Information processing systems, information processing methods, and programs

Abstract

We provide information processing systems and other tools that enable the creation of high-quality job postings. [Solution] According to one aspect of the present invention, an information processing system is provided, comprising at least one processor, wherein the processor is configured to perform the following steps by reading a program: in a first job information acquisition step, first job information including a job description document relating to the organization or business of the target job is acquired; in a similar job extraction step, similar job postings similar to the target job are extracted from job postings registered in the database based on the first job information; in an analysis step, improvement suggestion information including points for improvement of the target job is created based on a comparison between the job description document and second job information representing the content of the similar job postings; and in an improvement information display control step, the improvement suggestion information is displayed.

Description

Technical Field

[0001] The present invention relates to an information processing system, an information processing method, and a program.

Background Art

[0002] Patent Document 1 discloses a technique for appropriately matching applicants (job seekers) and employers.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] There is a need for a technique that can create high-quality job offers effective for recruiting job seekers.

[0005] In view of the above circumstances, the present invention aims to provide an information processing system and the like that can create high-quality job offers.

Means for Solving the Problems

[0006] According to one aspect of the present invention, there is provided an information processing system including at least one processor configured to execute the following steps by reading a program. In a first job information acquisition step, first job information including a job description text related to the organization or business of a target job offer is acquired. In a similar job extraction step, based on the first job information, similar jobs similar to the target job offer are extracted from the jobs registered in a database. In an analysis step, improvement proposal information including improvement points of the target job offer is created based on a comparison between the job description text and second job information representing the content of the similar jobs. In an improvement information display control step, the improvement proposal information is displayed.

[0007] This method allows for the creation of high-quality job postings. [Brief explanation of the drawing]

[0008] [Figure 1] This is a diagram showing the configuration of Information Processing System 1. [Figure 2] This is a block diagram showing the hardware configuration of server device 10. [Figure 3] This block diagram shows the hardware configuration of the job seeker terminal 20 and the job applicant terminal 30. [Figure 4] This is a block diagram showing the functions realized by the server device 10 (control unit 11), the job seeker terminal 20 (control unit 21), and the job seeker terminal 30 (control unit 31). [Figure 5] This figure shows an example of a job entry screen ID displayed on the job seeker terminal 20. [Figure 6] This figure shows an example of the response rate display screen RD shown on the job seeker terminal 20. [Figure 7] This figure shows an example of the evaluation display screen ED that is displayed on the job seeker terminal 20. [Figure 8] This figure shows an example of the SD screen displaying similar job postings on the job seeker terminal 20. [Figure 9] This figure shows an example of the revised proposal display screen CD that appears on the job seeker terminal 20. [Figure 10] This is an activity diagram showing an example of the flow of information processing (display processing of improvement suggestion information) performed by Information Processing System 1. [Modes for carrying out the invention]

[0009] Embodiments of the present invention will be described below with reference to the drawings. The various features shown in the embodiments below can be combined with each other.

[0010] Incidentally, the program for implementing the software appearing in one embodiment may be provided as a non-transitory computer-readable medium, or it may be provided as a downloadable medium from an external server, or it may be provided so that the program is launched on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] Furthermore, in various information processing according to one embodiment, an input and an output corresponding to the input can be realized. Here, as long as an output is obtained as a result of the input, the form of the information referenced in such information processing (hereinafter referred to as "reference information") is not limited. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a decision formula such as a regression equation constructed by a statistical method), or a trained model that has been pre-trained to learn the correlation between input and output, or a generative AI such as a large-scale language model that can output a desired result by inputting a prompt (these models include parameters that construct the correlation relationship between input and output) or a visual language model.

[0012] Furthermore, in one embodiment, "part" may include, for example, hardware resources implemented by a circuit in a broad sense, and the information processing of software that can be specifically realized by these hardware resources. Also, in one embodiment, various types of information are handled, and this information can be represented, for example, by the physical values ​​of signal values ​​representing voltage and current, the high or low values ​​of signal values ​​as a set of binary bits composed of 0s or 1s, or by quantum superposition (so-called qubits), and communication and calculations can be performed on a circuit in a broad sense.

[0013] Furthermore, a circuit in a broad sense is a circuit realized by combining at least a suitable combination of circuits, circuits, processors, and memory. The processor may be a general-purpose processor or a dedicated circuit. In other words, it includes application-specific integrated circuits (ASICs), programmable logic devices (for example, simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs)), etc.

[0014] 1. Hardware Configuration This section describes the hardware configuration.

[0015] <Information Processing System 1> Figure 1 is a configuration diagram representing information processing system 1. Information processing system 1 comprises a communication line 2, a server device 10, multiple employer terminals 20, and multiple job seeker terminals 30. The server device 10, employer terminals 20, and job seeker terminals 30 are configured to communicate with each other via the communication line 2. The connection between the server device 10, employer terminals 20, and job seeker terminals 30 may be wired or wireless. Furthermore, the server device 10, employer terminals 20, and job seeker terminals 30 are each examples of information processing devices.

[0016] The information processing system 1 constitutes at least a part of a job offer - job seeking system used by, for example, a plurality of job offerers (the first job offerer U1 and the second job offerer U2) and a plurality of job seekers (the first job seeker U3 and the second job seeker U4). The information processing system 1 mainly performs operations such as searching for job seekers by job offerers, searching for job offers by job seekers, and mediating communication between job offerers and job seekers. For example, the information processing system 1 provides and manages a talent matching platform, a talent matching service, etc. that are used by job offerers and job seekers. In one embodiment, the information processing system 1 is composed of one or more devices or components. Hereinafter, these components will be described.

[0017] <Server device 10> FIG. 2 is a block diagram showing the hardware configuration of the server device 10. As shown in FIG. 2, 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 via the communication bus 14 inside the server device 10.

[0018] <Control unit 11> The control unit 11 performs processing and control of the overall operations related to the server device 10. The control unit 11 is, for example, a central processing unit (CPU), which is an example of a processor. The control unit 11 realizes various functions related to the server device 10 by reading a predetermined program stored in the storage unit 12. That is, the information processing by 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. These will be described in more detail in the next section. Note that the control unit 11 is not limited to being single, and the server device 10 may have a plurality of control units 11 for each function. Also, the server device 10 may be composed of combinations of these.

[0019] <Storage unit 12> The storage unit 12 stores various types of information as defined above. This can be done, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the server device 10 executed by the control unit 11, or as memory such as random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The storage unit 12 stores various programs, variables, etc. related to the server device 10 executed by the control unit 11.

[0020] <Communications Department 13> The communication unit 13 preferably uses wired communication methods such as USB, IEEE1394, Thunderbolt®, and wired LAN network communication, but may also include wireless LAN network communication, mobile communication such as LTE / 5G, and Bluetooth® communication as needed. In other words, the communication unit 13 may be implemented as a collection of these multiple communication methods. Furthermore, the server device 10 may communicate various information with the outside world via the communication unit 13 and the network.

[0021] The server device 10 may be on-premises or in a cloud environment. The cloud-based server device 10 may provide the above-mentioned functions and processing in the form of, for example, SaaS (Software as a Service) or cloud computing.

[0022] <Job seeker terminal 20> Figure 3 is a block diagram showing the hardware configuration of the employer terminal 20 and the job seeker terminal 30. The employer terminal 20 is an information processing terminal used by employers and can access the server device 10.

[0023] "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), or their representatives. Representatives within employers may also be called hiring managers, and may include personnel from the organization's human resources department or the department responsible for hiring. Furthermore, employers may also include headhunters. A headhunter is an organization or its representative that acts as an intermediary between job seekers and employers (organizations) on behalf of the organization (employer). Headhunters are also known as recruitment agencies, hiring agents, or recruitment agencies.

[0024] As shown in Figure 3A, the job seeker terminal 20 comprises a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, an output unit 25, and a communication bus 26. The control unit 21, storage unit 22, communication unit 23, input unit 24, and output unit 25 are electrically connected within the job seeker terminal 20 via the communication bus 26. The descriptions of the control unit 21, storage unit 22, and communication unit 23 are the same as the descriptions of each part in the server device 10 and are therefore omitted.

[0025] <Input section 24> The input unit 24 receives operation inputs made by the user. The operation inputs are transmitted as command signals to the control unit 21 via the communication bus 26. The control unit 21 can perform predetermined controls or calculations based on the transmitted command signals as needed. The input unit 24 may be included in the housing of the job seeker terminal 20 or it may be an external component. 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, etc. to the input unit 24. Instead of a touch panel, the input unit 24 can be a switch button, mouse, trackpad, QWERTY keyboard, etc.

[0026] <Output section 25> The output unit 25 displays a graphical user interface (GUI) screen that can be operated by the user. The output unit 25 may be included in the housing of the job seeker terminal 20 or it may be an external device. Specifically, the output unit 25 can be implemented as a display device such as a CRT display, liquid crystal display, organic EL display, or plasma display. It is preferable that these display devices be used according to the type of job seeker terminal 20.

[0027] <Job seeker terminal 30> The job seeker terminal 30 is an information processing terminal used by job seekers and can access the server device 10. "Job seekers" include, for example, those who are looking for a new job or a new job (e.g., currently employed individuals (those seeking a new job), prospective graduates (job seekers), students, etc.), and those who are interested in changing jobs or finding employment.

