Job seeking support system, job seeking support method, and program

Abstract

To provide a job seeking support system, method and program for presenting recruitment information to a job seeker based on the interest of the job seeker.SOLUTION: In a job seeking support system 1, a server apparatus executes: an activity A102 which acquires information on candidate recruitment selected by a job seeker, out of recruitment information registered on a database; an activity A103 which estimates at least one recommendation category based on the contents of the candidate recruitment and first reference information; an activity A104 which extracts at least one recommendation recruitment to be recommended to the job seeker, out of the recruitment information registered on the database, based on the recommendation category or a recommendation keyword regarding the recommendation category, and a second reference information; and an activity A105 which presents the recommendation category and the recommendation recruitment in association with each other, to the job seeker.SELECTED DRAWING: Figure 8

Description

[Technical Field]

[0001] The present invention relates to a job search support system, a job search support method, and a program. [Background technology]

[0002] As disclosed in Patent Document 1, a technique is known in which a job seeker searches for a job offer that matches their desired conditions based on job information registered by a recruiter. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-269220 Summary of the Invention [Problem to be solved by the invention]

[0004] In such technology, job seekers are required to input keywords and the like as search conditions to search for desired job openings.

[0005] In view of the above circumstances, the present invention provides a job-seeking support system that can present job offers to job seekers based on the interests of the job seekers. [Means for solving the problem]

[0006] According to one aspect of the present invention, a job search support system is provided. The job search support system includes at least one processor. The processor is configured to execute the following steps by reading a program. In the acquisition step, information on candidate job openings selected by the job seeker is acquired from job openings registered in a database. In the estimation step, at least one recommended category is estimated based on the content of the candidate job openings and first reference information. The recommended category is a category in which the job seeker is interested, among categories for classifying information included in job openings based on attributes, and the first reference information includes a correlation between the content of the candidate job openings and the recommended category. In the extraction step, at least one recommended job opening to be recommended to the job seeker is extracted from the job openings registered in the database based on the recommended category or recommended keywords related to the recommended category and second reference information. The second reference information includes a correlation between the recommended category or recommended keywords and the recommended job opening. In the presentation step, the recommended category and the recommended job opening are associated and presented to the job seeker.

[0007] According to this embodiment, job offers that meet the job seeker's desired conditions can be presented based on the job seeker's selection (viewing, bookmarking, etc.) of job offers. Therefore, job offers can be presented using an approach that differs from the search conditions entered by the job seeker. Furthermore, by presenting recommended categories, the job seeker can grasp his or her desired conditions that have not been verbalized. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a configuration diagram illustrating a job search support system 1. FIG. [Figure 2] 2 is a block diagram showing the hardware configuration of the server device 10. FIG. [Figure 3] FIG. 2 is a block diagram showing the hardware configuration of a recruiting party terminal 20 and a job seeker terminal 30. [Figure 4] 1 is a block diagram showing functions realized by a server device 10 (control unit 11), a recruiting party terminal 20 (control unit 21), and a job seeker terminal 30 (control unit 31). [Figure 5]10 is a diagram showing an example of a recommended job offer presentation screen RD displayed on the job seeker terminal 30. FIG. [Figure 6] FIG. 10 is a diagram showing an example of a job search screen SD displayed on the job seeker terminal 30. [Figure 7] 10 is a diagram showing an example of a scout document list screen LD displayed on the job seeker terminal 30. FIG. [Figure 8] 1 is an activity diagram showing the flow of information processing (recommended job offer presentation processing) executed by job seeking support system 1. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings. Various features shown in the following embodiments can be combined with each other.

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

[0011] Furthermore, various information processing according to an embodiment may realize input and output corresponding to the input. Here, the form of information referenced in such information processing (hereinafter referred to as reference information) is not limited as long as an output is obtained as a result of the input. 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 formula constructed using a statistical method), a trained model that has previously trained the correlation between input and output, or a large-scale language model that can output a desired result by inputting a prompt.

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

[0013] Furthermore, a circuit in the broad sense is a circuit realized by at least an appropriate combination of a circuit, circuitry, processor, memory, etc. The processor may be a general-purpose processor or a dedicated circuit. That is, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

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

[0015] <Job Search Support System 1> Fig. 1 is a configuration diagram showing a job search support system 1. The job search support system 1 comprises a communication line 2, a server device 10, a plurality of recruiter terminals 20, and a plurality of job seeker terminals 30. The server device 10, the recruiter terminals 20, and the job seeker terminals 30 are configured to be able to communicate with each other via the communication line 2. The connection between the server device 10, the recruiter terminals 20, and the job seeker terminals 30 may be wired or wireless.

[0016] Job search support system 1 constitutes part of a recruitment and job search system used by multiple recruiters (first recruiter U1 and second recruiter U2) and multiple job seekers (first job seeker U3 and second job seeker U4). Job search support system 1 mainly manages job seeker registration information and job postings. In one embodiment, job search support system 1 is made up of one or more devices or components. These components are described below.

[0017] <Server device 10> 2 is a block diagram showing the hardware configuration of server device 10. Server device 10 includes a control unit 11, a storage unit 12, a communication unit 13, and a communication bus 14. Control unit 11, storage unit 12, and communication unit 13 are electrically connected within server device 10 via communication bus 14.

[0018] <Control unit 11> The control unit 11 processes and controls the overall operations related to the server device 10. The control unit 11 is, for example, a central processing unit (CPU). The control unit 11 realizes various functions related to the server device 10 by reading out predetermined programs stored in the storage unit 12. In other words, 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 multiple control units 11 may be provided for each function. A combination of these may also be used.

[0019] <Storage section 12> The memory unit 12 stores various pieces of information defined above. This can be implemented, for example, as a storage device such as a solid state drive (SSD) that stores various programs and the like related to the server device 10 executed by the control unit 11, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the program operations. The memory 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 is preferably a wired communication means such as USB, IEEE1394, Thunderbolt (registered trademark), or wired LAN network communication, but may also include wireless LAN network communication, mobile communication such as 3G / LTE / 5G, or BLUETOOTH (registered trademark) communication as needed. That is, it is more preferable to implement it as a collection of multiple communication means. That is, the server device 10 may communicate various information from the outside via the communication unit 13 and the network.

[0021] The server device 10 may be an on-premise server or a cloud server. The cloud server device 10 may provide the above-described functions and processes in the form of, for example, SaaS (Software as a Service) or cloud computing.

[0022] <Recruiter Terminal 20> Fig. 3 is a block diagram showing the hardware configuration of the recruiting party terminal 20 and the job seeker terminal 30. As shown in Fig. 3A, the recruiting party terminal 20 comprises a control unit 21, a memory unit 22, a communication unit 23, an input unit 24, an output unit 25, and a communication bus 26. The control unit 21, the memory unit 22, the communication unit 23, the input unit 24, and the output unit 25 are electrically connected within the recruiting party terminal 20 via the communication bus 26. The explanation of the control unit 21, the memory unit 22, and the communication unit 23 is the same as the explanation of each unit in the server device 10, and will therefore be omitted. Note that the recruiting party terminal 20 may also be a terminal operated by a recruitment agency that interacts with job seekers on behalf of the recruiter.

[0023] <Input section 24> The input unit 24 accepts operation inputs made by the user. The operation inputs are transferred as command signals to the control unit 21 via the communication bus 26. The control unit 21 can execute predetermined control or calculations based on the transferred command signals as necessary. The input unit 24 may be included in the housing of the recruiter terminal 20 or may be attached externally. For example, the input unit 24 may be implemented as a touch panel integrated with the output unit 25. When the input unit 24 is implemented as a touch panel, the user can input tap operations, swipe operations, etc. to the input unit 24. Instead of a touch panel, a switch button, a mouse, a trackpad, a QWERTY keyboard, etc. can be used as the input unit 24.

[0024] <Output section 25> The output unit 25 displays a screen of a graphical user interface (GUI) that can be operated by the user. The output unit 25 may be included in the housing of the recruiter terminal 20, or may be attached externally. Specifically, the output unit 25 may be implemented as a display device such as a CRT display, a liquid crystal display, an organic EL display, or a plasma display. It is preferable that these display devices are implemented by selectively using them depending on the type of recruiter terminal 20.

[0025] <Job Seeker Terminal 30> 3B, the job seeker terminal 30 includes a control unit 31, a memory unit 32, a communication unit 33, an input unit 34, an output unit 35, and a communication bus 36. The control unit 31, the memory unit 32, the communication unit 33, the input unit 34, and the output unit 35 are electrically connected via the communication bus 36 inside the job seeker terminal 30. The explanation of the control unit 31, the memory unit 32, the communication unit 33, the input unit 34, and the output unit 35 will be omitted as they are the same as the explanation of each unit in the recruiter terminal 20.

[0026] 2. Functional configuration This section describes the functional configuration of this embodiment. Information processing by the software stored in the storage unit 12 is specifically realized by the control unit 11, which is an example of hardware, and can be executed as each functional unit included in the control unit 11 (a processor provided in the job search support system 1).

[0027] FIG. 4 is a block diagram showing functions realized by the server device 10 (controller 11), the recruiter terminal 20 (controller 21), and the job seeker terminal 30 (controller 31).

[0028] As shown in Fig. 4A, the server device 10 (control unit 11) includes a basic display control unit 111, an acquisition unit 112, an estimation unit 113, an extraction unit 114, a presentation unit 115, a job search unit 116, a scout display control unit 117, and an artificial intelligence unit 120. As shown in Fig. 4B, the recruiter terminal 20 (control unit 21) includes a display unit 211 and an operation acquisition unit 212. As shown in Fig. 4C, the job seeker terminal 30 (control unit 31) includes a display unit 311 and an operation acceptance unit 312.

