics and the initial conditions included in the job posting into the first job posting correction model of the artificial intelligence unit 120, causes the first job posting correction model to output corrected initial conditions in which the initial conditions are corrected to correspond to the job seeker characteristics, and inputs the corrected initial conditions into the first job posting creation model and causes the first job posting creation model to output the corrected job posting. The first job posting correction model is a learning model that is trained to input the job seeker characteristics and the initial conditions and output the corrected initial conditions. That is, the first job posting correction model is a learning model that is trained using the job seeker characteristics and the initial conditions, and the corresponding corrected initial conditions as teacher data. The first job posting correction model may be a generation AI including a large-scale language model. In this case, the correction submission unit 118 inputs a prompt into which an instruction to create the corrected initial conditions, the corrected job seeker characteristics, and the initial conditions are inserted, to the first job advertisement correction model, and outputs the corrected initial conditions to the first job advertisement correction model. The correction submission unit 118 may also input a prompt into which, in addition to the instruction to create the corrected initial conditions, the corrected job seeker characteristics, and the initial conditions, for example, one or more samples of job seeker characteristics, one or more samples of initial conditions, and one or more samples of corrected initial conditions corresponding to the samples, are inserted, to the first job advertisement correction model. This makes it possible to correct the job advertisement by using the first job advertisement creation model when creating the job advertisement.

Furthermore, the correction suggestion unit 118 may input the correction history to the first job advertisement correction model in addition to the job seeker characteristics and the job advertisement, and cause the first job advertisement correction model to output corrected initial conditions in which the initial conditions are corrected to correspond to the job seeker characteristics and the correction history. This makes it possible to predict the matters that the employer values from the correction history and reflect them in the correction of the job advertisement. In this case, the first job advertisement correction model is a learning model that is trained to input the job seeker characteristics, the correction history, and the initial conditions, and to output the corrected initial conditions. In other words, the first job advertisement correction model is a learning model that is trained to use the job seeker characteristics, the correction history, and the initial conditions, and the corresponding corrected initial conditions as teacher data. The first job advertisement correction model may be a generation AI including a large-scale language model. In this case, the correction suggestion unit 118 inputs an instruction to create the corrected initial conditions, and a prompt in which the corrected job seeker characteristics, the correction history, and the initial conditions are inserted, to the first job advertisement correction model, and causes the first job advertisement correction model to output the corrected initial conditions. Furthermore, the correction presentation unit 118 may input a prompt into the first job advertisement correction model that includes, in addition to an instruction to create the corrected initial conditions, the corrected job seeker characteristics, the correction history, and the initial conditions, for example, one or more sample job seeker characteristics, one or more sample initial conditions, one or more sample correction histories, and one or more sample corrected initial conditions corresponding thereto.

If the job posting does not include the initial conditions, that is, if the job posting is not created using the initial conditions and the first job posting creation model (if the job posting is imported from outside), the modification submission unit 118 inputs the job posting into the initial condition estimation model of the artificial intelligence unit 120, and causes the initial condition estimation model to estimate the initial conditions corresponding to the job posting. The modification submission unit 118 then inputs the modified job seeker characteristics and the estimated initial conditions into the first job posting modification model, and causes the first job posting modification model to output modified initial conditions in which the initial conditions are modified to correspond to the job seeker characteristics, and inputs the modified initial conditions into the first job posting creation model, and causes the first job posting creation model to output the modified job posting. This makes it possible to modify a job posting that does not include initial conditions by using the first job posting creation model when creating the job posting, by calculating backwards and estimating the initial conditions based on the contents of the job posting. The correction submission unit 118 may also input the estimated initial conditions and the original job posting before correction to the first job posting creation model, and output a corrected job posting based on the estimated initial conditions and the job posting before correction.

For a job posting that does not include initial conditions, the correction suggestion unit 118 may accept input of initial conditions instead of estimating the initial conditions. For example, when the job posting does not include initial conditions, the correction suggestion unit 118 may display the initial condition input screen ID of Figures 5 and 6 and have the user input the initial conditions.