[0028] As shown in Figure 3B, the job seeker terminal 30 comprises a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, an output unit 35, and a communication bus 36. The control unit 31, storage unit 32, communication unit 33, input unit 34, and output unit 35 are electrically connected within the job seeker terminal 30 via the communication bus 36. The descriptions of the control unit 31, storage unit 32, communication unit 33, input unit 34, and output unit 35 are the same as the descriptions of each part in the employer terminal 20 and are therefore omitted.

[0029] 2. Functional Configuration This section describes the functional configuration of this embodiment. Information processing by software stored in the memory 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 (at least one processor provided by the information processing system 1).

[0030] Figure 4 is a block diagram showing the functions realized by the server device 10 (control unit 11), the job seeker terminal 20 (control unit 21), and the job seeker terminal 30 (control unit 31).

[0031] As shown in Figure 4A, the server device 10 (control unit 11) comprises a basic display control unit 111, a first job information acquisition unit 112, a similar job extraction unit 113, a second job information creation unit 114, an analysis unit 115, an improvement information display control unit 116, and an artificial intelligence unit 120.

[0032] As shown in Figure 4B, the job seeker terminal 20 (control unit 21) includes a display unit 211 and an operation acquisition unit 212. As shown in Figure 4C, the job seeker terminal 30 (control unit 31) includes a display unit 311 and an operation acquisition unit 312.

[0033] <Basic display control unit 111> The basic display control unit 111 is configured to display various information on the employer terminal 20 or the job seeker terminal 30. For example, in response to requests from each user (employers U1, U2 or job seekers U3, U4), the basic display control unit 111 displays the registration information of job seekers registered in the database on the display unit 211 of the employer terminal 20 or the display unit 311 of the job seeker terminal 30.

[0034] <1st job information acquisition department 112> The first job information acquisition unit 112 is configured to acquire first job information, which includes a job description document relating to the organization or business of the target job. The "job description document" is a document that describes at least one of the following: the organization, industry, department, job type, duties, position, annual salary, and work style of the target job. For example, the job description document may include a description of the organization's history, business environment, characteristics, a specific description of the job type or duties, and a description of working conditions.

[0035] The first job posting may include job requirements in addition to the job description. Job requirements define the conditions and qualities that are expected of job seekers for the position being advertised. Job requirements may include, for example, required skills, required experience, required qualifications, desired personality, character, values, and attitude. Furthermore, job requirements may include mandatory requirements, which are conditions that are not required, and desirable requirements, which are conditions that are not mandatory but are desirable. The first job posting is typically a job application form, but it does not necessarily have to be formatted like a job application form; it may be an unstructured document such as a memo listing the job requirements as items.

[0036] The first job information acquisition unit 112 may, for example, accept the upload of first job information from the recruiter terminal 20 and acquire the uploaded first job information. Alternatively, the first job information acquisition unit 112 may accept input from the recruiter terminal 20 such as information indicating the target job (for example, the title and ID of the target job) and the storage location of the first job information (network address, URL, path, etc.), and acquire the corresponding first job information from a database or the like.

[0037] The first job information acquisition unit 112 may receive input of information indicating the target job and specific keywords from the job seeker terminal 20, and acquire the first job information and specific keywords for the target job. "Specific keywords" are keywords that the similar job extraction unit 113, described later, refers to when extracting similar job postings. Examples of specific keywords include keywords related to organizations or operations (e.g., keywords representing organizations, industries, departments, job titles, duties, positions, annual income, work style, etc.) and keywords related to job requirements (e.g., keywords representing required skills, required experience, required qualifications, etc.).

[0038] The first job information acquisition unit 112 may accept input of multiple specific keywords in combination with information indicating a single target job. These specific keywords may also be referenced by the second job information creation unit 114, which will be described later, when creating the second job information.

[0039] Figure 5 shows an example of a job posting input screen ID displayed on the recruiter terminal 20. The job posting input screen ID includes the job posting input field RF, the keyword input field KF, the close button B11, and the create button B12.

[0040] The job posting input field RF accepts input of information indicating the target job posting (job ID in the example in Figure 5). The keyword input field KF accepts input of a specific keyword. When an input operation is performed on the close button B11, the contents entered in the job posting input field RF and the keyword input field KF are discarded, and the job posting screen ID is closed. When an input operation is performed on the create button B12, the information indicating the target job posting entered in the job posting input field RF and the specific keyword entered in the keyword input field KF are sent to the server device 10. Note that entering a specific keyword in the keyword input field KF is not mandatory.

[0041] <Similar job offer extraction section 113> The similar job extraction unit 113 is configured to extract similar job postings from the job postings registered in the job posting database that are similar to the target job posting, based on the first job posting information acquired by the first job posting information acquisition unit 112.

[0042] The similar job search unit 113 may extract one or more similar job postings. In other words, the similar job search unit 113 may extract multiple similar job postings. Furthermore, a fixed upper limit (for example, 5) may be set for the number of similar job postings to be extracted.

[0043] Similar job postings are extracted, for example, using job descriptions registered in a job database. Note that similar job postings may be from different employers than the one posting the target job. Furthermore, similar job postings may include postings that are no longer being advertised (e.g., filled positions, postings whose application period has ended).

[0044] The similarity between two job postings is determined, for example, by the difference in the features of the job information of the two postings (for example, the job description text and job requirements included in the job posting) (for example, the distance between vectors obtained by vectorizing words or sentences). The similar job posting extraction unit 113 determines that two job postings are similar if the difference in their features is small (for example, the cosine similarity, which is the distance between the two vectors, is above a threshold). Vectorization is performed by quantification using known methods such as natural language processing by morphological analysis or encoding. Alternatively, features may be calculated by referring to a table in which features are defined for each keyword. Furthermore, the similar job posting extraction unit 113 may input the two job postings into a similarity determination model, which is a pre-trained model (a dedicated learning model or a general-purpose learning model), and have the similarity determination model output a similarity score or similarity determination result.

[0045] The similar job search unit 113 may determine similarity between two job postings based on the matching, similarity, or mismatch of the content of corresponding items in the job information of each job posting. For example, the similar job search unit 113 may determine that two job postings are similar if their job titles and / or industries match or are similar. In particular, the similar job search unit 113 may extract similar job postings based on the similarity of at least one of the following: industry, job title, and organizational size. Furthermore, the similar job search unit 113 may determine that two job postings are similar if their annual salary ranges match. In addition, the similar job search unit 113 may determine that two job postings are similar if, for example, the number of matching items exceeds a predetermined threshold.

[0046] The similar job extraction unit 113 may extract similar job postings from the job database that are similar to the target job posting and related to the specific keyword, based on the first job posting information and specific keyword acquired by the first job posting information acquisition unit 112. This makes it possible for job seekers to adjust the similar job postings used for comparison by the analysis unit 115, described later, using specific keywords.

[0047] "Similar job postings related to a specific keyword" are, for example, job postings whose content (job description) contains keywords that match or are similar to the specific keyword. Here, keyword similarity is determined by methods such as comparison using feature quantities such as vector data, referencing a database containing synonyms, related words, and variations in spelling, or determination using a trained model. Furthermore, "keywords similar to a specific keyword" may include keywords that are higher-level (abstract) or lower-level (concretized) concepts of the specific keyword.

[0048] The similar job search unit 113 extracts multiple candidate job postings similar to the target job posting based on the first job posting information, and may further extract similar job postings from among the multiple candidate job postings according to the relationship between the content of the candidate job postings and specific keywords. This allows for more accurate adjustment of similar job postings based on specific keywords.

[0049] As described above, candidate job postings are determined by the difference in the features of the job postings compared to the first job posting. The number of candidate job postings extracted is, for example, between 20 and 100.

[0050] "Relevance to a specific keyword" refers to, for example, the degree of similarity between a specific keyword and the content of a candidate job posting (job description or job requirements). The similarity between a specific keyword and the content of a candidate job posting may be calculated based on, for example, the cosine similarity value, and may be expressed as, for example, 0% to 100%. The similar job posting extraction unit 113 may, for example, extract candidate job postings as similar job postings whose relevance to a specific keyword is above a predetermined threshold (for example, a similarity of XX or more). Alternatively, the similar job posting extraction unit 113 may, for example, extract candidate job postings as similar job postings whose rank is above a predetermined level (for example, 10th place or higher) when multiple candidate job postings are arranged in order of relevance.

[0051] Furthermore, the similar job search unit 113 may, as a "relevance to a specific keyword," for example, extract candidate job postings that contain a specific keyword or a keyword similar to the specific keyword as similar job postings, or it may extract candidate job postings that have a high frequency of occurrence of the specific keyword or a keyword similar to the specific keyword (for example, a frequency of occurrence above a predetermined threshold) as similar job postings.

[0052] The similar job search unit 113 may also extract similar job postings by considering the expected value of successful placement, as described below, in addition to the similarity. For example, the similar job search unit 113 may extract as similar job postings any candidate job postings where the similarity is above a threshold, the expected value of successful placement is above a certain rank (rank of XX or higher), or the parameter combining similarity and expected value of successful placement is above a threshold, or the rank of the parameter is above a certain rank.