[0029] <Basic display control unit 111> The basic display control unit 111 is configured to display various information on the recruiter terminal 20 and the job seeker terminal 30. For example, the basic display control unit 111 displays a resume and curriculum vitae prepared by the job seeker, a job posting prepared by the recruiter, etc. on the display unit 211 of the recruiter terminal 20 or the display unit 311 of the job seeker terminal 30.

[0030] Recruiters include organizations such as for-profit corporations (such as companies), non-profit corporations (such as cooperatives and foundations), and public corporations (such as local governments). Recruiters also include recruitment agencies that act as agents of organizations to mediate between job seekers and organizations. Recruitment agencies are also called headhunters or agents.

[0031] <Acquisition part 112> The acquisition unit 112 is configured to acquire information on candidate job offers selected by a job seeker from among the job offers registered in the job offer database. Specifically, the acquisition unit 112 accepts the selection of candidate job offers from the job seeker terminal 30, and acquires information including the details of the selected candidate job offer (job advertisement) from the job offer database stored in the storage unit 12, for example.

[0032] Examples of "candidate job offers" include job offers that a job seeker selects as a job to be viewed (i.e., a job offer for which the job seeker has instructed the job seeker terminal 30 to display the contents of the job posting), job offers that a job seeker selects as a job to be assigned a candidate label (bookmark) (i.e., a job offer for which the job seeker has instructed the job seeker to assign a candidate label), job offers to which a job seeker has applied, job offers for which a job seeker has received a scouting document, and job offers to which a job seeker has replied to a scouting document. Job offers that have been assigned a candidate label are registered, for example, in a bookmark list prepared for each job seeker. By referring to the bookmark list, a job seeker can select a job offer to take an action such as viewing or applying for from among the job offers that have been assigned a candidate label. In addition, job offers that are eligible for candidate job offers are not limited to job offers that satisfy the recruitment requirements of a job seeker, as long as they are registered in the job search support system 1, and include job offers for which a job seeker can perform an action such as viewing.

[0033] A job posting, which is the content of a job offer, lists job information for multiple items such as the name of the position being recruited, job content and working conditions (annual salary, job type, industry, work location, work style, work environment, etc.), application qualifications (skills), desired personality, selling points, etc. Job postings may also include items such as the job title, headline, and information about the employer (company size (sales, number of employees, etc.), industry, etc.).

[0034] Furthermore, the acquisition unit 112 acquires registration information of job seekers to be used by the estimation unit 113, which will be described later, from, for example, a job seeker database stored in the storage unit 12. The registration information of job seekers includes resumes, curriculum vitae, and other profile information of job seekers.

[0035] A "resume" is a document that mainly contains a job seeker's profile, current situation, educational background, work history, desired working conditions, etc., while a "curriculum vitae," also known as a resume, is a document in which a job seeker conveys to an employer his or her work history, experience, skills, qualifications, etc. In addition, the job seeker registration information may also include the job seeker's desired conditions (desired industry, desired job type, etc.).

[0036] <Estimation part 113> The estimation unit 113 is configured to estimate at least one recommended category based on the content of the candidate job offer and the first reference information. A "recommended category" is a category in which the job seeker is interested, among categories for classifying information included in a job offer based on attributes. Examples of attributes for classifying information included in a job offer into categories include business content, services or products offered, company size, working conditions, job title, industry, occupation, required qualifications and skills, etc. Examples of categories related to "business content" include "direction," "planning," "technology development," and "market analysis." Examples of categories related to "services or products offered" include "owned media," "SaaS," "precision machinery," and "food." Examples of categories related to "company size" include "startup," "small and medium-sized enterprises," "mid-sized enterprises," "large enterprises," and "listed companies." For example, if the estimation unit 113 determines that the job seeker is interested in information (keywords) belonging to "direction," the estimation unit 113 estimates "direction" as the recommended category. A specific estimation procedure will be described later.

[0037] The estimation unit 113 may estimate the recommended category based on the content of one candidate job offer, but it is preferable to estimate the recommended category based on the content of multiple candidate job offers (for example, multiple job postings viewed by the job seeker). This increases the accuracy of estimating the recommended category. Furthermore, the estimation unit 113 may estimate multiple recommended categories for one job seeker. This makes it possible to present multiple recommended categories and recommended job offers associated with these recommended categories to the user job seeker.

[0038] The first reference information includes a first correlation between the contents of candidate job offers and recommended categories. The first reference information is stored, for example, in the memory unit 12. The first reference information is an estimator constructed so that it can input the contents of candidate job offers (job postings) and output recommended categories. The first reference information includes, for example, a table, a function, a simple algorithm, etc., that indicates the correlation between feature amounts (vector data) extracted from multiple job postings and recommended categories. The first correlation included in the first reference information can be constructed, for example, by statistically analyzing data that records job postings viewed by a job seeker and the job seeker's behavioral history (keywords used to search for job postings, job postings applied for, job postings that have been concluded, etc.).

[0039] The first reference information may include a first category estimation model of the artificial intelligence unit 120 that has been trained to input the contents of candidate job offers and output recommended categories. In this case, the estimation unit 113 inputs the contents of candidate job offers to the first category estimation model and causes the first category estimation model to output recommended categories. This makes it possible to predict recommended categories based on an analysis of the contents of a large number of job offers.

[0040] The first category estimation model is a learning model trained using training job posting data and correct answer data for the recommended category to be extracted from the job posting as training data. The training job posting data includes master information (numeric values ​​or items selected from preset candidates) such as annual salary, job type, industry, work location, and skills, as well as text information consisting of sentences such as title and detailed information. In the first category estimation model, the parameters calculated or tuned through training correspond to the first correlation. The first category estimation model is subjected to additional training (updating of the learning model) based on newly registered job posting data at regular intervals (for example, once a day) by, for example, the artificial intelligence unit 120.

[0041] The first category estimation model may be trained to output a recommended category corresponding to a candidate job from among the multiple candidate categories by constructing a correspondence relationship between multiple candidate categories and keywords contained in the job that correspond to the candidate categories based on the contents of multiple job openings registered in the job opening database. A "candidate category" is a candidate category to be output as a recommended category, and each candidate category is associated with multiple keywords that fall into that candidate category. This improves the prediction accuracy of the candidate category.

[0042] Candidate categories may be generated by the first category estimation model itself during its training process, or may be defined before the first category estimation model is trained. When the first category estimation model itself generates candidate categories, the first category estimation model, for example, extracts keywords used in a large number of job postings registered in the job posting database and generates categories that group these keywords as candidate categories. On the other hand, when the first category estimation model is trained after candidate categories are defined, the first category estimation model is trained to associate keywords included in a large number of job postings with given candidate categories. The first category estimation model may also generate candidate categories by extracting keywords used in a plurality of job postings that have specific attributes, such as a specific occupation or industry, from among the job postings registered in the job posting database. The first category estimation model may also extract keywords for generating candidate categories based on the frequency of keyword usage in the plurality of job postings. In this case, the first category estimation model may, for example, extract frequently used keywords (those whose frequency of usage is equal to or greater than a predetermined value) from the plurality of job postings.

[0043] For example, a topic model is used as the first category estimation model. In the topic model construction (learning) stage, for example, first, for sentences included in all or some of the available job postings registered in the job posting database (for example, job postings with specific attributes, such as a specific job type or industry), the frequency of keyword appearance is calculated using morphological analysis or the like, and the co-occurrence of multiple keywords is analyzed to extract multiple topics. Next, the probability distribution (probability of appearance) of keywords in each extracted topic and the probability distribution (probability of appearance) of topics in each job posting are estimated. The topics in the topic model correspond to the "candidate categories" of the first category estimation model, and the probability distribution of keywords in each topic corresponds to the "correspondence between candidate categories and keywords included in job postings" of the first category estimation model. Here, for example, only job postings with specific attributes, such as a specific job type or industry, among the job postings registered in the job posting database may be used to construct the topic model.

[0044] The topic model constructed in this manner receives input of candidate job offers and outputs a probability distribution of topics in the candidate job offers based on keywords included in the candidate job offers. The topic model then outputs, for example, at least one topic from among the multiple topics included in the candidate job offers whose frequency of appearance (probability) is equal to or greater than a predetermined threshold as a recommended category. The topic model may receive input of multiple candidate job offers (e.g., multiple job offers viewed by a job seeker) and estimate multiple topics corresponding to each candidate job offer, and output recommended categories based on the frequency of appearance of the estimated multiple topics. For example, the topic model may output, as recommended categories, topics from among the estimated multiple topics whose frequency of appearance is equal to or greater than a predetermined value, or a predetermined number of topics (e.g., the top three) in descending order of frequency of appearance. The topic model also outputs, in the topic estimated as a recommended category, at least one keyword whose frequency of appearance is equal to or greater than a predetermined threshold, as a recommended keyword related to the topic (recommended category).

[0045] The first category estimation model may be a generative AI including a large-scale language model. In this case, the estimation unit 113 inputs the job posting of at least one candidate job, inputs a prompt including an instruction to output recommended categories to the first category estimation model, and causes the first category estimation model to output the recommended categories. Furthermore, the estimation unit 113 may input, to the first category estimation model, in addition to the instruction to output recommended categories and the job posting of the candidate job, a prompt in which, for example, samples of job postings of one or more candidate jobs and samples of one or more corresponding recommended categories are inserted as input and output samples.

