Retrieval support system, retrieval support method and program

Abstract

To provide a retrieval support system, a retrieval support method and a program capable of reducing the cost such as time and effort required for retrieval.SOLUTION: In the retrieval support system, a control unit of a server unit includes: a reception unit that receives an input of a retrieval query including at least one retrieval keyword; and an acquisition unit that acquires retrieval information corresponding to the retrieval query received in a reception step from a recommended job applicant database for retrieval from a recommended job applicant database for retrieval. The retrieval information is information that is registered in the recommended job applicant database for retrieval associated with the retrieval query including at least one keyword used for job applicant retrieval and identifies attributes of recommended job applicants for retrieval. The recommended job applicants for retrieval are job applicants who are expected to take actions related to closing a job offer on the basis of past retrieval results based on the retrieval query. A retrieval unit retrieves job applicants by using the retrieval query received by the reception unit and the retrieval information as retrieval conditions.SELECTED DRAWING: Figure 4

Description

[Technical Field]

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

[0002] As disclosed in Patent Document 1, a technique is known in which an employer searches for job seekers who meet certain conditions based on job seeker information registered by the job seeker. [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 the above-mentioned conventional technology, the employer may have to change the search conditions and repeat the search until he or she finds a job seeker who meets the requirements, which results in high search costs.

[0005] In view of the above circumstances, the present invention provides a search support system and the like that can reduce costs such as the effort and time required for searching. [Means for solving the problem]

[0006] According to one aspect of the present invention, a search support system is provided. The search support system includes at least one processor. The processor is configured to execute the following steps by reading a program. In the receiving step, input of a search query including at least one search keyword is received. In the acquiring step, search information corresponding to the search query received in the receiving step is obtained from a database of recommended job seekers for search. The search information is information that identifies attributes of recommended job seekers for search that are associated with a search query including at least one keyword and registered in the database of recommended job seekers for search, and the recommended job seekers for search are job seekers who are assumed to have a high expected value of occurring an action related to the conclusion of a job offer, based on past search results based on the search query. In the searching step, a job seeker is searched for using the search query and search information received in the receiving step as search conditions.

[0007] According to this aspect, when a recruiter inputs a search query to search for a job seeker, information (search information) about recommended job seekers for search who are expected to take action related to the conclusion of a job offer can be used to assist in narrowing down the job seekers, thereby reducing the cost required for searches by the recruiter. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a configuration diagram showing a 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] FIG. 10 is a schematic diagram showing the procedure for building a database of recommended job seekers for search. [Figure 6]10 is a flowchart showing an example of a process for building a database of recommended job seekers for search, which is executed by the database building unit 115. FIG. [Figure 7] 1 is an activity diagram showing an example of the flow of information processing (job seeker search processing) executed by search 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] <Search Support System 1> Fig. 1 is a configuration diagram showing a search support system 1. The 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] The search support system 1 constitutes, for example, 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). The search support system 1 mainly performs tasks such as allowing recruiters to search for job seekers and sending scouting documents from recruiters to job seekers. In one embodiment, the search support system 1 is comprised of one or more devices or components. These components will be 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 recruiting party terminal 20 is an information processing terminal used in the course of business by each recruiter who is a user belonging to an organization that receives services provided by the server device 10. The description of the control unit 21, the memory unit 22, and the communication unit 23 is omitted here, as they are the same as the description of each unit in the server device 10.

[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 within the job seeker terminal 30 via the communication bus 36. The job seeker terminal 30 is an information processing terminal used by each job seeker who is a user receiving services provided by the server device 10. The description of the control unit 31, the memory unit 32, the communication unit 33, the input unit 34, and the output unit 35 is omitted here as they are the same as the description of each unit in the recruiter terminal 20.

[0026] 2. Functional configuration This section describes the functional configuration of this embodiment. 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 (at least one processor included in the 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, server device 10 (control unit 11) includes a basic display control unit 111, a reception unit 112, an acquisition unit 113, a search unit 114, a database construction unit 115, and an artificial intelligence unit 120. As shown in Fig. 4B, recruiter terminal 20 (control unit 21) includes a display unit 211 and an operation acquisition unit 212. As shown in Fig. 4C, job seeker terminal 30 (control unit 31) includes a display unit 311 and an operation reception 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 job advertisement and scouting document created by the recruiter, registration information created by the job seeker, 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] <Reception Department 112> The reception unit 112 is configured to receive input from the recruiting provider terminal 20 and the job seeker terminal 30. Specifically, the reception unit 112 receives input from the recruiting provider terminal 20 of a search query including at least one search keyword.