[0053] The similar job search unit 113 may also extract job postings that are similar to the target job posting and that meet the extraction criteria in terms of the number of actions taken towards job seekers. This excludes job postings with low activity (approaches to job seekers) from the similar job postings, thereby increasing the effectiveness of the improvement suggestion information provided by the analysis unit 115, which will be described later.

[0054] "Actions directed at job seekers" include, for example, sending a recruitment letter (with the job posting attached) based on the job opening to the job seeker, and adding the job seeker who is a candidate for the job opening to a bookmark list. Here, a "recruitment letter" is a document sent by an employer (including headhunters) to a job seeker with the aim of encouraging them to apply for the selection process or proposing an interview, and may also be called a recruitment email. A "bookmark list" (also called a "favorites list" or "interesting list") is a list prepared by each employer where the employer can register any job seeker they wish to hire.

[0055] The "extraction criteria" could be, for example, "the number of actions is greater than or equal to a predetermined value." The extraction criteria may also include the period for counting the number of actions. For example, the extraction criteria could be "the number of actions is greater than or equal to a predetermined value (for example, the number of scouting documents sent is 30 or more) during a predetermined period (for example, the most recent year)."

[0056] The similar job search unit 113 may also extract job postings as similar jobs that are similar to the target job posting, related to specific keywords, and whose number of actions towards job seekers meets the extraction criteria. In other words, the similar job search unit 113 may use a combination of similarity to the target job posting, relevance to specific keywords, and the number of actions as criteria for similar jobs.

[0057] <Second Job Posting Creation Department 114> The second job posting creation unit 114 is configured to create a second job posting that includes the content extracted from each of the multiple similar job postings, based on the content of multiple similar job postings extracted by the similar job posting extraction unit 113 and the third reference information (reference information for job posting creation). This allows the analysis unit 115, described later, to create more meaningful improvement suggestion information.

[0058] "Second job information" is, for example, information extracted item by item from the content (job postings) of multiple similar job postings and integrated item by item. Second job information may be a fictitious job posting generated based on the content of multiple similar job postings, or it may be information with a different structure (format) than the job posting (for example, information that includes items in a different category than the job posting).

[0059] "Integration" here includes processes such as: creating a second job posting by simply adding up the contents of multiple similar job postings; creating a second job posting by adding up the parts common to the contents of multiple similar job postings; creating a second job posting by adding up the reference parts extracted from the contents of multiple similar job postings; and creating a second job posting by selecting the most appropriate and optimal content among multiple similar job postings in comparison with the target job posting (first job posting). Furthermore, the second job posting may also include, for example, a "winning pattern" extracted from multiple similar job postings that meet a predetermined criterion (e.g., expected value of successful placement is above a threshold). Note that the integration process may be performed on the entire contents (job postings) of multiple similar job postings, or on each item of multiple similar job postings (e.g., the section describing the organization).

[0060] The "reference portion" used to integrate the content of similar job postings is, for example, a portion that is more similar to the target job posting than other portions, or a portion that is more relevant to a specific keyword than other portions. The "optimal content" used to integrate the content of similar job postings is, for example, the content with the highest success score (indicating the likelihood of a successful placement) among multiple similar job postings for a given item. The "success score" is derived from the correlation between the content of a job posting and the likelihood of a successful placement (information included in the third reference information), which is constructed based on data from job postings registered in the job database to date and the history of actions taken by job seekers on that job posting (number of views, response rate to scouting documents, number of applications, whether a placement was made, etc.).

[0061] The third reference information is information relating to the correlation between the content of multiple similar job postings and the second job posting. The third reference information is stored, for example, in the memory unit 12. The third reference information may include, for example, tables, functions, simple algorithms, etc., that show the correlation between the content of multiple similar job postings and the second job posting. The correlation included in the third reference information can be constructed, for example, by statistically analyzing data that records the content of multiple similar job postings and the corresponding second job posting. For example, the third reference information may include a function that extracts information such as skills and qualifications from the content of multiple similar job postings and uses the skills and qualifications with the highest frequency of occurrence as the content of the second job posting, or an algorithm that calculates the average value of the salary range included in multiple similar job postings and uses it as the content of the second job posting.

[0062] The third reference information may include a set of parameters for generating second job information from the contents of multiple similar job postings. For example, the third reference information may be various pre-trained models. For example, the third reference information may include a second job information creation model which is a dedicated learning model or a general-purpose learning model that has been machine-trained to take the contents of multiple similar job postings as input and output second job information. In this case, the second job information creation unit 114 inputs the contents of multiple similar job postings into the second job information creation model and causes the second job information creation model to output second job information.

[0063] The second job posting creation model is included in the artificial intelligence unit 120. The second job posting creation model, which is a dedicated learning model, may be constructed, for example, by learning using data of the content of multiple similar job postings and the corresponding data of the second job postings as training data. In such a second job posting creation model, parameters calculated and tuned through learning construct a correlation between the content of multiple similar job postings and the second job postings. The dedicated learning model may also include a generative AI capable of generating output information (answers) that are not included in the training data based on the input information. The generative AI of the dedicated learning model is a limited-use generative AI that does not require input of instructions such as the content of the output information to be generated or the content of the task to be executed.

[0064] If the second job posting creation model is a general-purpose learning model (for example, a language model such as a large-scale language model), the second job posting creation unit 114 inputs a prompt to the second job posting creation model that includes the contents of multiple similar job postings and an instruction to output second job postings corresponding to the contents of the multiple similar job postings as input, causing the second job posting creation unit to output the second job postings. The second job posting creation unit 114 may also generate a prompt that gives the second job posting creation model an instruction to create second job postings and input this prompt to the second job posting creation model. In addition to the contents of multiple similar job postings and the instruction to create and output second job postings, the second job posting creation unit 114 may also input a prompt to the second job posting creation model that includes, for example, one or more samples of the contents of multiple similar job postings and one or more samples of corresponding second job postings as examples, samples, or training data of input and output pairs. Here, parameters for constructing a second job posting creation model and prompts containing instructions to output second job postings corresponding to the content of multiple similar job postings establish a correlation between the content of multiple similar job postings and the second job postings. The general-purpose learning model may include a generative AI capable of generating arbitrary output information based on input information. The generative AI in the general-purpose learning model is a general-purpose generative AI that requires input such as instructions for the content of the output information to be generated and the content of the task to be performed.

[0065] The second job information creation unit 114 may create second job information based on the content of multiple similar job postings extracted by the similar job posting extraction unit 113, as well as specific keywords acquired by the first job information acquisition unit 112. In this case, for example, if the second job information creation model is a general-purpose learning model (for example, a language model such as a large-scale language model), the second job information creation unit 114 inputs a prompt to the second job information creation model that includes the content of multiple similar job postings, specific keywords, and an instruction to output second job information based on the content related to the specific keywords included in the multiple similar job postings, causing the second job information creation model to output second job information.

[0066] The second job posting creation unit 114 may create the second job posting using multiple evaluation axes used by the analysis unit 115, which will be described later. For example, the second job posting creation unit 114 may input prompts to the second job posting creation model, which is a general-purpose learning model, that include instructions to extract effective writing styles that resonate with job seekers and are commonly used in multiple similar job postings as "winning patterns" (or instructions to extract ineffective writing styles that do not resonate with job seekers as "losing patterns") for each evaluation axis of specificity, comprehensiveness, appeal, and originality, and cause the second job posting creation model to output winning patterns (or losing patterns) for each evaluation axis. In this case, the second job posting will include a set of winning patterns (or losing patterns) for each of the multiple evaluation axes.

[0067] A "winning pattern" is a pattern obtained, for example, by extracting common writing styles from similar job postings that have a high expected conversion rate (above a predetermined threshold). A "losing pattern" is a pattern obtained, for example, by extracting common writing styles from similar job postings that have a low expected conversion rate (below a predetermined threshold).

[0068] The second job posting creation unit 114 may calculate the degree of sharing (implementation rate) among multiple similar job postings extracted for each item of the second job posting. The "degree of sharing" is, for example, the ratio of the number of similar job postings that contain content that matches or is similar to the content of the second job posting to the total number of similar job postings used to create the second job posting.

[0069] Here, the "items in the second job posting" may be, for example, items in the job posting (e.g., job type, industry, etc.), or they may be evaluation axes (e.g., specificity, comprehensiveness, etc.) used by the analysis unit 115, described later, to create improvement suggestion information. The second job posting creation unit 114 may calculate the degree of sharing for each of the four evaluation axes, such as specificity, comprehensiveness, appeal, and originality. The second job posting creation unit 114 may also create second job postings (winning patterns or losing patterns) only for items where the degree of sharing is above a predetermined threshold (e.g., 50%).

[0070] The second job posting creation unit 114 may input prompts to the second job posting creation model, which is a general-purpose learning model, that include instructions to extract winning patterns (or losing patterns) for each of the multiple evaluation axes mentioned above, as well as instructions to calculate the degree of sharing of the extracted winning patterns (or losing patterns), causing the second job posting creation model to output the winning patterns (or losing patterns) for each evaluation axis and their degree of sharing.