[0046] The estimation unit 113 may estimate, as recommended keywords, keywords that correspond to the recommended categories output by the first category estimation model. This enables the extraction unit 114, which will be described later, to extract recommended job offers using the recommended keywords.

[0047] The recommended keywords estimated by the estimation unit 113 may be keywords associated with recommended categories (candidate categories) in the correspondence relationships learned by the first category estimation model. That is, the estimation unit 113 may cause the first category estimation model to output recommended categories and recommended keywords.

[0048] Furthermore, the estimation unit 113 may estimate recommended keywords from recommended categories output by the first category estimation model using the correspondence between recommended categories and recommended keywords included in the first reference information. In this case, the recommended keywords do not necessarily match the keywords associated with the recommended categories in the correspondence learned by the first category estimation model. The correspondence between recommended categories and recommended keywords is prepared, for example, as a table defining at least one keyword corresponding to each category. The estimation unit 113 estimates, as the recommended keyword, at least one keyword associated with a category identical to or similar to the recommended category in the first reference information. Note that the similarity of categories is determined, for example, by comparing features such as vector data, referring to a database that defines similar categories, or by determining using a learning model (large-scale language model).

[0049] The estimation unit 113 may estimate recommended keywords that interest the job seeker based on the content of the candidate job offer and the first reference information, and may also estimate recommended categories based on the recommended keywords. In this case, the first reference information further includes a second correlation between the content of the candidate job offer and the recommended keywords, and a correspondence between the recommended keywords and the recommended categories. This enables the extraction unit 114, which will be described later, to extract recommended job offers based on the recommended keywords.

[0050] When estimating recommended keywords from candidate job offers, the estimation unit 113 may estimate recommended keywords based on the content of one candidate job offer, but it is preferable to estimate recommended keywords based on the content of multiple candidate job offers, which increases the accuracy of estimating recommended keywords.

[0051] The first reference information for estimating recommended keywords includes, for example, a table, a function, a simple algorithm, or the like, which indicates the correlation between the feature amounts (vector data) extracted from multiple job postings and the recommended keywords. The second correlation included in the first reference information can be constructed, for example, similar to the first correlation for estimating recommended categories, by statistically analyzing data recording the job postings viewed by a job seeker and the behavioral history of the job seeker.

[0052] The first reference information may include a keyword estimation model of the artificial intelligence unit 120 that has been trained to input the contents of candidate job offers and output recommended keywords. In this case, the estimation unit 113 inputs the contents of candidate job offers into the keyword estimation model and causes the keyword estimation model to output recommended keywords. This makes it possible to predict recommended keywords based on an analysis of the contents of a large number of job offers.

[0053] The keyword estimation model is a learning model trained using training job posting data and correct answer data of recommended keywords to be extracted from the job posting as training data. In the keyword estimation model, the parameters calculated or tuned by training correspond to the second correlation.

[0054] The keyword estimation model may also be a generation AI that includes a large-scale language model. In this case, the estimation unit 113 inputs at least one candidate job posting, inputs a prompt including an instruction to output recommended keywords to the keyword estimation model, and causes the keyword estimation model to output the recommended keywords. In addition to the instruction to output recommended keywords and the candidate job posting, the estimation unit 113 may also input, as input and output samples, a prompt that includes, for example, one or more sample candidate job postings and one or more corresponding sample recommended keywords to the keyword estimation model.

[0055] The correspondence between the recommended keywords and the recommended categories included in the first reference information is prepared, for example, as a table that defines at least one keyword that corresponds to each category, as in the case of estimating recommended keywords from recommended categories. The estimation unit 113 estimates, as the recommended category, a category that is associated with a keyword that is the same as or similar to the recommended keyword in the first reference information. Note that one recommended category may be estimated based on multiple recommended keywords, or multiple recommended categories may be estimated based on one recommended keyword.

[0056] The first reference information may also include a second category estimation model of the artificial intelligence unit 120 that has been trained to be able to input recommended keywords and output recommended categories. In this case, the estimation unit 113 inputs the recommended keywords to the second category estimation model and causes the second category estimation model to output recommended categories. The second category estimation model is a learning model that has trained using data on recommended keywords and correct answer data for recommended categories to which the recommended keywords belong as training data.

[0057] The second category estimation model may also be a generative AI including a large-scale language model. In this case, the estimation unit 113 receives recommended keywords as input, inputs a prompt including an instruction to output recommended categories to the second category estimation model, and causes the second category estimation model to output the recommended categories. The estimation unit 113 may also input predefined category candidates and a prompt including an instruction to infer a category corresponding to the recommended keywords from the category candidates and output the category as a recommended category to the second category estimation model. Furthermore, the estimation unit 113 may input a prompt into which, as input and output samples, for example, one or more samples of recommended keywords and one or more samples of recommended categories corresponding to the recommended keywords are inserted, in addition to the instruction to output the recommended categories and the recommended keywords, to the second category estimation model.

[0058] The estimation unit 113 may estimate recommended keywords based on the content of the candidate job offers, the registration information of the user job seeker, and the first reference information. In this case, the first reference information includes a third correlation between the content of the candidate job offers, the registration information of the job seeker, and the recommended keywords. This allows the recommended keywords to be estimated by further referring to the registration information of the user job seeker, in addition to the content of the candidate job offers. This makes it possible to present recommended categories that verbalize the potential desires of the user job seeker, and corresponding recommended job offers.

[0059] The third reference information when estimating recommended keywords using the job seeker's registration information includes, for example, a table, a function, a simple algorithm, etc., which indicates the correlation between the feature values ​​(vector data) extracted from multiple job postings, the feature values ​​extracted from the registration information, and the recommended keywords. The third correlation included in the first reference information can be constructed, for example, by statistically analyzing data recording the job postings viewed by the job seeker, the job seeker's registration information, and the job seeker's behavioral history.

[0060] The estimation unit 113 may determine recommended keywords based on the magnitude of the keyword score, which increases as the frequency of use in candidate job offers increases and decreases as the frequency of use in the job seeker's registered information increases. This allows keywords that are more likely to be listed in job postings than in the job seeker's registered information to be estimated as recommended keywords. Furthermore, such recommended keywords are more likely to be used by job seekers in job searches. Therefore, by extracting recommended job offers using the recommended keywords, it becomes possible to present recommended job offers that are more suitable for the user, the job seeker.

[0061] The keyword score is calculated, for example, by the following procedure. First, the estimation unit 113 calculates the frequency of use of each of the multiple job keywords included in the multiple candidate job openings across all the multiple candidate job openings (the number of times a keyword that is the same as or similar to the job keyword appears). The estimation unit 113 also calculates the frequency of use of each of the multiple job keywords in the registered information of the job seeker, who is the user. For each job keyword, the estimation unit 113 calculates the keyword score by subtracting a certain weighted value for the frequency of use in the registered information of the job seeker from the frequency of use across all the candidate job openings.

[0062] For example, the estimation unit 113 determines, as recommended keywords, job recruitment keywords whose keyword scores are equal to or greater than a predetermined threshold. The estimation unit 113 may also determine recommended keywords based on the ranking of keyword scores among the extracted job recruitment keywords. For example, the estimation unit 113 may determine, as recommended keywords, job recruitment keywords whose percentage (top percentage) of keyword scores from the top is equal to or greater than a predetermined threshold.

[0063] The estimation unit 113 may determine, from among the recommended keywords estimated from the recommended categories output by the first category estimation model or the recommended keywords output by the keyword estimation model, keywords having a keyword score equal to or greater than a predetermined threshold, the value of which decreases as the frequency of use in the job seeker's registration information increases, as recommended keywords to be used for extraction of recommended job offers by the extraction unit 114. This makes it possible to extract, as recommended keywords, keywords that are used in job offers but are not often used in the registration information such as the job seeker's resume.

[0064] The first reference information may include a keyword estimation model of the artificial intelligence unit 120 that has been trained to be able to output recommended keywords using the content of the candidate job offer and the registration information of the user, who is a job seeker, as input. In this case, the estimation unit 113 inputs the content of the candidate job offer and the registration information of the job seeker into the keyword estimation model and causes the keyword estimation model to output recommended keywords. The keyword estimation model is a learning model that has trained using training data of job postings, training data of the registration information of the job seeker, and correct answer data of recommended keywords that should be extracted from the job postings by referring to the registration information as training data. In the keyword estimation model, parameters calculated or tuned by training correspond to the third correlation.

[0065] The keyword estimation model may also be a generation AI that includes a large-scale language model. In this case, the estimation unit 113 receives as input at least one candidate job posting and the job seeker's registration information, inputs a prompt including an instruction to output recommended keywords to the keyword estimation model, and causes the keyword estimation model to output the recommended keywords. The estimation unit 113 may also input to the keyword estimation model a prompt that includes, as input and output samples, one or more samples of candidate job postings, one or more samples of the job seeker's registration information, and one or more samples of recommended keywords corresponding to the candidate job postings, in addition to the instruction to output recommended keywords, the candidate job postings, and the job seeker's registration information.

[0066] The estimation unit 113 may exclude keywords related to conditions designed to be selected from pre-prepared options in a job search from the recommended keyword candidates. This allows keywords that are unlikely to be entered as search keywords by job seekers in a job search to be excluded from the recommended keywords. This improves the accuracy of extracting recommended jobs. Examples of "conditions designed to be selected from options" include work location, availability of remote work, employment status (full-time employee, contract employee, etc.). These conditions are rarely entered as search keywords by job seekers when searching job postings.