The modification presentation unit 118 may directly modify the job posting instead of modifying the initial conditions of the job posting. That is, the modification presentation unit 118 may input the modified job posting and the job posting to the second job posting modification model of the artificial intelligence unit 120, and cause the second job posting modification model to output a modified job posting in which the job posting is modified to correspond to the job posting characteristics. The second job posting modification model is a learning model that is trained to input the job posting characteristics and the job posting and output the modified job posting. That is, the second job posting modification model is a learning model that is trained to use the job posting characteristics, the original job posting, and the corresponding modified job posting as training data. The second job posting modification model may be a generation AI including a large-scale language model. The modification presentation unit 118 inputs a prompt into which an instruction to create a modified job posting, the modified job posting characteristics, and the job posting are inserted, into the second job posting modification model, and causes the second job posting modification model to output the modified job posting. Furthermore, the correction submission unit 118 may input a prompt that includes, in addition to the instruction to create a corrected job posting, the corrected job seeker characteristics, and the job posting, for example, one or more sample job seeker characteristics, one or more sample job postings, and one or more corresponding samples of corrected job postings, to the second job posting correction model. This makes it possible to correct a job posting in a system that does not have the first job posting creation model.

Furthermore, the correction presentation unit 118 may input the correction history to the second job advertisement correction model in addition to the job seeker characteristics and the job advertisement, and cause the second job advertisement correction model to output a corrected job advertisement in which the job advertisement is corrected to correspond to the job seeker characteristics and the correction history. This allows the items that the employer values, which are inferred from the correction history, to be reflected in the correction of the job advertisement. In this case, the second job advertisement correction model is a learning model that is trained to input the job seeker characteristics, the correction history, and the job advertisement, and to output the corrected job advertisement. In other words, the second job advertisement correction model is a learning model that is trained using the job seeker characteristics, the correction history, and the job advertisement, and the corresponding corrected job advertisement as training data. The second job advertisement correction model may be a generation AI including a large-scale language model. The correction presentation unit 118 inputs an instruction to create a corrected job advertisement, and a prompt in which the corrected job seeker characteristics, the correction history, and the job advertisement are inserted, to the second job advertisement correction model, and causes the second job advertisement correction model to output the corrected job advertisement. Furthermore, the correction submission unit 118 may input a prompt into the second job posting correction model that, in addition to an instruction to create a corrected job posting, the corrected job seeker characteristics, the correction history, and the job posting, includes, for example, one or more sample job seeker characteristics, one or more correction histories, one or more sample job postings, and one or more corresponding samples of corrected job postings.

After creating the revised job posting, the revision presentation unit 118 displays the job posting before revision and the job posting after revision in a comparative manner. This allows the employer to easily confirm the revision content presented by the revision presentation unit 118. FIG. 10 is an example of the revision content display screen MD. The revision content display screen MD displays the revision summary MS, the pre-revision job posting JP1, and the revised job posting JP2. The pre-revision job posting JP1 and the revised job posting JP2 are displayed side by side on the left and right.

The correction presentation unit 118 may highlight the differences between the pre-correction job posting JP1 and the corrected job posting JP2 (the corrections in the corrected job posting JP2) so that they can be visually distinguished from other parts. The correction presentation unit 118 may also present only the corrections.

<Third Reception Unit 119>
The third reception unit 119 is configured to receive corrections to the job posting. Specifically, the third reception unit 119 receives input from the recruiter terminal 20 to reflect all or part of the corrections presented by the correction presentation unit 118. The third reception unit 119 also receives direct correction input (editing input) to the descriptions of the job posting from the recruiter terminal 20. The corrected job posting is stored in the database. The corrected job posting may also be evaluated again by the job posting evaluation unit 113.

<Artificial Intelligence Department 120>
The artificial intelligence unit 120 is configured to receive input from each functional unit and return the instructed output. The artificial intelligence used by the server device 10 in each functional unit may be a common one, or may be prepared individually for each functional unit or learning model.

The artificial intelligence unit 120 is an AI (Artificial Intelligence) equipped with transformers including GPT (Generative Pretrained Transformer, including GPT-1, GPT-2, and GPT-3), BERT (Bidirectional Encoder Representations from Transformers), and BART (Bidirectional and Auto-regressive Transformer), and language models such as a recurrent neural network (RNN), and includes a generative AI.

The language model is an example of a learning model based on a machine learning algorithm. Specific examples of machine learning algorithms include nearest neighbor methods, naive Bayes methods, decision trees, support vector machines, and deep learning using neural networks. The artificial intelligence unit 120 can apply the above algorithms as appropriate.