[0071] The second job posting creation unit 114 may create a second job posting in which similar job postings cannot be identified (for example, in which proper nouns are not included). For example, the second job posting creation unit 114 may insert the condition "Create a second job posting in which similar job postings cannot be identified" into the prompts input to the second job posting creation model, which is a general-purpose learning model. More specifically, the second job posting creation unit 114 may insert the instruction "Mask proper nouns to general expressions" into the prompts input to the second job posting creation model, which is a general-purpose learning model, thereby replacing proper nouns such as company names, product names, and service names with general expressions such as "major IT company" or "payment service," and creating a second job posting that does not contain proper nouns.

[0072] The second job posting creation unit 114 may create effect explanation documents that describe the effects on job seekers for each item of the second job posting, based on the second job posting and the fourth reference information (reference information for effect explanation documents). Here, "items of the second job posting" may be items in the job posting form or evaluation axes. For example, the second job posting creation unit 114 may create effect explanation documents for each of the four evaluation axes: specificity, comprehensiveness, appeal, and originality.

[0073] The "Effect Explanation Document" is, for example, a document that explains the "winning patterns" extracted by the Second Job Information Creation Department 114. This explanation may include the basis for the winning patterns (e.g., numerical data), the psychological effects on candidates, etc. Psychological effects include, for example, an approach to concerns that candidates may have (e.g., mention of career paths after joining the company), and expectations that candidates may have (e.g., opportunities for growth, sense of contribution, etc.).

[0074] "Effects on job seekers" include, for example, resolving job seekers' questions, creating a sense of expectation, and having a special appeal to specific targets. In addition to these effects, the effect description text may also include explanations of the content of each item in the second job posting, the basis (facts) for the effect occurring, and the target audience for which the effect occurs.

[0075] The fourth reference information is information regarding the correlation between the second job posting and the effect description text. The fourth reference information is stored, for example, in the memory unit 12. The fourth reference information may include, for example, a table, a function, a simple algorithm, etc., that shows the correlation between the second job posting and the effect description text. The correlation included in the fourth reference information can be constructed, for example, by statistically analyzing data that records the second job posting and the corresponding effect description text.

[0076] The fourth reference information may include a set of parameters for generating an effect description text from the second job information. For example, the fourth reference information may be various pre-trained models. For example, the fourth reference information may include an effect description text generation model which is a dedicated training model or a general-purpose training model that has been machine-trained to take the second job information as input and output an effect description text. In this case, the second job information generation unit 114 inputs the second job information into the effect description text generation model and causes the effect description text generation model to output an effect description text.

[0077] The effect description text generation model is included in the artificial intelligence unit 120. The effect description text generation model, which is a dedicated learning model, may be constructed, for example, by training using data from the second job posting and the corresponding effect description text data as training data. In such an effect description text generation model, parameters calculated and tuned through learning construct a correlation between the second job posting and the effect description text.

[0078] If the effect description text generation model is a general-purpose learning model, the second job information generation unit 114 inputs a prompt to the effect description text generation model that includes the second job information and an instruction to output an effect description text corresponding to the second job information, causing the effect description text generation model to output the effect description text. The second job information generation unit 114 may also generate a prompt that gives the effect description text generation model an instruction to create an effect description text and input this prompt to the effect description text generation model. In addition to the second job information and the instruction to create and output the effect description text, the second job information generation unit 114 may also input a prompt to the effect description text generation model that includes, for example, one or more samples of the second job information and one or more samples of the corresponding effect description text as examples, samples, or training data of input and output pairs. Here, the parameters that construct the effect description text generation model and the prompt that includes an instruction to output an effect description text corresponding to the second job information construct the correlation between the second job information and the effect description text.

[0079] For example, the second job posting creation unit 114 may input prompts to the general-purpose learning model, the effect description document creation model, which outputs information such as: a categorization of winning patterns (numerical data, project phase, scope of work, organizational structure, daily schedule, business challenges, etc.) as an explanation of winning patterns for each of the multiple evaluation axes; and the psychological effects of the winning patterns on job seekers (for example, the job seeker's questions that are resolved (reality after joining the company, career path, interpersonal relationships, etc.) and the expectations provided to job seekers (growth opportunities, sense of contribution, stability, etc.)). The effect description document creation model may then output this information. Furthermore, the second job posting creation unit 114 may input prompts to the effect description document creation model which outputs targets (job types) for which this information is particularly effective, and the effect description document creation model may output the trends of targets for which the winning patterns are effective.

[0080] If the second job posting contains information that allows for the individual identification of similar job postings, the second job posting creation unit 114 may create an effect description document in which similar job postings are not identified (for example, proper nouns are replaced with general expressions). For example, the second job posting creation unit 114 may insert a condition such as "create an effect description document in a way that similar job postings are not identified" or an instruction such as "mask proper nouns with general expressions" into the prompts input to the effect description document creation model, which is a general-purpose learning model.

[0081] <Analysis Department 115> The analysis unit 115 is configured to create improvement suggestion information, including points for improvement of the target job, based on a comparison between the job description text included in the first job information acquired by the first job information acquisition unit 112 and the second job information representing the content of similar jobs extracted by the similar job extraction unit 113. The second job information used to create the improvement suggestion information may be created by the second job information creation unit 114, or it may be the job posting of the similar job itself.

[0082] "Improvement suggestion information" is information that contributes to improving the target job posting. Specifically, improvement suggestion information may include, for example, at least one of the "evaluation" and "revision proposals" of the target job posting, or a combination thereof.

[0083] The "areas for improvement" included in the improvement suggestion information may include, for example, items that are included in the second job posting but not in the job description of the first job posting, or expressions in the second job posting that are more effective for job seekers than those in the job description of the first job posting.

[0084] The analysis unit 115 may, for example, extract the differences between the job description text of the first job posting and the second job posting (particularly the job description text in the second job posting), and create improvement suggestion information that includes these differences as points for improvement.

[0085] The analysis unit 115 may create improvement suggestion information that includes proposed revisions to the target job posting, based on the second job posting, as points for improvement, based on the combination of the job description text included in the first job posting and the second job posting, and the first reference information (reference information for proposed revisions). This allows employers to easily and accurately revise the target job posting by referring to the proposed revisions.

[0086] The "revision proposal" included in the improvement suggestion information may be, for example, a revised version of the job description text, or a document explaining the policy for revising the job description text. The "revision text" may be, for example, a document that fills the gap identified by the analysis department 115 when comparing the winning patterns included in the second job information with the job description text of the first job information, and the "document explaining the policy for revision" may be, for example, a document explaining the policy for revision to fill that gap.

[0087] The first reference information is information relating to the correlation between the combination of the job description text and the second job information and the proposed revisions. The first reference information is stored, for example, in the memory unit 12. The first reference information may include, for example, tables, functions, simple algorithms, etc., that show the correlation between the combination of the job description text and the second job information and the proposed revisions. The correlations included in the first reference information can be constructed, for example, by statistically analyzing data that records the combination of the job description text and the second job information and the corresponding proposed revisions.

[0088] The first reference information may include a set of parameters for generating revised job descriptions from a combination of job description text and second job information. For example, the first reference information may be various pre-trained models. For example, the first reference information may include a revision proposal generation model, which is either a dedicated learning model or a general-purpose learning model, that has been machine-trained to take a combination of job description text and second job information as input and output revised job descriptions. In this case, the analysis unit 115 inputs the combination of job description text and second job information into the revision proposal generation model and causes the revision proposal generation model to output revised job descriptions.

[0089] The revision proposal generation model is included in the artificial intelligence unit 120. The revision proposal generation model, which is a dedicated learning model, may be constructed by training it using, for example, data of combinations of job description texts and second job information, and data of corresponding revision proposals, as training data. In such a revision proposal generation model, parameters calculated and tuned through learning construct a correlation between the combination of job description texts and second job information and the revision proposals.

[0090] If the revision proposal generation model is a general-purpose learning model, the analysis unit 115 inputs a prompt to the revision proposal generation model that includes a combination of job description text and second job information, and an instruction to output a revision proposal corresponding to that combination as input, causing the revision proposal generation model to output a revision proposal. The analysis unit 115 may also generate a prompt that gives the revision proposal generation model an instruction to create a revision proposal, and input this prompt to the revision proposal generation model. In addition to the combination of job description text and second job information and the instruction to create and output a revision proposal, the analysis unit 115 may also input a prompt to the revision proposal generation model that includes, for example, one or more sample combinations of job description text and second job information, and one or more sample revision proposals corresponding to them, as examples, samples, or training data of input and output pairs. Here, the parameters that construct the revision proposal generation model and the prompt that includes an instruction to output a revision proposal corresponding to the combination of job description text and second job information construct a correlation between the combination of job description text and second job information and the revision proposal.

[0091] The analysis unit 115 may create improvement suggestion information that includes the evaluation of the target job as points for improvement, based on the combination of the job description text included in the first job information and the second job information, and the second reference information (evaluation reference information). This allows the employer to consider the parts of the target job that should be modified and the modification policy, etc., based on the evaluation of the target job.