[0067] The above-mentioned keyword estimation model that receives as input the content of the candidate job offer and the keyword estimation model that receives as input the content of the candidate job offer and the registration information of the user (job seeker) may be trained so as to be able to output both recommended categories and recommended keywords. Furthermore, if the keyword estimation model is a generation AI that includes a large-scale language model, the estimation unit 113 may receive as input the candidate job offer advertisement or the candidate job offer advertisement and the registration information of the job seeker, and input a prompt that includes an instruction to output recommended categories and recommended keywords to the keyword estimation model, thereby causing the keyword estimation model to output the recommended categories and recommended keywords.

[0068] The estimation unit 113 may estimate the recommended category and / or recommended keyword by referring to only the candidate job offers selected (viewed, etc.) by the job seeker user within a predetermined period. That is, the acquisition unit 112 may acquire the content of the candidate job offers selected by the job seeker within a predetermined period. For example, the estimation unit 113 may estimate the recommended category and / or recommended keyword based on the content of the candidate job offers viewed by the job seeker user within a certain period (e.g., seven days) from the current time. This makes it possible to present recommended categories and recommended job offers that reflect the most recent preferences of the job seeker user.

[0069] The estimation unit 113 may estimate the recommended categories and / or recommended keywords at predetermined intervals. The interval at which the estimation unit 113 estimates the recommended categories and / or recommended keywords (the update interval for the recommended categories and / or recommended keywords) is, for example, one day. In this way, by the estimation unit 113 performing estimation once a day, the behavioral history of the job seeker, who is the user, regarding job offers, can be immediately reflected in the estimation result.

[0070] <Extraction part 114> The extraction unit 114 is configured to extract at least one recommended job offer to be recommended to the job seeker from among the job offers registered in the job offer database, based on the recommended category or the recommended keyword related to the recommended category and the second reference information.

[0071] When the estimation unit 113 extracts multiple recommended categories, the extraction unit 114 may extract at least one recommended job offer for each of the multiple recommended categories. This allows specific recommended job offers for each of the multiple recommended categories to be presented to the user (job seeker), making it easier for the job seeker to get an idea of ​​the recommended categories. Furthermore, the extraction unit 114 may extract multiple recommended job offers for each recommended category.

[0072] The second reference information includes a fourth correlation between the recommended categories or recommended keywords and the recommended job postings. The second reference information is an estimator constructed to be able to output recommended job postings using at least one of the recommended categories and recommended keywords as input. The second reference information includes, for example, a table, a function, a simple algorithm, etc., that indicates the correlation between multiple recommended categories or multiple recommended keywords and feature quantities (vector data) extracted from the job postings.

[0073] The extraction unit 114 may extract recommended job offers based on the similarity between the recommended category and the content of the job offer. This makes it possible to extract recommended job offers using an approach different from a job search by a job seeker who is a user, without estimating recommended keywords.

[0074] The similarity between the recommended category and the content of a job offer is defined, for example, by the difference (vector distance) between the feature obtained by vectorizing the recommended category and the feature obtained by vectorizing the words or sentences included in the job advertisement of the job offer to be compared. For example, cosine similarity is used as this similarity, and the closer the cosine similarity is to 1, the greater the similarity. In this case, the second reference information includes a similarity range for determining that the job offer compared with the recommended category is a recommended job offer. The extraction unit 114 extracts, as recommended jobs, job offers whose similarity with the recommended category is within a predetermined range (for example, the value obtained by subtracting the cosine similarity from 1 is equal to or less than a threshold).

[0075] Furthermore, the extraction unit 114 may calculate the similarity between the recommended category and the content of the job offer using a similarity calculation model of the artificial intelligence unit 120 that has been trained to be able to output the similarity between two pieces of input information. That is, the second reference information may include the similarity calculation model. The similarity calculation model may be a generative AI that includes a large-scale language model. In this case, the extraction unit 114 inputs two pieces of input information (the recommended category and the job advertisement) as input, inputs a prompt including an instruction to output the similarity to the similarity calculation model, and causes the similarity calculation model to output the similarity. Furthermore, the extraction unit 114 may input a prompt to the similarity calculation model that includes, as input and output samples, one or more samples of input information and one or more corresponding similarity samples, in addition to the instruction to output the similarity and the two pieces of input information.

[0076] Furthermore, the extraction unit 114 may extract, as recommended job offers, job offers that include information classified into recommended categories (recommended categories extracted from job postings). In this case, the second reference information includes, for example, a list of keywords corresponding to the recommended categories (including the names of the recommended categories themselves). The extraction unit 114 extracts, from job offers registered in the job offer database, job offers that include keywords that are the same as or similar to keywords corresponding to recommended categories, as recommended job offers corresponding to the recommended categories.

[0077] Furthermore, the extraction unit 114 may extract recommended job offers by referring to category labels previously attached to job offers. For example, the extraction unit 114 may extract job offers attached with category labels identical or similar to recommended categories as recommended job offers corresponding to the recommended categories. Category labels are attached to job offers based on keywords included in the job offer advertisement, for example. Furthermore, multiple category labels may be attached to one job offer advertisement. In this case, the second reference information includes, for example, information indicating category labels identical or similar to the recommended categories.

[0078] The extraction unit 114 may extract recommended job offers using recommended keywords instead of recommended categories. That is, the extraction unit 114 may extract recommended job offers based on recommended keywords associated with recommended categories and the second reference information. This makes it possible to extract recommended job offers based on keywords used in the recommended job offers.

[0079] Here, the recommended keywords used by the extraction unit 114 to extract recommended job offers do not necessarily have to be those estimated by the estimation unit 113. For example, the extraction unit 114 may extract keywords related to the recommended category (for example, associated in the second reference information) from the registration information of the user, who is a job seeker, as recommended keywords, and further extract recommended job offers based on the recommended keywords. Alternatively, the extraction unit 114 may extract keywords included in the name of the recommended category as recommended keywords, and further extract recommended job offers based on the recommended keywords.

[0080] The extraction unit 114 may extract recommended job offers by keyword search using the recommended keywords as search conditions. This makes it possible to extract recommended job offers using an existing job search search flow. In this case, the extraction unit 114 extracts, as recommended job offers, job offers that include keywords that are the same as or similar to the recommended keywords estimated by the estimation unit 113 from job offers registered in the job offer database. Furthermore, the extraction unit 114 associates the extracted recommended job offers with recommended categories that are associated with the recommended keywords used to extract the recommended job offers.

[0081] If multiple recommended keywords are estimated by the estimation unit 113, the extraction unit 114 may perform a search for each recommended keyword, or may perform a search by combining multiple recommended keywords associated with the same recommended category (for example, as an AND condition or an OR condition).

[0082] The extraction unit 114 may extract recommended job offers based on the similarity between the recommended keywords and the content of the job offers. This makes it possible to present recommended job offers extracted using an approach different from keyword search to the job seeker, who is the user.

[0083] The similarity between the recommended keywords and the content of the job offer is defined, for example, by the difference (vector distance) between the feature value obtained by vectorizing the recommended keywords and the feature value obtained by vectorizing the words or sentences included in the job advertisement of the job offer to be compared. In this case, the second reference information includes a similarity range for determining that the job offer compared with the recommended keywords is a recommended job offer. The extraction unit 114 extracts, as a recommended job offer, a job offer whose similarity with the recommended keywords is within a predetermined range. The extraction unit 114 may also calculate the similarity between the recommended keywords and the content of the job offer using a similarity calculation model (including a generation AI) that has been trained to be able to output the similarity between two pieces of input information.

[0084] The extraction unit 114 may extract recommended job offers using both the recommended category and the recommended keyword. That is, the second reference information may include, as the fourth correlation, both the correlation between the recommended category and the recommended job offer and the correlation between the recommended keyword and the recommended job offer. For example, the extraction unit 114 may extract, as recommended job offers, job offers whose similarity to both the recommended category and the recommended keyword are within their respective predetermined ranges (that is, satisfying the condition of high similarity).

[0085] The second reference information may include a job vacancy extraction model of the artificial intelligence unit 120 that has been trained to be able to input recommended categories and / or recommended keywords and output recommended job vacancies. In this case, the extraction unit 114 inputs the recommended categories and / or recommended keywords into the job vacancy extraction model and causes the job vacancy extraction model to output recommended job vacancies. The job vacancy extraction model is a learning model that has trained using data on recommended categories and / or recommended keywords and data on job postings corresponding to the recommended categories and / or recommended keywords as training data. In the job vacancy extraction model, parameters calculated or tuned by learning correspond to the fourth correlation.

[0086] The job offer extraction model may also be a generation AI including a large-scale language model. In this case, the extraction unit 114 inputs recommended categories and / or recommended keywords, inputs a prompt including an instruction to output recommended job offers to the job offer extraction model, and causes the job offer extraction model to output the recommended job offers. Furthermore, the extraction unit 114 may input, to the job offer extraction model, a prompt in which, as input and output samples, for example, one or more samples of recommended categories and / or recommended keywords and one or more samples of recommended job offers corresponding thereto, in addition to the instruction to output recommended job offers and the recommended categories and / or recommended keywords.

[0087] The extraction unit 114 may extract a plurality of recommended job offers for each recommended category and calculate a recommendation score for each recommended job offer. The recommendation score is calculated based on at least one of the similarity between the content of the recommended job offer and the registered information of the user (job seeker) and the probability of an action related to the recommended job offer occurring. This makes it possible to control the display of recommended job offers based on a recommendation score that takes into account the aptitude of the user (job seeker), the ease of taking action, etc.