The artificial intelligence unit 120 has a trained model constructed by a known learning method such as supervised learning, unsupervised learning, or self-supervised learning. In supervised learning, machine learning is performed using training data (training data). The training data is composed of a pair of input data (explanatory variables) for learning and output data (target variables, correct answer labels) associated with the input data. When using supervised learning, the training data is converted in advance into a format suitable for machine learning by preprocessing. The artificial intelligence unit 120 performs machine learning processing using the training data to determine internal parameters of the learning model and stores the determined parameters as the learning model. For each learning model, the artificial intelligence unit 120 prepares the necessary training data according to the data to be input and the data to be output, and uses it for supervised learning of the learning model. In addition, the language model may not only be one trained for a specific task, but also a general-purpose model that can be used for a wide range of tasks.

The artificial intelligence unit 120 includes, as its artificial intelligence, a general-purpose natural language processing learning model such as a large-scale language model (LLM) that has learned a huge amount of data. Such a general-purpose learning model includes a language model that can handle various tasks without fine tuning by one-shot learning, few-shot learning, etc. The general-purpose learning model can also handle various tasks by zero-shot learning. In this case, the artificial intelligence unit 120 prepares a prompt according to the input data and the output data, and issues instructions to the learning model. The artificial intelligence used in each functional unit of the control unit 11 may be a separate learning model, or may be a common general-purpose learning model.

The learning model included in the artificial intelligence unit 120 is capable of additional learning. For example, the artificial intelligence unit 120 learns whether the corrections to the job posting that it presented have been reflected in the job posting by the employer. In other words, the artificial intelligence unit 120 performs additional learning and fine-tuning using a teacher dataset in which the presented corrections are labeled with the actual correction results as feedback for the corrections created and presented by the learning model. This allows the corrections output from the learning model to be optimized and presented to the user.

<Display section>
The display unit 211 of the recruiting party terminal 20 displays a screen indicated by the screen data transmitted from the server device 10.

<Operation reception section>
The operation reception unit 212 of the recruiting party terminal 20 receives operations by a user (recruiter) who uses the recruiting party terminal 20.

3. Information Processing Method This section describes an information processing method of the server device 10. This information processing method is executed by a computer, with each unit of the server device 10 acting as each step.

Specifically, this information processing method includes a first reception step, a first display control step, a second reception step, a second display control step, and a correction presentation step. In the first reception step, the input of a job posting is accepted. In the first display control step, information on a plurality of virtual job seekers corresponding to the job posting and job seeker characteristics corresponding to the job posting are displayed based on the contents of the job posting. In the second reception step, corrections to the job seeker characteristics are accepted. In the second display control step, information on a virtual job seeker corresponding to the corrected job seeker characteristics is displayed. In the correction presentation step, the corrections to the job posting based on the corrected job seeker characteristics are presented.

Figure 11 is an activity diagram showing the flow of information processing (job posting revision processing) executed by the recruitment support system 1. Below, the information processing is explained along with each activity in this activity diagram.

The job posting revision process begins with the employer inputting the job posting to be revised. The employer selects the job posting to be revised from a list presented by the server device 10 on the employer terminal 20 (activity A101). The server device 10 accepts the job posting input from the employer terminal 20 (first reception step, activity A102). Next, the server device 10 displays information on multiple virtual job seekers corresponding to the job posting and the job seeker characteristics corresponding to the job posting on the employer terminal 20 based on the contents of the job posting (first display control step, activity A103). This causes the virtual job seekers and job seeker characteristics to be displayed on the employer terminal 20 (activity A104).

The employer modifies the displayed job seeker characteristics on the employer terminal 20 on which the virtual job seeker and job seeker characteristics are displayed (activity A105). The server device 10 accepts the modification of the job seeker characteristics from the employer terminal 20 (second reception step, activity A106). Next, the server device 10 updates the information of the virtual job seeker to be displayed on the employer terminal 20 based on the modification of the job seeker characteristics (second display control step, activity A107). As a result, the virtual job seeker after the modification of the job seeker characteristics is displayed on the employer terminal 20 (activity A108).