[0092] The "evaluation" included in the improvement suggestion information is determined, for example, by the degree of content satisfaction when comparing the job description in the first job posting with that of the second job posting (particularly the job description in the second job posting). The evaluation may include information indicating the evaluation rank and an evaluation statement showing the comparison result between the job description and the second job posting. This allows employers to easily check the evaluation of the target job and understand the basis for the evaluation.

[0093] "Information indicating the evaluation rank" may include, for example, symbols, numbers, or keywords that indicate a high level of evaluation. The analysis unit 115 may assign evaluation ranks such as "○" to those with a high degree of content satisfaction compared to the second job posting, "△" to those that require improvement, and "×" to those with insufficient content. The evaluation rank may also be a multi-level scale such as "S, A, B" or "5, 4, 3," or a score or percentage such as "80 out of 100 points" or "80% achievement rate." Furthermore, the information indicating the evaluation rank is not limited to text or symbols, but may also be an object such as color (for example, information color-coded according to the evaluation) or a gauge that indicates a high level of evaluation. The evaluation rank may be calculated by comparing the job description text included in the first job posting with the content of similar job postings (second job postings), or by comparing the job description text included in the first job posting with the content of other job postings other than similar job postings.

[0094] The second reference information is information relating to the correlation between the combination of the job description text and the second job information and the evaluation. The second reference information is stored, for example, in the memory unit 12. The second reference information may include, for example, tables, functions, simple algorithms, etc., that show the correlation between the combination of the job description text and the second job information and the evaluation. The correlation included in the second reference information can be constructed, for example, by statistically analyzing data that records the combination of the job description text and the second job information and the corresponding evaluation.

[0095] The second reference information may include a set of parameters for generating an evaluation from a combination of the job description and the second job information. For example, the second reference information may be various pre-trained models. For example, the second reference information may include an evaluation generation model which is a machine learning model that has been trained to take a combination of the job description and the second job information as input and output an evaluation, or a general-purpose machine learning model. In this case, the analysis unit 115 inputs the combination of the job description and the second job information into the evaluation generation model and causes the evaluation generation model to output an evaluation.

[0096] The evaluation creation model is included in the artificial intelligence unit 120. The evaluation creation model, which is a dedicated learning model, may be constructed by training using, for example, data of combinations of job description texts and second job information, and corresponding evaluation data, as training data. In such an evaluation creation model, parameters calculated and tuned through learning construct a correlation between the combination of job description texts and second job information and the evaluation.

[0097] If the evaluation creation model is a general-purpose learning model, the analysis unit 115 inputs a prompt to the evaluation creation model that includes a combination of job description text and second job information, and an instruction to output an evaluation corresponding to that combination, causing the evaluation creation model to output an evaluation. The analysis unit 115 may also generate a prompt that gives the evaluation creation model an instruction to create an evaluation, and input this prompt to the evaluation creation model. In addition to the combination of job description text and second job information and the instruction to create and output an evaluation, the analysis unit 115 may also input a prompt to the evaluation creation model that includes, for example, one or more sample combinations of job description text and second job information, and one or more sample evaluations corresponding to them, as examples, samples, or training data of input and output pairs. Here, the parameters for constructing the evaluation creation model and the prompt that includes an instruction to output an evaluation corresponding to the combination of job description text and second job information construct a correlation between the combination of job description text and second job information and the evaluation.

[0098] For example, the analysis unit 115 may input prompts to the evaluation creation model, which is a general-purpose learning model, that include instructions to determine an evaluation rank based on the degree to which the job description text in the first job posting is implemented or insufficient in relation to the winning pattern in the second job posting (or the degree to which the job description text in the first job posting matches the losing pattern in the second job posting), and instructions to create an evaluation text that quotes the wording of the job description text in the first job posting and compares it with the winning pattern to explain the shortcomings, gaps, etc. (or similarities, common parts, etc. with the losing pattern), causing the evaluation creation model to output an evaluation including an evaluation rank and an evaluation text. Furthermore, the analysis unit 115 may add instructions to the prompts input to the evaluation creation model to create an evaluation that describes the impact on the psychology of target job seekers (e.g., job type) of the gap between the job description text in the first job posting and the winning pattern (or common parts with the losing pattern).

[0099] The analysis unit 115 may create a revised proposal based on the evaluation of the target job posting it has created. For example, the analysis unit 115 may create improvement proposal information, including the revised proposal, based on the job description text included in the first job posting information, the evaluation of the job description text, the combination with the second job posting information, and the first reference information. In this case, the first reference information is information regarding the correlation between the job description text, its evaluation, the combination with the second job posting information, and the revised proposal.

[0100] The analysis unit 115 may, for example, input prompts to a general-purpose learning model, which is a revision proposal generation model, including a combination of a job description text, its evaluation, and a second job information, and an instruction to output a revision proposal corresponding to that combination as input, causing the revision proposal generation model to output the revision proposal.

[0101] For example, the analysis unit 115 may input prompts to the revision proposal creation model, which is a general-purpose learning model, that include creating improvement suggestions to bridge the gap between the job description text and the winning pattern, and creating revised job description texts in line with those improvement suggestions, and have the revision proposal creation model output revised proposals. Furthermore, the analysis unit 115 may add instructions to the prompts input to the revision proposal creation model to create reasons why the revised proposal will resonate with target job seekers (for example, job type) (for example, "Job type XX tends to be XX. Therefore, adding XX to the job description will stimulate the target's desire to take on challenges," or "This revision will help avoid the tendencies (losing patterns) that job postings with low expected results have").

[0102] The analysis unit 115 may create improvement suggestion information for each of the multiple evaluation axes. Each of the multiple evaluation axes includes at least one of the following: specificity, comprehensiveness, appeal, and originality. This allows the recruiter to be presented with improvement points (revision proposals, evaluations, etc.) based on multiple different evaluation axes, making it easier for the recruiter to make revisions that objectively enhance the value of the target job posting.

[0103] Furthermore, in the analysis unit 115, the evaluation axes used when creating an evaluation of the target job posting compared to the second job posting as an area for improvement, and the evaluation axes used when creating a revised proposal for the target job posting based on the second job posting as an area for improvement, may be different from each other. For example, an evaluation may be created for all four evaluation axes (specificity, comprehensiveness, appeal, and originality), and a revised proposal may be created only for the evaluation axes for which the evaluation met the prescribed criteria.

[0104] "Specificity" refers to whether the description in the job posting is specific or not. The criteria for judging specificity may differ depending on the job attributes such as industry, job type, and duties. For example, a criterion for judging specificity may be, "Is the job posting that describes the work detailed enough for job seekers to realistically imagine what it would be like to work?"

[0105] "Comprehensiveness" refers to whether the job description includes the required basic items (for example, company introduction, business introduction, recruitment background, job description (short-term to medium- to long-term), skills to be acquired, attractiveness of the position, etc.).

[0106] "Appeal" refers to whether the job description includes elements that make it attractive to job seekers or a value proposition. "Attractiveness" includes, for example, the social impact of the business, personal growth, technological challenges, and organizational culture. "Value proposition" includes, for example, strategic descriptions that address job seekers' pain points, competition for recruitment, and unique value propositions; a compelling narrative (a grand purpose) that goes beyond a simple list of tasks (e.g., "We handle everything from XX to XX in one go"); a narrative that conveys the background, challenges, solutions, and significance (e.g., "We are challenging the XX of our client's industry with our unique XX as a weapon, and the XX that will result from solving it will be significant. We are looking for colleagues to join us in this challenge"); and information that alleviates potential pain points for job seekers (e.g., "Performance is not fairly evaluated") and concerns about changing jobs (e.g., "Onboarding after joining the company"). Note that the value proposition may also be an EVP (Employee Value Proposition).

[0107] "Uniqueness" refers to whether the job description includes unique perspectives that will attract job seekers beyond the standard basic items (for example, specific experiences of team members, lessons learned from failures, example daily schedules, details of future business plans, etc.).

[0108] For example, the analysis unit 115 may input prompts to a general-purpose learning model, such as a modification proposal creation model or a value creation model, including the names of the evaluation axes and instructions to create evaluations or modification proposals for each evaluation axis, causing the modification proposal creation model or value creation model to output evaluations or modification proposals for each evaluation axis. The modification proposal creation model and the value creation model may be an integrated single improvement proposal model. For example, the analysis unit 115 may input prompts to an improvement proposal model including the names of the evaluation axes and instructions to create evaluations and modification proposals for each evaluation axis, causing the improvement proposal model to output evaluations and modification proposals for each evaluation axis.

[0109] The analysis unit 115 may create improvement suggestion information for a composite evaluation axis that combines multiple evaluation axes (for example, "specificity and comprehensiveness"). Alternatively, the analysis unit 115 may create improvement suggestion information for each sub-evaluation axis that subdivides one evaluation axis into multiple axes (for example, "embodiment of EVP" and "attractiveness of the position").

[0110] The analysis unit 115 may input prompts into the revision proposal creation model, value creation model, or improvement proposal model that include instructions to determine whether to refer to and incorporate the appeal of the second job posting (e.g., the attractiveness of the position) as an evaluation axis, or to differentiate using a different axis, and to create an evaluation or revision proposal.