[0088] The similarity between the content of the recommended job offer and the registered information of the job seeker is defined, for example, by the difference (vector distance) between the feature amount obtained by vectorizing the words or sentences included in the job posting of the recommended job offer and the feature amount obtained by vectorizing the words or sentences included in the registered information of the user job seeker. Furthermore, the extraction unit 114 may calculate the similarity between the content of the recommended job offer and the registered information of the job seeker using a similarity calculation model trained to be able to output the similarity between two pieces of input information. In other words, the second reference information may include the similarity calculation model.

[0089] "Actions related to recommended job offers" are actions that can be performed by multiple job seekers registered for a recommended job offer, and include viewing the job offer, applying for the job offer by the job seeker, replying to a scouting document based on the job offer, and closing a deal for the job offer. "Probability of an action occurring" includes, for example, the probability that a job seeker will go from viewing a job offer to applying, the probability that a job seeker will reply to a scouting document, the probability that a scouting document will be sent and closed, and the probability of passing the screening (document screening, interview screening, etc.). These probabilities are calculated, for example, based on the history (actual values) of multiple actions related to the job offer stored in the memory unit 12. For example, the "probability that a job seeker will go from viewing a job offer to applying" can be obtained by dividing the number of applications by job seekers by the number of views by job seekers.

[0090] The extraction unit 114 may estimate the occurrence probability of an action related to a recommended job offer using an occurrence probability calculation model of the artificial intelligence unit 120 that has been trained to take the content of the recommended job offer as input and output the occurrence probability of the action. In this case, the extraction unit 114 inputs the job posting of the recommended job offer into the occurrence probability calculation model and causes the occurrence probability calculation model to output the occurrence probability of the action. The occurrence probability calculation model is a learning model that has trained using training job posting data and data regarding the occurrence of actions in the job posting (e.g., data on whether or not a job seeker has viewed, applied, replied to a scouting document, passed a screening, or concluded a job offer) as training data. The occurrence probability calculation model may be a generative AI that includes a large-scale language model. In this case, the extraction unit 114 takes the job posting of the recommended job offer as input, inputs a prompt including an instruction to output the occurrence probability of the action into the occurrence probability calculation model, and causes the occurrence probability calculation model to output the occurrence probability of the action. In addition, the extraction unit 114 may input a prompt into the occurrence probability calculation model that includes, as input and output samples, for example, samples of job postings for one or more recommended jobs and samples of the occurrence probabilities of one or more corresponding actions, in addition to an output instruction for the probability of the action occurring and the job posting for the recommended job.

[0091] The extraction unit 114 may calculate the recommendation score based only on the similarity between the content of the recommended job offer and the registered information of the job seeker, may calculate the recommendation score based only on the probability of an action related to the recommended job offer occurring, or may calculate the recommendation score based on both the similarity and the probability of an action occurring. When using the similarity and the probability of an action occurring, the extraction unit 114 may calculate the recommendation score by, for example, adding the value obtained by multiplying the similarity by the first weighting coefficient and the value obtained by multiplying the probability of an action occurring by the second weighting coefficient.

[0092] The extraction unit 114 may extract at least one auxiliary recommended job offer to be auxiliary recommended to the job seeker from among the job offers registered in the job offer database, based on at least one of the content of the candidate job offer and the registration information of the user job seeker, and the third reference information. The auxiliary recommended job offer is a job offer extracted without using the recommended category and recommended keyword estimated by the estimation unit 113. The third reference information includes a fifth correlation between the auxiliary recommended job offer and at least one of the content of the candidate job offer and the registration information of the job seeker, and the third reference information.

[0093] The extraction unit 114 extracts, for example, job offers whose content is highly similar to that of the candidate job offers as job offers recommended for assistance. The similarity between the content of job offers is defined, for example, by the difference in feature values ​​(vector distance) obtained by vectorizing words or sentences included in the job advertisement of the job offer being compared. In this case, the third reference information includes a range of similarity for determining that the job offer compared to the candidate job offer is a job offer recommended for assistance. The extraction unit 114 extracts, as job offers recommended for assistance, job offers whose similarity with the candidate job offer is within a predetermined range.

[0094] The extraction unit 114 may extract, as a supplementary recommended job offer, a job offer whose content is highly similar to the job seeker's registered information. The similarity between the job seeker's registered information and the content of the job offer is defined, for example, by the difference (vector distance) between the feature amount obtained by vectorizing the words or sentences included in the job seeker's registered information and the feature amount obtained by vectorizing the words or sentences included in the job posting of the job offer to be compared. In this case, the third reference information includes a similarity range for determining that the job offer compared with the job seeker's registered information is a supplementary recommended job offer. The extraction unit 114 extracts, as a supplementary recommended job offer, a job offer whose similarity with the job seeker's registered information falls within a predetermined range.

[0095] The extraction unit 114 may extract the auxiliary recommended job offers based on both the first similarity to the content of the candidate job offers and the second similarity to the registered information of the job seeker. For example, the extraction unit 114 may extract, as the auxiliary recommended job offers, job offers for which both the first similarity and the second similarity are within their respective predetermined ranges (i.e., satisfying the condition of high similarity).

[0096] The third reference information may include an auxiliary job offer extraction model of the artificial intelligence unit 120 that has been trained to be able to output auxiliary recommended job offers using the content of candidate job offers and / or the job seeker's registration information as input. In this case, the extraction unit 114 inputs the content of candidate job offers and / or the job seeker's registration information into the auxiliary job offer extraction model and causes the auxiliary job offer extraction model to output auxiliary recommended job offers. The auxiliary job offer extraction model is a learning model trained using training job postings and / or job seeker registration information data and data on the job postings and / or job postings similar to the training information as training data. In the auxiliary job offer extraction model, parameters calculated or tuned through learning correspond to the fifth correlation. The auxiliary job offer extraction model may be a generative AI including a large-scale language model. In this case, the extraction unit 114 inputs the job postings of candidate job offers and / or the job seeker's registration information into the auxiliary job offer extraction model and inputs a prompt including an instruction to output auxiliary recommended job offers into the auxiliary job offer extraction model, causing the auxiliary recommended job offers to be output. In addition to the output instruction for the auxiliary recommended job offer and the job postings and / or job seeker registration information of the candidate job offers, the extraction unit 114 may also input a prompt into the auxiliary job offer extraction model as input and output samples, for example, samples of the job postings and / or job seeker registration information of one or more candidate job offers and samples of one or more corresponding auxiliary recommended job offers.

[0097] <Presentation part 115> The presentation unit 115 is configured to present the recommended categories estimated by the estimation unit 113 and the recommended job offers extracted by the extraction unit 114 in association with each other to a job seeker.

[0098] The presentation unit 115 presents the recommended job by, for example, displaying on the job seeker terminal 30 part of the information included in the job posting of the recommended job (job title (position name), job number (ID), conditions, etc.).

[0099] "Presenting in association" includes, for example, setting an object (e.g., a frame) or display area for each recommended category and displaying the corresponding recommended job offers within this object or display area, or assigning the corresponding recommended category to each recommended job offer and displaying it.

[0100] The presentation unit 115 may present all of the multiple recommended job offers extracted by the extraction unit 114, or may present some of the multiple recommended job offers. Furthermore, if the estimation unit 113 has estimated multiple recommended categories, the presentation unit 115 may present the recommended job offers to the job seeker in association with each of the multiple recommended categories. This allows the job seeker (user) to focus on a recommended category that interests him or her and closely examine the recommended job offers within that recommended category. Furthermore, by presenting recommended job offers for each recommended category, it becomes easier for the job seeker (user) to get a sense of the recommended categories, making it easier for the job seeker (user) to verbalize his or her desired conditions. This makes it easier for the job seeker (user) to recall search keywords when searching for job postings himself or herself.

[0101] The presentation unit 115 may present all of the multiple recommended categories estimated by the estimation unit 113, or may present only a predetermined number of recommended categories (for example, three). The recommended categories to be presented are determined based on, for example, the number of associated recommended job offers, the maximum or average recommendation scores of the associated recommended job offers, a predetermined priority for each category, etc.

[0102] The presentation unit 115 may present the extracted recommended job offers to the job seeker in an order based on their respective recommendation scores. The recommendation score is a score calculated for each recommended job offer by the extraction unit 114 as described above. This allows recommended job offers that are highly suitable for the user (job seeker) to be preferentially displayed, making it easier for the job seeker to take action such as applying for the presented recommended job offers.

[0103] The presentation unit 115 presents the recommended job offers in the form of a recommended job offer list in which the recommended job offers are sorted in descending order of recommendation score for each recommended category, for example. The presentation unit 115 may select a predetermined number of recommended job offers (for example, three) in descending order of recommendation score for each recommended category and display them on the job seeker terminal 30. Furthermore, the presentation unit 115 may not present recommended job offers whose recommendation scores are less than a predetermined threshold.

[0104] The presentation unit 115 may present recommended job offers selected at random from the extracted plurality of recommended job offers to the job seeker. This can prevent job seekers from concentrating their applications or other actions on a specific job offer or organization. As a result, the probability of closing a job offer based on the presented recommended job offers can be increased. For example, the presentation unit 115 may randomly select a predetermined number of recommended job offers for each recommended category and display them on the job seeker terminal 30. The presentation unit 115 may also set the probability of a job offer being selected as a recommended job offer to be presented (selection probability) depending on the recommendation score (the degree of similarity with the job seeker's registered information or the probability of an action related to the candidate job offer occurring). For example, the presentation unit 115 may set the selection probability to 40% for a recommended job offer with the highest recommendation score and 30% for a recommended job offer with the second highest recommendation score.