On the employer terminal 20 where the revised virtual job seeker is displayed, the employer either revises the job seeker characteristics again (activity A105), or instructs the job posting to be revised using the current job seeker characteristics (activity A109). The server device 10 accepts the revision instruction from the employer terminal 20 and revises the job posting (activity A110). The server device 10 then presents the revisions to the job posting to the employer terminal 20 (revision presentation step, activity A111). This causes the revisions to the job posting to be displayed on the employer terminal 20 (activity A112).

4. Operation The operation of this embodiment can be summarized as follows. That is, the characteristics of the job seeker desired by the employer can be adjusted based on the information of the virtual job seeker who is the subject of the job advertisement. Therefore, an effective job advertisement can be created so that the employer can easily hire the job seeker desired.

Furthermore, when actual job seeker information is presented to an employer, it may lead to an action of sending a scout email to the job seeker, but it is unlikely to lead to revisions or improvements to the job posting. By presenting a virtual job seeker to an employer, as in this embodiment, it is possible to present the employer with a concrete image of the job seeker, without disclosing the personal information of the job seeker, and to have the employer revise the job posting.

The above describes an embodiment of the present invention, but the present invention is not limited to this and can be modified as appropriate without departing from the technical concept of the invention.

5. Others In the above embodiment, the server device 10 performs various storage and control, but multiple external devices may be used instead of the server device 10. That is, various information and programs may be distributed and stored in multiple external devices using block chain technology or the like. In particular, the artificial intelligence unit 120 may be an external configuration of the server device 10. In that case, the artificial intelligence unit 120, which is an external configuration, is configured to receive input from each functional unit of the server device 10 and return the instructed output to the server device 10.

The aspect of this embodiment is not limited to the recruiting support system 1, and may be an information processing method or a program. The recruiting support method includes each step of the recruiting support system 1. The program causes a computer to execute each step of the recruiting support system 1.

The control unit 11 of the recruitment support system 1 does not necessarily have to include the second display control unit 117. In other words, after accepting modifications to the job seeker characteristics, the control unit 11 may modify the job posting without updating the virtual job seeker displayed on the employer terminal 20.

<Other embodiments>
Fig. 12 is a block diagram showing functions realized by the server device 10 (control unit 11) in another embodiment. As shown in Fig. 12, the server device 10 in the other embodiment may further include a third display control unit 121. That is, the processor of the server device 10 may further execute a third display control step.

The third display control unit 121 is configured to display information on a plurality of virtual job seekers corresponding to the initial conditions and characteristics corresponding to the initial conditions based on the initial conditions for creating a job posting. The input of the initial conditions is accepted, for example, by the first acceptance unit 114.

For example, when creating a new job posting, the third display control unit 121 displays the job seeker display screen JD of FIG. 9 on the employer terminal 20, and allows the user to use the displayed virtual job seeker information and job seeker characteristics when creating a new job posting. Specifically, for example, after initial conditions are input on the initial condition input screen ID of FIG. 5 and FIG. 6, the third display control unit 121 extracts the virtual job seeker information and job seeker characteristics corresponding to the input initial conditions, and displays them on the job seeker display screen JD. The third display control unit 121 may also display the job seeker display screen JD on the employer terminal 20 without displaying the initial condition input screen ID of FIG. 5 and FIG. 6.

For example, the third display control unit 121 inputs the initial conditions and the database of virtual job seekers into the second virtual job seeker extraction model of the artificial intelligence unit 120, and causes the second virtual job seeker extraction model to output a virtual job seeker that matches the content of the initial conditions. The second virtual job seeker extraction model is a learning model that is trained to input the initial conditions and the database of virtual job seekers and output a virtual job seeker that matches the content of the initial conditions. In other words, the second virtual job seeker extraction model is a learning model that is trained using the initial conditions and the corresponding virtual job seekers as teacher data. The second virtual job seeker extraction model may also be a generative AI including a large-scale language model. In this case, the third display control unit 121 inputs a prompt into the second virtual job seeker extraction model that includes an instruction to extract a virtual job seeker that matches the content of the initial conditions by referring to the virtual job seeker database, and the initial conditions, and causes the second virtual job seeker extraction model to output a virtual job seeker that matches the content of the initial conditions. Furthermore, the third display control unit 121 may input a prompt that inserts, for example, one or more samples of the initial conditions and one or more samples of virtual job seekers corresponding thereto, in addition to the extraction instructions and initial conditions for the virtual job seeker to the second virtual job seeker extraction model.