[0111] <Improvement Information Display Control Unit 116> The improvement information display control unit 116 is configured to display the improvement suggestion information created by the analysis unit 115 on the job seeker terminal 20. However, the second job information itself does not necessarily have to be presented to the job seeker. In other words, the improvement information display control unit 116 does not need to display the second job information itself on the job seeker terminal 20.

[0112] The improvement information display control unit 116 may display the first job posting information and the improvement suggestion information in a way that allows for comparison. This makes it easier for employers to revise the target job posting by allowing them to check the improvement points (suggested revisions, evaluations, etc.) while comparing them with the original first job posting information (job description text).

[0113] If improvement suggestion information has been created for each of the multiple evaluation axes, the improvement information display control unit 116 may display the job description text of the first job posting and the corresponding improvement suggestion information (revision proposal or evaluation) side by side on the job seeker terminal 20 for each evaluation axis.

[0114] The improvement information display control unit 116 displays the second job information created by the second job information creation unit 114 on the recruiter terminal 20 for each of several items, and may also display the degree of sharing among multiple similar job postings on the recruiter terminal 20 for each item. This allows recruiters to check the implementation rate (degree of sharing) of the content (winning pattern) used to create the improvement suggestion information for the target job posting in similar job postings, thereby increasing the persuasiveness of the improvement suggestion information. The degree of sharing is calculated by the second job information creation unit 114 as described above.

[0115] For example, the improvement information display control unit 116 may display on the recruiter terminal 20, for each of the multiple evaluation axes, the content corresponding to the evaluation axis of the second job posting (winning pattern) and the degree to which that content is shared.

[0116] The improvement information display control unit 116 may display the degree of sharing along with the effect explanation text created by the second job information creation unit 114 on the job seeker terminal 20 for each item (for example, evaluation axis). This increases the reliability of the second job information (winning pattern) presented to the job seeker.

[0117] The improvement information display control unit 116 may display on the employer terminal 20 the result of comparing the expected success rate of the target job posting with the expected success rate of similar job postings extracted by the similar job posting extraction unit 113. The expected success rate is a value that indicates the likelihood of a job posting being filled, calculated based on the actions of the employer or job seeker regarding the job posting. This makes it possible to quantitatively present to the employer the need to improve the target job posting.

[0118] "Actions by employers or job seekers regarding job postings" include "actions by employers towards job seekers" and "actions by job seekers towards job postings." "Actions by employers towards job seekers" are synonymous with "actions by job seekers" in the similar job posting extraction unit 113. "Actions by job seekers towards job postings" include, for example, viewing job postings, adding job postings to a bookmark list, applying for jobs, and replying to recruitment documents sent by employers based on job postings. The "bookmark list" here is a list prepared for each job seeker, where job seekers can register any job postings they wish to use.

[0119] The expected conversion rate is an indicator calculated, for example, by statistical processing, normalization, etc., of the number of actions (one or more) over a specified period. The expected conversion rate may be the number of actions themselves. Alternatively, the expected conversion rate may be a numerical value representing actual recruitment results, such as the number of successful placements.

[0120] If the similar job extraction unit 113 extracts multiple similar job postings, the "expected success rate of similar job postings" is, for example, a statistical value (mean, median, maximum, minimum, etc.) of the expected success rates of multiple similar job postings.

[0121] The expected conversion rate can also be the response rate to recruitment letters sent to job seekers. This allows employers to identify improving the response rate to recruitment letters as a specific challenge, thereby encouraging improvements to increase the recruitment conversion rate.

[0122] The "response rate" is the ratio of the number of responses to a recruitment letter sent for a job posting to the total number of recruitment letters sent for that job posting (i.e., the number of job seekers to whom a recruitment letter was sent for a job posting divided by the number of job seekers who responded).

[0123] The "comparative results" are displayed on the recruiter terminal 20 in the form of objects such as tables or graphs that arrange the expected success rates of each job in a comparable manner, text containing each expected success rate, or objects or text representing the differences (differences, ratios, etc.) between each expected success rate. The results may also include an evaluation of the expected success rate of the target job (whether it is a negative or positive number compared to the expected success rates of similar job postings).

[0124] The control unit 11 may create improvement suggestion information, etc., based on the comparison result between the expected value of the target job opening and the expected value of the similar job openings extracted by the similar job opening extraction unit 113. For example, if the expected value of the similar job openings is higher than the expected value of the target job opening, or if the difference between the expected value of the similar job openings and the expected value of the target job opening is greater than or equal to a predetermined value, the second job information creation unit 114 may create an effect explanation document, the analysis unit 115 may create an evaluation, and the revision proposal may be created, etc.

[0125] Figure 6 shows an example of the response rate display screen RD displayed on the recruiter terminal 20. The response rate display screen RD displays the response rate to the scout document for the target job, the response rate (average value) to the scout documents for similar job, and a text that includes the difference between these two, as a comparison of the expected conversion value between the target job and similar job. In addition, in the example in Figure 6, the response rate display screen RD also displays a text that explains the summary (policy) of the improvement suggestion.

[0126] Figure 7 shows an example of the evaluation display screen ED displayed on the recruiter terminal 20. The evaluation display screen ED displays multiple evaluation axes EA, and for each evaluation axis EA, the evaluation rank ER and the evaluation text ES are displayed. In the example in Figure 7, the evaluation rank ER is displayed with symbols such as "○", "△", and "×".

[0127] Figure 8 shows an example of a similar job display screen SD displayed on the recruiter terminal 20. The similar job display screen SD displays multiple evaluation axes EA that are common to the evaluation display screen ED in Figure 7, and for each evaluation axis EA, the degree of sharing DS and the effect description text RS are displayed.

[0128] Figure 9 shows an example of the revised proposal display screen CD displayed on the recruiter terminal 20. The revised proposal display screen CD displays multiple evaluation axes EA, which are the same as the evaluation display screen ED in Figure 7. For each evaluation axis EA, the revised policy CP, the original text OS (text before revision) included in the first job posting, and the revised text CS are displayed. The original text OS and the revised text CS are displayed side by side for comparison.

[0129] The order in which the response rate display screen RD in Figure 6, the evaluation display screen ED in Figure 7, the similar job posting display screen SD in Figure 8, and the proposed revision display screen CD in Figure 9 are displayed on the recruiter terminal 20 does not matter. In other words, the improvement information display control unit 116 may display these screens on the recruiter terminal 20 in any order (for example, in the order instructed by the recruiter).

[0130] <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 each functional unit of the server device 10 may be common to all units, or it may be prepared individually for each functional unit.

[0131] The artificial intelligence unit 120 may be an AI (Artificial Intelligence) equipped with pre-trained models such as transformers including GPT (Generative Pretrained Transformer, including GPT-1 to GPT-5), BERT (Bidirectional Encoder Representations from Transformers), BART (Bidirectional and Auto-regressive Transformer), and language models such as recurrent neural networks (RNNs). The artificial intelligence unit 120 may be, for example, a general-purpose learning model including various language models, large-scale language models, and generative AI, or an AI agent, and may include specific models such as OpenAI's GPT, Google's Gemini, and models provided through services and platforms such as Microsoft's Azure AI Studio. Generative AI may be, for example, text generation AI, image generation AI, or multimodal generation AI. The pre-trained model may be called an artificial intelligence model, machine learning model, or deep learning model. In addition, the artificial intelligence unit 120 can include any pre-trained model.

[0132] Specific machine learning algorithms used to build trained models include nearest neighbors, naive Bayes, decision trees, support vector machines, and deep learning using neural networks. The artificial intelligence unit 120 can apply these algorithms as appropriate.

[0133] The artificial intelligence unit 120 may have a trained model constructed by a learning method such as supervised learning, unsupervised learning, or self-supervised learning. In supervised learning, machine learning is performed using training data. Training data consists of pairs of input data and output data (correct answer data) for training. Furthermore, the trained model may not only be one trained for a specific task, but also a general-purpose learning model that can be used universally for a wide range of tasks.

[0134] The artificial intelligence unit 120 may include a natural language model as artificial intelligence, or it may be a general-purpose learning model such as a Large Language Model (LLM). An LLM is a learning model that has been pre-trained on a large amount of large data consisting of text data, etc. (for example, (i) web content on the internet, or (ii) data stored in a predetermined database), and can perform various language processing tasks by being given a task. According to the given prompt, it can perform a wide range of natural language processing tasks, such as understanding sentence patterns and context, responding to questions, and generating sentences. Such a general-purpose learning model may include a pre-trained model that can handle various tasks without fine-tuning by One-shot Learning or Few-shot Learning. Furthermore, the general-purpose learning model may also be configured to handle various tasks by Zero-shot Learning. The artificial intelligence used in each functional unit of the control unit 11 may be a separate pre-trained model, or it may be a common general-purpose pre-trained model. In addition, the artificial intelligence unit 120 may include a small-scale language model or a medium-scale language model that is smaller in scale than a large-scale language model as a pre-trained model. Small-scale and medium-scale language models are natural language processing models that are trained on less data (and constructed with fewer parameters) compared to large-scale language models.