[0105] If the recommended job offers extracted by the extraction unit 114 include candidate job offers (for example, job offers that have been viewed by the user job seeker), the presentation unit 115 may exclude the recommended job offers from the recommended job offers to be presented (may not present them).

[0106] The presentation unit 115 may accept input of an instruction to present recommended job offers from a job seeker who is a user, and present the recommended categories and recommended job offers to the job seeker terminal 30, or may present the recommended categories and recommended job offers to the job seeker terminal 30 in response to the occurrence of a predetermined event or at a predetermined timing. Examples of predetermined events include the user job seeker logging in to the job search support system 1, and the display of a management screen on the job seeker terminal 30 for registration information of the job seeker.

[0107] The presentation unit 115 may select and present some of the recommended job offers from the extracted plurality of recommended job offers at predetermined intervals. The period for switching the recommended job offers is, for example, one day. That is, the presentation unit 115 may present recommended job offers on a daily basis. For example, the presentation unit 115 may select job offers to be presented from the plurality of recommended job offers in descending order of recommendation score, and after a predetermined interval, may repeat the process of selecting job offers to be presented from the remaining recommended job offers in descending order of recommendation score again so that the same job offer to be presented is not selected.

[0108] The presentation unit 115 may select and present some recommended categories from the estimated plurality of recommended categories for each predetermined period. The period for switching recommended categories may be, for example, one day. That is, the presentation unit 115 may present recommended categories on a daily basis. The period for switching recommended categories may be the same as or different from the period for switching recommended job offers. For example, the presentation unit 115 may change only the recommended job offers presented in association with the recommended categories without changing the presented recommended categories by making the period for switching recommended categories longer than the period for switching recommended job offers.

[0109] The estimation of recommended categories by the estimation unit 113 and / or the extraction of recommended job offers by the extraction unit 114 may be performed in accordance with the period during which the recommended categories and / or recommended job offers presented by the presentation unit 115 are switched. In other words, the timing of switching the presented recommended categories and / or recommended job offers and the timing of estimating recommended categories and / or extracting recommended job offers may be synchronized. For example, if the presentation unit 115 switches the recommended categories and / or recommended job offers on a daily basis, the estimation unit 113 and / or the extraction unit 114 estimates recommended categories and / or extracts recommended job offers once a day. The presentation unit 115 selects recommended categories and / or recommended job offers for presentation from the recommended categories estimated on a daily basis and / or the recommended job offers extracted on a daily basis.

[0110] When a non-presented category exists among the multiple recommended categories, the presentation unit 115 may present supplementary recommended job offers to the job seeker instead of the non-presented category and the recommended job offers associated with the non-presented category. A non-presented category is a recommended category from which no recommended job offers are extracted or from which the recommendation scores of all extracted recommended job offers are less than a predetermined value. This makes it possible to recommend a certain number of job offers to the user (job seeker) even when there is a recommended category from which an insufficient number of recommended job offers are extracted.

[0111] A non-presented category is typically a recommended category from which the extraction unit 114 was unable to extract any recommended job offers based on the recommended category or recommended keyword (zero search results were obtained). Also included in the non-presented category are recommended categories from which at least one recommended job offer was extracted but none of the recommended job offers had a recommendation score that met the standard and were insufficiently suitable for the job seeker (in other words, recommended categories from which no recommended job offers that were sufficiently suitable for the job seeker could be extracted).

[0112] Specifically, the presentation unit 115 first selects the number of recommended categories to be presented from the recommended categories excluding non-presentable categories (hereinafter referred to as "presentable categories"). If the number of presentable categories is smaller than the number of presented categories (for example, if the number of presentable categories is three and the presentable categories are two), the presentation unit 115 presents the recommended job offers associated with the recommended categories selected for presentation and the auxiliary recommended job offers to the user (job seeker). Since the auxiliary recommended job offers are not associated with the recommended categories, they are presented in association with a title such as "recommended job offers" instead of the recommended category name.

[0113] FIG. 5 is a diagram showing an example of a recommended job offer presentation screen RD displayed on the job seeker terminal 30. The recommended job offer presentation screen RD has a recommended recommendation name RC and recommended job information RI arranged thereon. In the example of FIG. 5, three recommended recommendation names RC are presented. The recommended recommendation name RC includes the name of the recommended recommendation estimated by the estimation unit 113. For example, in the recommended recommendation name RC in FIG. 5, "Startup Jobs," "Startup" corresponds to the name of the recommended recommendation. The recommended job information RI is a card-shaped object showing information about one recommended job offer. In the example of FIG. 5, the recommended job information RI includes the organization name JC, the job title JT of the recommended job offer, the expected annual salary JI, etc.

[0114] On the recommended job presentation screen RD, multiple recommended job information RI are arranged in a row on the left and right below the recommended recommendation name RC, and are associated with the recommended recommendation name RC. A job seeker can switch the display of recommended job information RI corresponding to each recommended recommendation name RC by sliding the area on the job seeker terminal 30 where the recommended job information RI is displayed left and right. In addition, a job seeker can select any recommended job information RI on the job seeker terminal 30 to view the job corresponding to the recommended job information RI, register it in a bookmark list, apply for it, etc.

[0115] <Job Search Department 116> The job search unit 116 is configured to accept a job search from a job seeker targeting job openings registered in the job database. Specifically, the job search unit 116 displays information about the recommended categories estimated by the estimation unit 113 as search condition candidates, accepts input of search conditions from the job seeker, and executes a job search based on the search conditions. This makes it possible to present new search conditions to the job seeker (user) based on the job seeker's own potential desired conditions.

[0116] The "information related to recommended categories" includes recommended keywords associated with the recommended categories in addition to the recommended categories themselves. The job search unit 116 may present only the recommended categories estimated by the estimation unit 113 as search condition candidates, or may present only the recommended keywords estimated by the estimation unit 113, or may present both the recommended categories and the recommended keywords. The job search unit 116 may also present the names of the recommended categories as search condition candidate keywords (part of the recommended keywords).

[0117] 6 is a diagram showing an example of a job search screen SD displayed on the job seeker terminal 30. The job search screen SD includes a search keyword input field IF, a search button B1, a keyword presentation area KA, and a category presentation area CA. The search keyword input field IF accepts input of any keyword used by the job seeker to search for a job. When a search keyword is entered in the search keyword input field IF and the search button B1 is operated, a job search is performed using the entered search keyword as a search condition.

[0118] In the keyword presentation area KA, recommended keywords are displayed as keyword tags KT, which are a type of object. By selecting and inputting a keyword tag KT, a job seeker can add the recommended keyword displayed as the keyword tag KT to the search keyword input field IF.

[0119] In the category presentation area CA, recommended categories are displayed as category items CI. By selecting a check box CB provided for each category item CI, a job seeker can add the recommended category displayed as the category item CI to the search criteria. That is, the job search unit 116 searches for jobs based on the selected recommended category and the search keyword entered in the search keyword input field IF. For example, the job search unit 116 searches for jobs that include information corresponding to the recommended category selected as a search criteria and that include the entered search keyword, and displays the results on the job seeker terminal 30.

[0120] <Scout display control unit 117> The scout display control unit 117 is configured to display scout documents sent from employers to the user (job seeker) on the job seeker terminal 30. Specifically, the scout display control unit 117 assigns a label indicating that the scout document is related to the recommended category to the scout document received by the job seeker that includes information about the recommended category estimated by the estimation unit 113. This makes it possible to present the job seeker with scout documents that match the potential desired conditions of the user (job seeker).

[0121] A "scout document containing information related to recommended categories" (hereinafter referred to as a "recommended scout document") is a scout document in which the body of the scout document or the job posting attached to the scout document contains information related to the recommended categories estimated by the estimation unit 113, or keywords that are identical to or similar to the recommended keywords estimated by the estimation unit 113. A recommended scout document is typically a scout document created based on a recommended job posting.

[0122] The scout display control unit 117 assigns to the recommended scout document, for example, a label indicating a recommended category or a recommended keyword related to the scout document. The scout display control unit 117 may assign to the recommended scout document a label indicating the name of a recommended category, may assign to the recommended scout document a label indicating a recommended keyword, or may assign to the recommended scout document a label indicating both a recommended category and a recommended keyword.

[0123] FIG. 7 is a diagram showing an example of a scout document list screen LD displayed on the job seeker terminal 30. The scout document list screen LD displays a list of multiple scout document information SI. In the example of FIG. 7, the scout document information SI includes the sender (organization name), a summary of the scout document (part of the main text), etc. Furthermore, a label LB of the related recommended category is attached to the scout document information SI of the recommended scout document. By selecting any scout document information SI on the job seeker terminal 30, the job seeker can view the scout document corresponding to the scout document information SI, reply to the scout document, etc. Note that as the label LB attached to the scout document information SI, a label indicating a recommended keyword may be displayed instead of a label indicating a recommended category.

[0124] <Artificial Intelligence Department 120> The artificial intelligence unit 120 is configured to receive input from each functional unit and return the instructed output. The artificial intelligence used by the server device 10 in each functional unit may be a common one, or may be prepared individually for each functional unit.

[0125] The artificial intelligence unit 120 is an AI (Artificial Intelligence) equipped with learning models such as Transformers including GPT (Generative Pretrained Transformer, including GPT-1, GPT-2, and GPT-3), BERT (Bidirectional Encoder Representations from Transformers), and BART (Bidirectional and Auto-regressive Transformer), and language models such as Recurrent Neural Networks (RNNs), and may include generative AI.