The third display control unit 121 may, for example, input the initial conditions to the third feature extraction model of the artificial intelligence unit 120 and output the job seeker characteristics. The third feature extraction model is a learning model trained to input the initial conditions and output the job seeker characteristics. In other words, the third feature extraction model is a learning model trained using the initial conditions and the corresponding job seeker characteristics as teacher data. The third feature extraction model may also be a generative AI including a large-scale language model. In this case, the third display control unit 121 may input a prompt into which an instruction to extract the job seeker characteristics and the initial conditions are inserted into the third feature extraction model and output the job seeker characteristics that match the contents of the initial conditions to the third feature extraction model. The third display control unit 121 may also input, in addition to the instruction to extract the job seeker characteristics and the initial conditions, a prompt into which, for example, one or more samples of the initial conditions and one or more samples of the corresponding job seeker characteristics are inserted into the third feature extraction model.

The third display control unit 121 may further extract job seeker characteristics based on the extracted information of the virtual job seekers. In other words, the third display control unit 121 may input multiple virtual job seekers into the second feature extraction model of the artificial intelligence unit 120 and output the job seeker characteristics.

The second reception unit 116 accepts modifications to the job seeker characteristics displayed by the third display control unit 121. The second reception unit 116 may accept input of job seeker characteristics, for example, by the user selecting from a selection of job seeker characteristics prepared in advance and displayed on the job seeker display screen JD. Information on the virtual job seeker corresponding to the modified job seeker characteristics is displayed by the second display control unit 117.

The job posting creation unit 112 creates a job posting based on the revised job seeker information. That is, the job posting creation unit 112 creates a job posting that includes a description of the tags (keywords) and parameter conditions included in the revised job seeker characteristics. The job posting creation unit 112 inputs the revised job seeker characteristics and the initial conditions into a second job posting creation model of the artificial intelligence unit 120, and causes the second job posting creation model to output a job posting. The second job posting creation model is a learning model that is trained to receive the initial conditions and job seeker characteristics as input and to output a job posting. The second job posting creation model may also be a generative AI that includes a large-scale language model.

This allows a new job posting to be created based on the information of the job seeker characteristics selected while checking the information of the virtual job seeker, and allows a job posting that matches the profile of the person desired by the employer to be created. In addition to the initial conditions and job seeker characteristics, the revision history of the job seeker characteristics may be input to the second job posting creation model, and a job posting corresponding to the initial conditions, job seeker characteristics, and revision history may be output. This allows the employer to predict the matters that are important from the revision history and reflect them in the job posting. In this case, the second job posting creation model is a learning model that has been trained to input the initial conditions, job seeker characteristics, and revision history, and output a job posting corresponding to them. In addition, the initial conditions may not be input, and the selected job seeker characteristics and revision history (or only the job seeker characteristics) may be input to the second job posting creation model, and the corresponding job posting may be output. In this case, the second job posting creation model is a learning model that has been trained to input the job seeker characteristics and revision history (or only the job seeker characteristics) and output a corresponding job posting. In these cases, the second job posting creation model may be a generative AI that includes a large-scale language model. In this case, prompts that match the input data and output data are appropriately input to the second job posting creation model.

Note that the matters described in the embodiment of FIG. 4, such as accepting modifications to job seeker characteristics and displaying information about a virtual job seeker corresponding to the modified job seeker characteristics, can also be appropriately applied to the embodiment of FIG. 12. Also, in the embodiment of FIG. 12, the first display control unit 115 is not a required component.

It may be provided in the following ways:

(1) A recruitment support system comprising a processor, the processor being configured to execute the following steps: a first reception step receives input of a job posting; a first display control step displays information on a plurality of virtual job seekers corresponding to the job posting and characteristics corresponding to the job posting based on the contents of the job posting; a second reception step receives corrections to the characteristics; and a correction presentation step presents corrections to the job posting based on the corrected characteristics.

(2) A recruiting support system as described in (1) above, in which the feature relates to the skills or experience of the virtual job seeker.

(3) In the recruitment support system described in (1) or (2) above, the processor is further configured to execute the following steps, and in a second display control step, the recruitment support system displays information about the virtual job seeker that corresponds to the modified characteristics.