[0032] A search keyword is a keyword that is used to determine whether identical or similar keywords exist (keyword matching) in the registration information of job seekers registered in the job seeker database. A search query may consist of one search keyword or multiple search keywords. In other words, the receiving unit 112 may receive input of one search keyword or multiple search keywords as a search query.

[0033] A search query may include search conditions other than search keywords, such as a range of a specific parameter (e.g., annual income) or a condition selected from pre-prepared options (e.g., whether or not remote work is available).

[0034] Furthermore, search keywords may be selected from options prepared in advance, or may be manually input by the user (recruiter) from the recruiter terminal 20. For example, search keywords may be input by the user (recruiter) selecting a search keyword from a list of preset job types, industry types, school names, organization names, etc.

[0035] The job seeker database is stored, for example, in the storage unit 12. The registered information of a job seeker includes the job seeker's resume, curriculum vitae, and other profile information. A "resume" is a document that mainly describes the job seeker's profile, current situation, educational background, work history, desired working conditions, etc., while a "curriculum vitae," also known as a résumé, is a document in which a job seeker conveys to a recruiter his or her work history, experience, skills, qualifications, etc., related to his or her past work. The registered information of a job seeker may also include the industry and occupation the job seeker desires.

[0036] Job seekers have attributes extracted from the registered information, such as the organization in which they have experience, the type of job they have experienced, the skills they possess, the qualifications they possess, their educational background, their annual income, and their age.

[0037] <Acquisition part 113> The acquiring unit 113 is configured to acquire search information corresponding to the search query accepted by the accepting unit 112 from the search recommended job seeker database.

[0038] The search information is information that identifies the attributes of recommended job seekers for search, which are linked to a search query containing at least one keyword used in a job seeker search and registered in a database of recommended job seekers for search. Furthermore, recommended job seekers for search are job seekers who are assumed to have a high expected value of occurrence of an action related to the conclusion of a job offer, based on past search results based on the search query.

[0039] A recommended job seeker for search purposes is, for example, a virtual job seeker having attributes generated based on the registration information of a job seeker registered in the job seeker database. Specifically, a recommended job seeker for search purposes is a job seeker defined by attributes generated by the database construction unit 115, which will be described later.

[0040] Actions related to the conclusion of a job offer include, for example, receiving a scouting document sent from a recruiter, replying to a scouting document sent from a recruiter, passing the screening process for the job offer (document screening, interview screening, etc.), concluding a job offer, etc. The attributes of recommended job seekers for search are the same as the attributes of job seekers registered in the job seeker database, and include, for example, the organization of experience, the type of experience, the skills held, the qualifications held, the educational background, the annual salary, the age, etc.

[0041] The search information includes keywords or sentences that directly represent attributes (e.g., job types experienced) or information about attributes (e.g., work history including descriptions of job types experienced), and feature quantities (e.g., vector data) of the attributes or information about the attributes. The search information may be information that identifies multiple attributes possessed by recommended job seekers for search purposes. In other words, the search information may include keywords, sentences, or feature quantities corresponding to each of the multiple attributes or information about the multiple attributes. The search information may also be the resume and / or curriculum vitae of the recommended job seeker for search purposes.

[0042] In particular, the search information is preferably vector data. In other words, the search information is preferably vector data obtained by converting attributes or information related to attributes of recommended job seekers for search purposes. This makes it possible to reduce the calculation load and calculation time when comparing the search information with the registered information of job seekers in a job seeker database to be searched. Note that the vector data is a vector value obtained by quantifying text data using known techniques such as natural language processing using morphological analysis or encoding of categorical variables. The search information may be expressed as the sum of vector values ​​of multiple attributes (a vector obtained by combining multiple vectors). Furthermore, the vector data constituting the search information may be a dense vector or a sparse vector.

[0043] The recommended job seeker database for search is stored, for example, in the storage unit 12. In the recommended job seeker database for search, information associating multiple search queries (typically a single search keyword or a combination of multiple search keywords) used in searches in the past with search information (recommended job seekers for search) corresponding to each of these search queries is registered in the form of, for example, a table or a list. For example, for the search query "Java" (registered trademark), search information (for example, vector data) that identifies attributes of recommended job seekers for search who are likely to be included in the search results for the search query and who also have a high expected value for the occurrence of actions related to the conclusion of a job offer is linked and recorded in the recommended job seeker database for search. The method of constructing the recommended job seeker database for search will be described later in the explanation of the database construction unit 115.