[0126] The language model is an example of a learning model based on a machine learning algorithm. Specific examples of machine learning algorithms include nearest neighbor methods, naive Bayes methods, decision trees, support vector machines, and deep learning using neural networks. The artificial intelligence unit 120 can apply the above algorithms as appropriate.

[0127] 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). The training data consists of pairs of input data for learning and output data (correct answer data). Furthermore, the language model may not only be trained for a specific task, but also be a general-purpose model that can be used for a wide range of tasks.

[0128] The artificial intelligence unit 120 may be a general-purpose natural language processing learning model, such as a large language model (LLM), which has learned a huge amount of data. An LLM is a learning model that has previously learned a large amount of data, such as text data (e.g., (i) web content on the Internet, or (ii) data stored in a specified database), and can execute various language processing tasks when given a task. Such a general-purpose learning model includes a language model that can handle various tasks without fine-tuning, using one-shot learning, few-shot learning, etc. Furthermore, a general-purpose learning model may also be configured to handle various tasks using zero-shot learning. The artificial intelligence used in each functional unit of the control unit 11 may be a separate learning model, or a common general-purpose learning model.

[0129] The learning models included in the artificial intelligence unit 120 (learning models used in each functional unit, such as a category estimation model and a keyword estimation model) can undergo additional learning as transfer learning or fine tuning. For example, each time new job seeker registration information and new job postings are registered, the artificial intelligence unit 120 may perform additional learning and fine tuning using these as new training data. This improves the accuracy of the information output from the learning models.

[0130] The learning model included in the artificial intelligence unit 120 may be a learning model (distilled model) obtained by knowledge distillation using an original learning model. In knowledge distillation, a trained model such as a large-scale language model is used as a teacher model, and the parameters of the student model (distilled model) are adjusted to minimize the output loss (Soft Target Loss) of the student model (distilled model) relative to the output (Soft Target) of the teacher model, thereby training the student model, which becomes the distilled model. Alternatively, the student model may be trained to minimize the output loss (Hard Target Loss) of the student model relative to the correct label (Hard Target) of the teacher data (combination of input data and output data of the learning model). Compared to the original learning model (teacher model), the distilled model has a smaller number of parameters and a smaller processing load while maintaining performance similar to the learning model. Therefore, using a distilled model can reduce the cost of the job search support system 1.

[0131] For example, the category estimation model, keyword estimation model, etc. may be distilled models trained using a combination of input data and output data in a large-scale language model as training data. Furthermore, when the job search support system 1 is introduced, a large-scale language model may be used as the category estimation model, keyword estimation model, etc., and once training data from the large-scale language model has been accumulated, the distilled models obtained by knowledge distillation using the training data may be used as the category estimation model, keyword estimation model, etc.

[0132] <Display section> The display unit 211 of the recruiting party terminal 20 and the display unit 311 of the job seeker terminal 30 each display a screen indicated by the screen data transmitted from the server device 10.

[0133] <Operation acquisition section> The operation acquisition unit 212 of the recruiting party terminal 20 accepts operations by a user (recruiter) who uses the recruiting party terminal 20. The operation acceptance unit 312 of the job seeker terminal 30 accepts operations by a user (job seeker) who uses the job seeker terminal 30.

[0134] 3. Information Processing Method This section describes an information processing method of the server device 10. This information processing method is executed by a computer, with each unit of the server device 10 acting as each step.

[0135] This information processing includes an acquisition step, an estimation step, an extraction step, and a presentation step. In the acquisition step, information on candidate job openings selected by the job seeker is acquired from job openings registered in the job opening database. In the estimation step, at least one recommended category is estimated based on the content of the candidate job openings and the first reference information. In the extraction step, at least one recommended job opening to be recommended to the job seeker is extracted from the job openings registered in the job opening database based on the recommended category or recommended keywords related to the recommended category and the second reference information. In the presentation step, the recommended category and the recommended job opening are associated and presented to the job seeker.

[0136] 8 is an activity diagram showing the flow of information processing (recommended job offer presentation processing) executed by job search support system 1. Below, the information processing will be explained along with each activity in this activity diagram.

[0137] The process of presenting recommended job offers is initiated by a predetermined operation by the job seeker. The predetermined operation is an operation by which server device 10 starts the process of presenting recommended job offers, and may be an indirect operation such as the job seeker logging in to job search support system 1, or a direct operation such as the job seeker issuing an instruction to extract recommended job offers. The job seeker inputs such a predetermined operation on the job seeker terminal 30 (activity A101). In response to the predetermined operation on the job seeker terminal 30, server device 10 acquires information on candidate job offers (activity A102).

[0138] The server device 10 estimates recommended categories and recommended keywords from the acquired information on candidate job offers (activity A103). Next, the server device 10 extracts recommended job offers based on the recommended categories and / or recommended keywords (activity A104). Furthermore, the server device 10 associates the extracted recommended job offers with the recommended categories and outputs them to the job seeker terminal 30 (activity A105). As a result, the recommended job offers are associated with the recommended categories and displayed (presented) on the job seeker terminal 30 (activity A106).

[0139] 4. Effect The operation of this embodiment can be summarized as follows. That is, job offers that meet the job seeker's desired conditions can be presented based on the job offer selection (viewing, bookmarking, etc.) made by the job seeker. Therefore, job offers can be presented using an approach that differs from the search conditions entered by the job seeker. Furthermore, by presenting recommended categories, the job seeker can grasp his or her desired conditions that have not been verbalized.

[0140] Although the embodiment of the present invention has been described above, the present invention is not limited to this and can be modified as appropriate within the scope of the technical idea of ​​the invention.

[0141] 5.Other In the above embodiment, the server device 10 performs various storage and control functions. However, multiple external devices may be used instead of the server device 10. That is, various information and programs may be distributed and stored in multiple external devices using blockchain technology or the like. In particular, the artificial intelligence unit 120 may be an external component of the server device 10. In this case, the external artificial intelligence unit 120 may be provided, for example, by an artificial intelligence service server and configured to receive inputs 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 LLM. The artificial intelligence service server receives inputs of prompts such as text, images, and voice, and generates and responds to the prompts.

[0142] The control unit 11 does not necessarily have to display information about recommended categories as search condition candidates in the job search unit 116. In addition, the control unit 11 does not necessarily have to assign a label indicating that the scout document is related to a recommended category in the scout display control unit 117.

[0143] The aspect of this embodiment is not limited to the job search support system 1, but may also be an information processing method or a program. The job search support method includes steps executed by the job search support system 1. The program causes a computer to execute the steps of the job search support system 1.

[0144] It may be provided in the following manner.

[0145] (1) A job search support system comprising at least one processor, the processor being configured to execute the following steps by reading a program: in the acquisition step, information on candidate job openings selected by a job seeker from among job openings registered in a database is acquired; in the estimation step, at least one recommended category is estimated based on the content of the candidate job openings and first reference information, wherein the recommended category is a category in which the job seeker is interested from categories for classifying information contained in job openings based on attributes, and the first reference information includes a correlation between the content of the candidate job openings and the recommended category; in the extraction step, at least one recommended job opening to be recommended to the job seeker from among job openings registered in the database based on the recommended category or recommended keywords related to the recommended category and second reference information, wherein the second reference information includes a correlation between the recommended category or the recommended keyword and the recommended job opening; and in the presentation step, the recommended category and the recommended job opening are associated and presented to the job seeker.

[0146] (2) In the job search support system described in (1) above, the first reference information includes a category estimation model that has been trained to receive input of the content of the candidate job offer and output the recommended category, and in the estimation step, the content of the candidate job offer is input to the category estimation model and the category estimation model is caused to output the recommended category.

[0147] (3) In the job search support system described in (2) above, the category estimation model is trained to output the recommended category corresponding to the candidate job from among the multiple candidate categories by building a correspondence between multiple candidate categories and keywords contained in the job, which correspond to the candidate categories, based on the contents of multiple registered job openings; in the estimation step, the keyword corresponding to the recommended category output by the category estimation model is estimated as the recommended keyword; and in the extraction step, the recommended job opening is extracted based on the recommended keyword and the second reference information.

[0148] (4) In the job search support system described in (1) or (2) above, in the estimation step, the recommended keywords in which the job seeker is interested are estimated based on the content of the candidate job offers and the first reference information, and the recommended categories are estimated based on the recommended keywords, wherein the first reference information further includes a correlation between the content of the candidate job offers and the recommended keywords and a correspondence between the recommended keywords and the recommended categories, and in the extraction step, the recommended job offers are extracted based on the recommended keywords associated with the recommended categories and the second reference information.

[0149] (5) In the job search support system described in (4) above, in the estimation step, the recommended keywords are estimated based on the content of the candidate job offers, the job seeker's registration information, and the first reference information, wherein the first reference information includes correlations between the content of the candidate job offers, the job seeker's registration information, and the recommended keywords.

[0150] (6) In the job search support system described in (5) above, in the estimation step, the recommended keywords are determined based on the magnitude of a keyword score, the value of which increases as the frequency of use in the candidate job offers increases, and the value of which decreases as the frequency of use in the job seeker's registration information increases.

[0151] (7) In the job search support system described in any one of (4) to (6) above, in the estimation step, keywords related to conditions designed to be selected from pre-prepared options in a job search are excluded from the candidates for the recommended keywords.