(4) In the recruitment support system described in any one of (1) to (3) above, in the first display control step, the virtual job seeker and the characteristics about which information is to be displayed are extracted based on the contents of the job posting.

(5) In the recruitment support system described in any one of (1) to (3) above, in the first display control step, the virtual job seeker whose information is to be displayed is extracted based on the contents of the job posting, and the characteristics are further extracted based on the information of the extracted virtual job seeker.

(6) In the recruiting support system described in any one of (1) to (5) above, in the second reception step, the recruiting support system receives modifications to the characteristics by adding, deleting, or moving tags corresponding to the characteristics that have been prepared in advance.

(7) In the recruiting support system described in (6) above, the tag includes a tag word related to the characteristic, and in the second reception step, the tag including the tag word that is a superordinate or subordinate conceptualization of the tag word included in the selected tag is presented as a recommended tag.

(8) In the recruitment support system described in any one of (1) to (7) above, in the second reception step, the recruitment support system receives modifications to the characteristics by setting parameter conditions related to the information of the virtual job seeker.

(9) In the recruiting support system described in (8) above, the parameter conditions are set by adjusting the positions of two knobs arranged on a slider that respectively set the lower and upper limits.

(10) A recruitment support system according to any one of (1) to (9) above, wherein the information on the virtual job seeker is generated based on information on a job seeker registered in advance.

(11) In the recruitment support system described in (10) above, the information on the virtual job seeker is generated by combining information on multiple job seekers who have been registered in advance.

(12) In the recruitment support system described in (10) above, the information on the virtual job seeker is generated by modifying or deleting part of the information on a previously registered job seeker.

(13) In the hiring support system described in any one of (1) to (12) above, the displayed information on the virtual job seeker includes a summary text created by the writing model by inputting attributes of the virtual job seeker into the writing model.

(14) In the recruitment support system described in any one of (1) to (13) above, in the revision presentation step, the revised characteristics and the initial conditions included in the job posting are input to a first job posting revision model, and the first job posting revision model is caused to output revised initial conditions in which the initial conditions have been revised to correspond to the characteristics, and the revised initial conditions are input to a job posting creation model, and the job posting creation model is caused to output a revised job posting, where the initial conditions include at least one of the profile of the person desired by the employer and the recruitment background.

(15) In the recruitment support system described in any one of (1) to (14) above, in the revision presentation step, the job posting is input to an initial condition estimation model, and the initial condition estimation model is caused to estimate initial conditions corresponding to the job posting, where the initial conditions include at least one of the profile of the person desired by the employer and the recruitment background, the revised characteristics and the estimated initial conditions are input to a first job posting revision model, and the first job posting revision model is caused to output revised initial conditions in which the initial conditions are revised to correspond to the characteristics, and the revised initial conditions are input to a job posting creation model, and the job posting creation model is caused to output the revised job posting.

(16) In the recruitment support system described in (14) or (15) above, the processor is further configured to execute the following steps: in the storage step, a revision history of the characteristics in the second reception step is stored; in the revision presentation step, in addition to the characteristics and the job posting, the revision history is further input to the first job posting revision model; and the first job posting revision model is caused to output the revised initial conditions in which the initial conditions have been revised to correspond to the characteristics and the revision history. A recruitment support system.

(17) In the recruitment support system described in any one of (1) to (16) above, in the correction presentation step, the corrected characteristics and the job posting are input into a second job posting correction model, and a corrected job posting in which the job posting is corrected to correspond to the characteristics is output to the second job posting correction model.

(18) In the recruitment support system described in (17) above, the processor is further configured to execute the following steps: in the storage step, a revision history of the characteristics in the second reception step is stored; in the revision presentation step, in addition to the characteristics and the job posting, the revision history is further input to the second job posting revision model; and the second job posting revision model is caused to output the revised job posting in which the job posting has been revised to correspond to the characteristics and the revision history. A recruitment support system.

(19) In the recruitment support system described in any one of (1) to (18) above, in the revision suggestion step, the job posting before revision and the job posting after revision are displayed in comparison.

(20) A recruitment support method comprising steps executed by a recruitment support system described in any one of (1) to (19) above.

(21) A program that causes a computer to execute each step of the recruitment support system described in any one of (1) to (19) above.
Of course, this is not the case.