[0135] The pre-trained models included in the artificial intelligence unit 120 (such as the model for creating revised proposals, which are used in each functional unit) can undergo additional training using methods such as transfer learning and fine-tuning. For example, whenever new data is registered, the artificial intelligence unit 120 may perform additional training and fine-tuning using this new data as training data. This improves the accuracy of the information output from the pre-trained models.

[0136] The trained model included in the artificial intelligence unit 120 may be a trained model (distilled model) obtained by knowledge distillation using the original trained model. In knowledge distillation, a trained model such as a large-scale language model is used as the teacher model, and the student model is trained by adjusting the parameters of the student model so that the loss of the student model's output relative to the teacher model's output (Soft Target Loss) is small, and that student model becomes the distilled model. Alternatively, the student model may be trained so that the loss of the student model's output relative to the correct labels (Hard Target) of the teacher data (combinations of input and output data of the training model) is small. Compared to the original training model (teacher model), the distilled model has performance close to that of the trained model, but with fewer parameters and a lower processing load. Therefore, by using the distilled model, the cost of the information processing system 1 can be reduced.

[0137] For example, the trained model used in each functional unit may be a distilled model trained using combinations of input and output data from a large-scale language model as training data. Alternatively, when the information processing system 1 is introduced, a large-scale language model may be used as the trained model in each functional unit, and once training data from the large-scale language model has been accumulated, the distilled model obtained by knowledge distillation using that training data may be used as the trained model in each functional unit.

[0138] An AI agent (also called an autonomous agent) is a model that, upon input of a goal (objective, purpose, etc.) such as "Teach me about XX" or a task such as "Output XX," breaks down the processes necessary to reach the goal or accomplish the task into subtasks, actions, etc., and performs necessary data collection and analysis, program generation and execution, etc. The AI ​​agent takes the information and instructions input by the user as its goal, autonomously selects and executes tasks and actions according to the goal, outputs information according to the goal, and does not require user intervention (operation input). Furthermore, the AI ​​agent may autonomously plan and execute, evaluate the execution results itself, and autonomously learn in order to aim for goal achievement. For example, the AI ​​agent may autonomously update itself based on the execution results of subtasks (e.g., collected information, results of information analysis, etc.).

[0139] <Display section> The display unit 211 of the job seeker terminal 20 shown in Figure 4B, and the display unit 311 of the job seeker terminal 30 shown in Figure 4C, respectively, display the screen (information) indicated by the data transmitted from the server device 10.

[0140] <Operation acquisition part> The operation acquisition unit 212 of the employer terminal 20 receives operations from the employer using the employer terminal 20. The operation acquisition unit 312 of the job seeker terminal 30 receives operations from the job seeker using the job seeker terminal 30.

[0141] 3. Information Processing Methods This section describes the information processing method of the server device 10. In this information processing method, each part of the server device 10 is executed by a computer as a step.

[0142] The above-described information processing method comprises a first job information acquisition step, a similar job extraction step, an analysis step, and an improvement information display control step. In the first job information acquisition step, first job information is acquired, which includes a job description text about the organization or business of the target job. In the similar job extraction step, similar job postings similar to the target job are extracted from the job postings registered in the database based on the first job information. In the analysis step, improvement suggestion information, which includes points for improvement of the target job, is created based on a comparison between the job description text and second job information representing the content of the similar job postings. In the improvement information display control step, the improvement suggestion information is displayed.

[0143] Figure 10 is an activity diagram showing an example of the flow of information processing (display processing of improvement suggestion information) performed by the information processing system 1. Below, the information processing will be explained according to each activity in this activity diagram.

[0144] The process of displaying improvement suggestion information begins with the employer selecting the target job. The employer selects the target job for which improvement suggestions will be received (entering the information of the target job) on the employer terminal 20 (Activity A101). The server device 10 obtains the first job information based on the information of the target job entered from the employer terminal 20 (Activity A102). Subsequently, the server device 10 extracts similar job information based on the first job information (Activity A103).

[0145] After extracting similar job postings, the server device 10 creates improvement suggestion information based on the job description text included in the first job posting information and the second job posting information (Activity A104). The creation of improvement suggestion information here may include at least one of the following: calculation of expected conversion value, calculation of degree of sharing, creation of effect explanation text, creation of evaluation, and creation of revised proposals. Subsequently, the server device 10 outputs the created improvement suggestion information to the job seeker terminal 20 (Activity A105). As a result, the improvement suggestion information is displayed on the job seeker terminal 20 (Activity A106).

[0146] 4. Effect The function of this embodiment can be summarized as follows: it enables the creation of high-quality job postings.

[0147] Although embodiments of the present invention have been described above, the present invention is not limited thereto and can be modified as appropriate without departing from the technical spirit of the invention.

[0148] 5. Others In the above embodiment, the server device 10 performed various storage and control functions, but instead of the server device 10, multiple external devices may be used. That is, various information and programs may be stored in a distributed manner across multiple external devices using blockchain 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 external artificial intelligence unit 120 may be provided by, for example, an artificial intelligence service server, and is configured to receive input from each functional unit of the server device 10, receive requests to execute artificial intelligence services, and return the instructed output as a processing result to the server device 10. The artificial intelligence service server may be a server that provides services using a language model as a learning model, or a server that executes language processing tasks using a language model. The artificial intelligence service server may be constructed using an LLM. The artificial intelligence service server receives prompt input in the form of text, images, audio, etc., and generates and responds with answers to the prompts.

[0149] At least one of the devices included in the information processing system 1 may be located outside the country in which the functions of the information processing system 1 are performed.

[0150] The embodiments of this model are not limited to the information processing system 1, but may also be an information processing method or a program. In the information processing method, the information processing device executes each step of the information processing system 1. In the program, the computer causes the computer to execute each step of the information processing system 1.

[0151] The control unit 11 does not necessarily have to include a second job information creation unit 114. In other words, the information processing system 1 does not need to have a function to create second job information based on the content of multiple similar job postings.

[0152] The product may be provided in any of the following embodiments.

[0153] (1) An information processing system comprising at least one processor, wherein the processor is configured to perform the following steps by reading a program, the first job information acquisition step involves acquiring first job information including a job description document relating to the organization or business of the target job; the similar job extraction step involves extracting similar job postings similar to the target job from among job postings registered in the database based on the first job information; the analysis step involves creating improvement suggestion information including points for improvement of the target job based on a comparison between the job description document and second job information representing the content of the similar job postings; and the improvement information display control step involves displaying the improvement suggestion information.

[0154] (2) An information processing system as described in (1) above, wherein in the analysis step, based on the combination of the job description text and the second job information and the first reference information, the system creates the improvement proposal information which includes a proposed revision of the target job that is based on the second job information as the improvement points, and the first reference information is information relating to the correlation between the combination of the job description text and the second job information and the proposed revision.

[0155] (3) An information processing system as described in (1) or (2) above, wherein the analysis step creates improvement suggestion information that includes an evaluation of the target job compared with the second job information as improvement points, based on the combination of the job description text and the second job information and the second reference information, wherein the second reference information is information relating to the correlation between the combination of the job description text and the second job information and the evaluation.

[0156] (4) An information processing system as described in (3) above, wherein the evaluation includes information indicating an evaluation rank and an evaluation document indicating the result of comparing the job description document with the second job information.

[0157] (5) An information processing system according to any one of the above items (1) to (4), wherein the improvement information display control step displays the first job information and the improvement proposal information in a comparative manner.

[0158] (6) An information processing system according to any one of (1) to (5) above, wherein in the similar job extraction step, a plurality of the aforementioned similar job openings are extracted, and in the second job information creation step, the second job information is created based on the contents of the plurality of the aforementioned similar job openings and the third reference information, wherein the third reference information is information relating to the correlation between the contents of the plurality of the aforementioned similar job openings and the second job information.

[0159] (7) An information processing system as described in (6) above, wherein in the improvement information display control step, the second job information is displayed for each of the items, and the degree of sharing among the multiple similar job postings is displayed for each of the items.

[0160] (8) An information processing system as described in (7) above, wherein in the second job information creation step, based on the second job information and the fourth reference information, an effect description document is created that explains the effect of each item of the second job information on job seekers, where the fourth reference information is information relating to the correlation between the second job information and the effect description document, and in the improvement information display control step, the effect description document is displayed for each item along with the degree of sharing.

[0161] (9) An information processing system according to any one of (1) to (8) above, wherein the analysis step creates the improvement suggestion information for each of the multiple evaluation axes, and each of the multiple evaluation axes includes at least one of specificity, comprehensiveness, appeal, and originality.

[0162] (10) An information processing system according to any one of (1) to (9) above, wherein in the first job information acquisition step, the system accepts input of information indicating the target job and a specific keyword, acquires the first job information and the specific keyword for the target job, and in the similar job extraction step, based on the first job information and the specific keyword, extracts similar job postings from the job postings registered in the database that are similar to the target job and related to the specific keyword.

[0163] (11) An information processing system as described in (10) above, wherein in the similar job extraction step, a plurality of candidate job postings similar to the target job posting are extracted based on the first job posting information, and further, from the plurality of candidate job postings, the similar job postings are extracted according to the relationship between the content of the candidate job postings and the specific keywords.