[0152] (8) In the job search support system described in any one of (1) to (7) above, in the extraction step, the recommended job offers are extracted by a keyword search using the recommended keywords as search conditions.

[0153] (9) In the job search support system described in any one of (1) to (7) above, in the extraction step, the recommended job offers are extracted based on the similarity between the recommended category or the recommended keyword and the content of the job offer.

[0154] (10) In the job search support system described in any one of (1) to (9) above, in the estimation step, a plurality of the recommended categories are estimated, in the extraction step, the recommended job offers are extracted for each of the plurality of recommended categories, and in the presentation step, the recommended job offers are associated with each of the plurality of recommended categories and presented to the job seeker.

[0155] (11) In the job search support system described in (10) above, in the extraction step, at least one auxiliary recommended job offer to be auxiliary recommended to the job seeker is extracted from the job offers registered in the database based on at least one of the content of the candidate job offers and the registration information of the job seeker, and third reference information, wherein the third reference information includes a correlation between the auxiliary recommended job offer and at least one of the content of the candidate job offer and the registration information of the job seeker, and in the presentation step, if a non-presented category exists among the multiple recommended categories, the auxiliary recommended job offer is presented to the job seeker instead of the non-presented category and the recommended job offers associated with the non-presented category, wherein the non-presented category is a recommended category from which no recommended job offer is extracted or from which the recommendation scores of all extracted recommended job offers are less than a predetermined value, and the recommendation score is calculated based on at least one of the similarity between the content of the recommended job offer and the registration information of the job seeker and the probability of occurrence of an action related to the recommended job offer.

[0156] (12) In the job search support system described in any one of (1) to (11) above, in the extraction step, a plurality of recommended job offers are extracted for each of the recommended categories, and a recommendation score for each of the recommended job offers is calculated based on at least one of the similarity between the content of the recommended job offers and the registered information of the job seeker and the probability of an action related to the recommended job offers occurring, and in the presentation step, the extracted plurality of recommended job offers are presented to the job seeker in an order based on their respective recommendation scores.

[0157] (13) In the job search support system described in any one of (1) to (12) above, in the extraction step, a plurality of recommended job offers are extracted for each of the recommended categories, and in the presentation step, a recommended job offer selected at random from the extracted plurality of recommended job offers is presented to the job seeker.

[0158] (14) In the job search support system described in any one of (1) to (13) above, the processor is configured to further execute the following steps: in the job search step, information about the recommended categories is displayed as search condition candidates, while input of the search conditions is accepted from the job seeker, and a job search is performed based on the search conditions.

[0159] (15) In the job search support system described in any one of (1) to (14) above, the processor is configured to further perform the following steps: in the scout display control step, a label indicating that the scout document received by the job seeker is related to the recommended category is assigned to the scout document that includes information about the recommended category.

[0160] (16) A job search support method comprising the steps executed by the job search support system described in any one of (1) to (15) above.

[0161] (17) A program for causing a computer to execute each step of the job search support system described in any one of (1) to (15) above. Of course, this is not the case.

[0162] Finally, while various embodiments of the present disclosure have been described, they are presented as examples and are not intended to limit the scope of the invention. The novel embodiments may be embodied in various other forms, and various omissions, substitutions, and modifications may be made without departing from the spirit of the invention. Such embodiments and modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as defined in the claims. [Explanation of symbols]

[0163] 1: Job search support system 2: Communication line 10: Server device 11: Control section 12: Storage section 13: Communications Department 14: Communication bus 20: Recruiter terminal 21: Control unit 22: Storage section 23: Communications Department 24: Input section 25: Output section 26: Communication bus 30: Job seeker terminal 31: Control unit 32: Storage section 33: Communications Department 34: Input section 35: Output section 36: Communication bus 111: Basic display control section 112: Acquisition Department 113: Estimation section 114:Extraction part 115: Presentation part 116: Job Search Department 117: Scout display control unit 120: Artificial Intelligence Department 211:Display section 212: Operation acquisition section 311: Display section 312: Operation reception section

Claims

1. A job search support system, at least one processor; The processor is configured to execute the following steps by reading the program: In the acquisition step, information on the candidate job that the job seeker selected from the job listings registered in the database is acquired, In the estimation step, at least one recommended category is estimated based on the content of the candidate job offer and first reference information, wherein the recommended category is a category in which the job seeker is interested among categories for classifying information included in the job offer based on attributes, and the first reference information includes a correlation between the content of the candidate job offer and the recommended category; In the extraction step, at least one recommended job offer to be recommended to the job seeker is extracted from the job offers registered in the database based on the recommended category or the recommended keyword related to the recommended category and second reference information, wherein the second reference information includes a correlation between the recommended category or the recommended keyword and the recommended job offer; In the presenting step, the job search support system presents the recommended categories and the recommended job offers in association with each other to the job seeker.

2. The job-seeking support system according to claim 1, the first reference information includes a category estimation model that is trained to receive content of the candidate job offer as an input and output the recommended category; In the estimation step, the job-seeking support system inputs the contents of the candidate job offers into the category estimation model, and causes the category estimation model to output the recommended category.

3. 3. The job-seeking support system according to claim 2, the category estimation model is trained to build correspondences between a plurality of candidate categories and keywords included in the job postings that correspond to the candidate categories based on the contents of a plurality of registered job postings, and to output the recommended category that corresponds to the candidate job posting from among the plurality of candidate categories; In the estimation step, the keyword corresponding to the recommended category output by the category estimation model is estimated as the recommended keyword; In the extracting step, the recommended job offers are extracted based on the recommended keywords and the second reference information.

4. The job-seeking support system according to claim 1, In the estimation step, the recommended keywords in which the job seeker is interested are estimated based on the content of the candidate job offer and the first reference information, and the recommended categories are estimated based on the recommended keywords, wherein the first reference information further includes a correlation between the content of the candidate job offer and the recommended keywords, and a correspondence between the recommended keywords and the recommended categories; In the extracting step, the recommended job offers are extracted based on the recommended keywords associated with the recommended categories and the second reference information.

5. 5. The job-seeking support system according to claim 4, A job search support system in which, in the estimation step, the recommended keywords are estimated based on the content of the candidate job offers, the job seeker's registration information, and the first reference information, wherein the first reference information includes correlations between the content of the candidate job offers, the job seeker's registration information, and the recommended keywords.

6. 6. The job-seeking support system according to claim 5, In the estimation step, the job search support system determines the recommended keywords based on the magnitude of a keyword score, which increases the value the more frequently the keyword is used in the candidate job offers and decreases the value the more frequently the keyword is used in the job seeker's registered information.

7. 5. The job-seeking support system according to claim 4, In the estimation step, the job search support system excludes keywords related to conditions designed to be selected from options prepared in advance in a job search from the recommended keyword candidates.

8. The job-seeking support system according to claim 1, In the extraction step, the job-seeking support system extracts the recommended job offers by a keyword search using the recommended keywords as search conditions.

9. The job-seeking support system according to claim 1, In the extraction step, the recommended job offers are extracted based on the similarity between the recommended category or the recommended keyword and the content of the job offer.

10. The job-seeking support system according to claim 1, In the estimation step, a plurality of the recommendation categories are estimated, In the extraction step, the recommended job offers are extracted for each of the plurality of recommended categories; In the presenting step, the job search support system associates the recommended job offers with each of the plurality of recommended categories and presents them to the job seeker.

11. The job-seeking support system according to claim 10, In the extraction step, at least one supplementary recommended job offer to be supplementarily recommended to the job seeker is extracted from the job offers registered in the database based on at least one of the contents of the candidate job offers and the registered information of the job seeker, and third reference information; wherein the third reference information includes a correlation between the content of the candidate job offer and at least one of the registered information of the job seeker and the supplementary recommended job offer, In the presentation step, if a non-presented category exists among the plurality of recommended categories, the supplementary recommended job offer is presented to the job seeker in place of the non-presented category and the recommended job offer associated with the non-presented category; Here, the non-presented category is a recommended category from which no recommended job offers are extracted or from which all of the extracted recommended job offers have recommendation scores less than a predetermined value, and the recommendation score is calculated based on at least one of the similarity between the content of the recommended job offers and the registered information of the job seeker and the probability of an action related to the recommended job offers occurring.

12. The job-seeking support system according to claim 1, In the extraction step, a plurality of recommended job offers are extracted for each of the recommended categories, and a recommendation score for each of the recommended job offers is calculated; wherein the recommendation score is calculated based on at least one of a degree of similarity between the content of the recommended job offer and the registered information of the job seeker and a probability of an action related to the recommended job offer occurring; In the presenting step, the extracted recommended job offers are presented to the job seeker in an order based on the respective recommendation scores.

13. The job-seeking support system according to claim 1, In the extraction step, a plurality of recommended job offers are extracted for each of the recommended categories; In the presenting step, the job seeker is presented with the recommended job offers selected at random from the extracted plurality of recommended job offers.

14. The job-seeking support system according to claim 1, The processor is further configured to perform the steps of: In the job search step, the job search support system displays information about the recommended categories as search condition candidates, accepts input of the search conditions from the job seeker, and executes a job search based on the search conditions.

15. The job-seeking support system according to claim 1, The processor is further configured to perform the steps of: In the scout display control step, a job search support system assigns a label indicating that the scout document received by the job seeker includes information related to the recommended category.

16. A job search support method, comprising: A job search support method comprising the steps executed by the job search support system according to any one of claims 1 to 15.

17. A program, A program for causing a computer to execute each step of the job search support system according to any one of claims 1 to 15.

Images