Finally, various embodiments of the present disclosure have been described, but these are presented as examples and are not intended to limit the scope of the invention. The novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. The embodiments and modifications thereof are within the scope and spirit of the invention, and are included in the scope of the invention and its equivalents as set forth in the claims.

1: Recruitment support system 2: Communication line 10: Server device 11: Control unit 12: Memory unit 13: Communication unit 14: Communication bus 20: Recruiter terminal 21: Control unit 22: Memory unit 23: Communication unit 24: Input unit 25: Output unit 26: Communication bus 111: Basic display control unit 112: Job posting creation unit 113: Job posting evaluation unit 114: First reception unit 115: First display control unit 116: Correction suggestion unit 117: Second display control unit 118: Correction suggestion unit 119: Third reception unit 120: Artificial intelligence unit 121: Third display control unit 211: Display unit 212: Operation reception unit

Claims (16)

A recruitment support system,
A processor is provided.
The processor is configured to execute the following steps:
In the first reception step, the job posting is received,
In the first display control step, the job advertisement is input into a virtual job seeker extraction model, a plurality of virtual job seekers corresponding to the job advertisement are output from the virtual job seeker extraction model, the plurality of virtual job seekers are input into a feature extraction model, the feature extraction model is caused to output features which are keywords representing careers, experiences, skills, or qualifications of the plurality of virtual job seekers, and information and the features of the plurality of virtual job seekers are further displayed;
Here, the virtual job seeker extraction model is a learning model trained to receive the job advertisement and the virtual job seeker database as inputs and to output the virtual job seeker that matches the content of the job advertisement, and the feature extraction model is a learning model trained to receive a plurality of the virtual job seekers as inputs and to output the features,
In the second receiving step, a correction to the feature is received;
In the modification submission step, the job posting and the modified features are input to a modification model, and a modified job posting including a keyword included in the features or a keyword similar to the keyword is output from the modification model;
Here, the revised model is a learning model that is trained to take the job posting and the features as input and to output the revised job posting , in this recruitment support system. A recruitment support system,
A processor is provided.
The processor is configured to execute the following steps:
In the first reception step, the job posting is received,
In the first display control step, the job advertisement is inputted into a virtual job seeker extraction model, and a plurality of virtual job seekers corresponding to the job advertisement are outputted from the virtual job seeker extraction model, and the job advertisement is inputted into a feature extraction model, and features which are keywords representing careers, experiences, skills, or qualifications described in the job advertisement are outputted from the feature extraction model, and information and the features of the plurality of virtual job seekers are displayed;
Here, the virtual job seeker extraction model is a learning model trained to receive the job advertisement and the virtual job seeker database as inputs and output the virtual job seeker that matches the content of the job advertisement, and the feature extraction model is a learning model trained to receive the job advertisement as inputs and output the features,
In the second receiving step, a correction to the feature is received;
In the modification submission step, the job posting and the modified features are input to a modification model, and a modified job posting including a keyword included in the features or a keyword similar to the keyword is output from the modification model;
Here, the revised model is a learning model that is trained to take the job posting and the features as input and output the revised job posting, in this recruitment support system. 3. The recruitment support system according to claim 1,
The processor is further configured to perform the steps of:
In the second display control step, the recruitment support system displays information on the virtual job seeker searched from the database using the modified characteristics as a search condition . 3. The recruitment support system according to claim 1,
In the second reception step, a recruitment support system prepares in advance tags including tag words extracted from the job posting or information on the virtual job seeker, presents specific tags among the tags in which the tag words are included in the characteristics, and accepts addition or deletion of the tag words to the characteristics by adding, deleting, or moving the specific tags. 3. The recruitment support system according to claim 1,
In the second receiving step, a correction of the numerical range of the characteristic is received by setting a parameter condition indicating an amount of experience, a level of expertise, or a numerical range of annual income of the virtual job seeker ;
In the revision presentation step, the revised model is caused to output the revised job posting in which the numerical range included in the revised characteristic is the recruitment target . 6. The recruitment support system according to claim 5 ,
A recruitment support system in which the parameter conditions are set by adjusting the positions of two knobs arranged on a slider that set a lower limit and an upper limit, respectively. 3. The recruitment support system according to claim 1,
The information of the virtual job seeker is obtained by inputting information of a job seeker registered in advance into a virtual job seeker creation model and outputting the information to the virtual job seeker creation model ,