[0164] (12) An information processing system according to any one of (1) to (11) above, wherein in the similar job extraction step, the information processing system extracts job postings that are similar to the target job posting and whose number of actions toward job seekers satisfies the extraction criteria as the similar job posting.

[0165] (13) An information processing system according to any one of (1) to (12) above, wherein the improvement information display control step displays the result of comparing the expected value of the target job with the expected value of the similar job, and the expected value of the job is a value that indicates the likelihood of the job being filled, calculated based on the actions of the employer or job seeker regarding the job.

[0166] (14) An information processing system as described in (13) above, wherein the expected value of contracts is the response rate to the recruitment documents sent to job seekers.

[0167] (15) An information processing system according to any one of (1) to (14) above, comprising a server device having the processor and a terminal that can access the server device.

[0168] (16) An information processing method wherein an information processing device performs each step of the information processing system described in any one of (1) to (15) above.

[0169] (17) A program that causes a computer to perform each step of the information processing system described in any one of (1) to (15) above. Of course, this is not always the case.

[0170] Finally, while various embodiments relating to this disclosure have been described, these are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0171] 1: Information Processing System 2: Communication lines 10: Server device 11: Control Unit 111: Basic Display Control Unit 112: 1st Recruitment Information Acquisition Department 113: Similar job offer extraction department 114: Second Job Posting Creation Department 115:Analysis Department 116: Improvement Information Display Control Unit 120: Artificial Intelligence Department 12: Storage section 13: Communications Department 14: Communications bus 20: Job seeker terminal 21: Control Unit 211:Display section 212: Operation acquisition section 22: Storage section 23: Communications Department 24: Input section 25: Output section 26: Communications bus 30: Job seeker terminal 31: Control Unit 311: Display section 312: Operation acquisition section 32: Storage section 33: Communications Department 34: Input section 35: Output section 36: Communications bus B11: Close button B12: Create button CD: Amendment display screen CP: Correction policy CS: Corrected text DS: Degree of sharing EA: Evaluation axis ED: Evaluation display screen ER: Rating Rank ES: Evaluation document ID: Job posting screen KF: Keyword input field OS: Original text RD:Reply rate display screen RF: Job posting input field RS: Effect description text SD:Similar job offer display screen

Claims

1. An information processing system, Equipped with at least one processor, The aforementioned processor is configured to perform the following steps by reading a program: In the first job information acquisition step, the first job information, including a job description document regarding the organization or duties of the target job, is acquired. In the similar job extraction step, similar job postings that are similar to the target job posting are extracted from the job postings registered in the database based on the first job posting information. In the analysis step, the combination of the job description text and the second job information representing the content of similar job postings is input into the revision proposal creation model, and the revision proposal creation model is made to output a revised version of the target job posting based on the second job information, thereby creating improvement suggestion information that includes the revised version as improvements to the target job posting. Here, the revision proposal creation model is a dedicated learning model or a general-purpose learning model that has been machine-trained to take the combination of the job description text and the second job information as input and output the revised version. The improvement information display control step involves an information processing system that displays the improvement suggestion information.

2. An information processing system, Equipped with at least one processor, The aforementioned processor is configured to perform the following steps by reading a program: In the first job information acquisition step, the first job information, including a job description document regarding the organization or duties of the target job, is acquired. In the similar job extraction step, similar job postings that are similar to the target job posting are extracted from the job postings registered in the database based on the first job posting information. In the analysis step, the combination of the job description text and the second job information representing the content of similar job postings is input to the evaluation creation model, and the evaluation creation model outputs an evaluation of the target job posting compared with the second job posting information, thereby creating improvement suggestion information that includes the evaluation as points for improvement of the target job posting. Here, the evaluation creation model is a dedicated learning model or a general-purpose learning model that has been machine-trained to take the combination of the job description text and the second job information as input and output the evaluation. The improvement information display control step involves an information processing system that displays the improvement suggestion information.

3. In the information processing system described in claim 2, The aforementioned evaluation includes information indicating an evaluation rank and an evaluation document showing the results of comparing the job description document with the second job information, in an information processing system.

4. In the information processing system described in claim 1, The information processing system in the improvement information display control step displays the first job information and the improvement proposal information in a way that allows for comparison.

5. An information processing system, Equipped with at least one processor, The aforementioned processor is configured to perform the following steps by reading a program: In the first job information acquisition step, the first job information, including a job description document regarding the organization or duties of the target job, is acquired. In the similar job extraction step, based on the first job information, multiple similar job postings that are similar to the target job posting are extracted from the job postings registered in the database. In the second job posting creation step, the contents of multiple similar job postings are input into the second job posting creation model, and the second job posting creation model is made to output a second job posting containing the contents extracted from each of the multiple similar job postings. Here, the second job posting creation model is a dedicated learning model or a general-purpose learning model that has been machine-trained to take the contents of multiple similar job postings as input and output the second job posting. In the analysis step, the differences between the job description and the second job information are extracted, and improvement suggestion information is created that includes the differences as points for improvement for the target job. The improvement information display control step involves an information processing system that displays the improvement suggestion information.

6. In the information processing system described in claim 5, An information processing system that, in the improvement information display control step, displays the second job information for each of several items, and for each item displays a degree of sharing, which is calculated as the ratio of the number of similar job postings that contain content that matches or is similar to the content of the second job information to the total number of similar job postings used to create the second job information.

7. In the information processing system described in claim 6, In the second job information creation step, the second job information is input into the effect description document creation model, and the effect description document creation model is made to output effect description documents explaining the effects of each item of the second job information on job seekers. Here, the effect description document creation model is a dedicated learning model or a general-purpose learning model that has been machine-trained to take the second job information as input and output the effect description documents. The information processing system in the improvement information display control step displays the effect description text along with the degree of sharing for each item.

8. In the information processing system described in claim 1, In the analysis step, the system inputs a prompt to the revision proposal creation model, which includes a combination of the job description text and the second job information, and an instruction to create the revision proposal for each of the multiple evaluation axes, thereby creating the improvement suggestion information for each of the multiple evaluation axes, wherein each of the multiple evaluation axes includes at least one of specificity, comprehensiveness, appeal, and originality.

9. In the information processing system described in Claim 2, In the analysis step, the system inputs a prompt to the evaluation creation model that includes a combination of the job description text and the second job information, and an instruction to create the evaluation for each of the multiple evaluation axes, thereby creating the improvement suggestion information for each of the multiple evaluation axes, wherein each of the multiple evaluation axes includes at least one of specificity, comprehensiveness, appeal, and originality.

10. In the information processing system described in claim 1, In the first job information acquisition step, the input of information indicating the target job and a specific keyword is accepted, and the first job information and the specific keyword for the target job are acquired. An information processing system that, in the similar job extraction step, extracts similar job postings from the job postings registered in the database that are similar to the target job posting and related to the specific keyword, based on the first job posting information and the specific keyword.

11. In the information processing system according to claim 10, An information processing system that, in the similar job extraction step, extracts multiple candidate job postings similar to the target job posting based on the first job posting information, and further extracts similar job postings from among the multiple candidate job postings according to the relationship between the content of the candidate job postings and the specific keywords.

12. In the information processing system described in claim 1, An information processing system that, in the similar job search step, extracts job postings that are similar to the target job posting and whose number of actions towards job seekers meets the extraction criteria as similar job postings.

13. An information processing system, Equipped with at least one processor, The aforementioned processor is configured to perform the following steps by reading a program: In the first job information acquisition step, the first job information, including a job description document regarding the organization or duties of the target job, is acquired. In the similar job extraction step, similar job postings that are similar to the target job posting are extracted from the job postings registered in the database based on the first job posting information. In the analysis step, the differences between the job description text and the second job information representing the content of the similar job are extracted, and improvement suggestion information is created that includes the differences as points for improvement of the target job. In the improvement information display control step, the improvement suggestion information and the result of comparing the first expected contract value for the target job posting with the second expected contract value for similar job postings are displayed. The first expected value of contracts is an index calculated by statistical processing or normalization of the number of first actions during a predetermined period, and the first action is at least one of the following: the job seeker viewing the job posting for the target job, the job seeker adding the target job to their bookmark list, the job seeker applying for the target job, and the job seeker responding to a scouting document sent to the job seeker based on the target job. The information processing system wherein the second expected value of the contract is an index calculated by statistical processing or normalization of the number of second actions during a predetermined period, and the second action is at least one of the following: a job seeker viewing a job posting for a similar job; a job seeker adding a similar job to their bookmark list; a job seeker applying for a similar job; and a job seeker responding to a scouting document sent to the job seeker based on the similar job.

14. In the information processing system described in claim 13, An information processing system in which the first expected contract value and the second expected contract value are, respectively, the response rate to the scouting document.

15. In the information processing system described in claim 1, A server device having the aforementioned processor, A terminal that can access the aforementioned server device, An information processing system equipped with the following features.

16. Information processing method, An information processing method comprising an information processing device performing each step of the information processing system described in any one of claims 1 to 15.

17. It is a program, A program for causing a computer to perform each step of the information processing system described in any one of claims 1 to 15.

Images