Here, the virtual job seeker creation model is a learning model that is trained to take information of a pre-registered job seeker as input, synthesize information on a plurality of pre-registered job seekers, and modify or delete part of the information on the pre-registered job seeker to generate information on the virtual job seeker, said recruitment support system . 3. The recruitment support system according to claim 1,
The displayed information of the virtual job seeker includes a summary sentence generated by a writing model by inputting attributes of the virtual job seeker into the writing model ,
A recruitment support system in which the attributes are the virtual job seeker's experience in an organization, the type of job experience, the skills and qualifications held, the annual income, or the age, the summary text is a text summarizing the virtual job seeker's career, work history, experience, skills, or qualifications, and the text creation model is a learning model trained to use the attributes as input and output the summary text . 3. The recruitment support system according to claim 1,
The correction model includes a first job posting correction model and a job posting creation model,
In the modification presentation step, the modified features and the initial conditions included in the job posting are input to the first job posting modification model, and the first job posting modification model is caused to output modified initial conditions in which the initial conditions are modified to correspond to the features, and the modified initial conditions are input to the job posting creation model, and the job posting creation model is caused to output the modified job posting;
Here, the initial conditions include at least one of the personality profile desired by the employer and the recruitment background, the first job posting correction model is a learning model trained to take the characteristics and the initial conditions as input and output the modified initial conditions, and the job posting creation model is a learning model trained to take the modified initial conditions as input and output the modified job posting , a recruitment support system. 3. The recruitment support system according to claim 1,
The correction model includes an initial condition estimation model, a first job posting correction model, and a job posting creation model;
In the correction suggestion step,
Inputting the job posting including at least one of a position name, a job description, working conditions, application qualifications, and appealing points into the initial condition estimation model, and having the initial condition estimation model estimate initial conditions corresponding to the job posting;
Here, the initial conditions include at least one of a personality profile desired by the recruiter and a recruitment background , and the initial condition estimation model is a learning model trained to use the job posting as an input and the initial conditions as an output,
The corrected features and the estimated initial conditions are input to the first job posting correction model, and the first job posting correction model is caused to output corrected initial conditions in which the initial conditions are corrected to correspond to the features, and the corrected initial conditions are input to the job posting creation model, and the job posting creation model is caused to output the corrected job posting ;
Here, the first job posting correction model is a learning model trained to take the features and the initial conditions as input and output the modified initial conditions, and the job posting creation model is a learning model trained to take the modified initial conditions as input and output the modified job posting , a recruitment support system. The recruitment support system according to claim 9 ,
The processor is further configured to perform the steps of:
In the storing step, a correction history of the feature in the second receiving step is stored;
In the modification suggestion step, in addition to the features and the initial conditions , the modification history is further input to the first job advertisement modification model, and the first job advertisement modification model is caused to output the modified initial conditions obtained by modifying the initial conditions so as to correspond to the features and the modification history ;
Here, the first job posting correction model is a learning model that is trained to input the features, the initial conditions, and the correction history and output the corrected initial conditions , in this recruitment support system. 3. The recruitment support system according to claim 1,
The modification model includes a second job posting modification model;
In the modification submission step, the modified characteristics and the job posting are input to the second job posting modification model, and the second job posting modification model is caused to output the modified job posting in which the job posting is modified to correspond to the characteristics ;
Here, the second job posting correction model is a learning model that is trained to input the features and the job posting and output the corrected job posting , in this recruitment support system. The recruitment support system according to claim 12 ,
The processor is further configured to perform the steps of:
In the storing step, a correction history of the feature in the second receiving step is stored;
In the correction suggestion step, in addition to the characteristics and the job posting, the correction history is further input to the second job posting correction model, and the second job posting correction model is caused to output the corrected job posting in which the job posting is corrected to correspond to the characteristics and the correction history ,
Here, the second job posting correction model is a learning model that is trained to input the features, the job posting, and the correction history, and output the corrected job posting , in this recruitment support system. 3. The recruitment support system according to claim 1,
In the correction suggestion step, the recruitment support system displays the job posting before correction and the job posting after correction in a comparative manner. A recruitment support method, comprising:
A recruitment support method comprising the steps executed by the recruitment support system according to claim 1 or 2 . A program,
A program for causing a computer to execute each step of the recruitment support system according to claim 1 or 2 .