[0044] The acquisition unit 113 searches the database of recommended job seekers for searches for search queries that are identical or similar to the search query entered by the recruiter performing the search, and acquires search information associated with the searched search query from the database of recommended job seekers for searches as search information corresponding to the search query entered by the recruiter. The similarity of search queries is determined, for example, by comparison using feature quantities such as vector data, by referring to a database in which synonyms, similar words, and spelling variations of words are registered, by judgment using a learning model (large-scale language model), etc.

[0045] Furthermore, when a search query includes multiple search keywords or search conditions other than search keywords (for example, conditions specifying attributes such as industry or occupation), the acquisition unit 113 treats the search query as a string of characters in which multiple search keywords or search conditions converted into keywords are arranged in a predetermined format. The predetermined format is, for example, a format that lists combinations of search query types and their contents.

[0046] For example, if "Java" (registered trademark) is entered as the search keyword and "IT / Internet" is selected as the industry of the search conditions, the search query will be a string that lists a combination of a string representing the type of search query (keyword, selection conditions such as industry, etc.) and its content (input or selected keyword), such as "keyword="Java", industry="IT / Internet". In addition to search queries made up of individual search keywords, the recommended job seeker database for search also stores search queries made up of character strings in the predetermined format as described above, and the acquisition unit 113 acquires search information by comparing search queries made up of such character strings with the recommended job seeker database for search.

[0047] If a search query identical or similar to the search query entered by the employer conducting the search does not exist in the database of recommended job seekers for search, the acquisition unit 113 does not acquire search information, or acquires dummy search information indicating that no search information exists.

[0048] <Search section 114> The search unit 114 is configured to execute a job seeker search based on the search query input by the recruiter. Specifically, the search unit 114 searches for job seekers using the search query received by the reception unit 112 (i.e., input by the user, the recruiter) and the search information acquired by the acquisition unit 113 as search conditions.

[0049] The search unit 114 may search for job seekers by matching the search query received by the reception unit 112 with keywords contained in the registration information of job seekers registered in the job seeker database and by comparing the search information with the registration information. This makes it possible to search for job seekers using the search query while also narrowing down the search results using the search information. This helps reduce search costs.

[0050] In detail, the search unit 114 may extract, as search targets, job seekers whose registered information contains keywords that are the same as or similar to the search keywords contained in the search query received by the reception unit 112. In a keyword search, the similarity of keywords is determined, for example, by comparing features such as vector data, referring to a database in which synonyms, similar words, and spelling variations are registered, or by making a judgment using a learning model (large-scale language model).

[0051] When a search query includes multiple search keywords, the search unit 114 may extract job seekers whose registered information includes all of the multiple search keywords (i.e., keyword search using AND conditions), or may extract job seekers whose registered information includes any of the multiple search keywords (i.e., keyword search using OR conditions).The search unit 114 may also perform a keyword search based on the selection of either the AND condition or the OR condition by the employer performing the search.

[0052] Furthermore, the search unit 114 may adjust the display of job seekers extracted as search targets based on a similarity score that indicates the degree of similarity between the search information and the registered information. The similarity score is, for example, a value obtained by comparing the search information with vector data obtained by vectorizing the registered information. This makes it possible to filter job seekers extracted based on a general keyword search using the search information.

[0053] The vector data of the registration information is obtained by vector-converting the attributes or information related to the attributes included in the registration information in the same manner as the vector data of the search information. The search unit 114 may compare the vector data for each attribute of the search information and the registration information, or may compare the vector data indicating the overall attributes of the search information and the registration information (for example, vector data obtained by combining the vector data for each attribute).

[0054] The similarity score is a score that increases as the vector distance (difference in feature amounts) between the search information and registered information decreases. For example, cosine similarity is used as the vector distance. In this case, the closer the cosine similarity is to 1, the larger the similarity score becomes.

[0055] "Adjusting the display" of job seekers extracted as search targets includes adjusting the display order of job seekers, adjusting whether or not to display job seekers, etc. The search unit 114 may determine the display order of extracted job seekers based on the magnitude of the similarity score. For example, the search unit 114 may display search results on the employer terminal 20 in which job seekers are sorted in descending order of similarity score. This gives priority to job seekers who are more likely to take action related to closing a job offer, thereby reducing the cost of further narrowing down search results, etc., required by employers.

[0056] The search unit 114 may exclude job seekers whose similarity scores are less than a predetermined threshold from the search results to be displayed on the recruiter terminal 20. In other words, the search unit 114 may cause the recruiter terminal 20 to display, as search results, only job seekers whose similarity scores are equal to or greater than a predetermined value.

[0057] If the acquisition unit 113 has not acquired search information or has acquired dummy search information (i.e., if the search information corresponding to the search query entered by the employer does not exist in the database of recommended job seekers for search), the search unit 114 searches for job seekers using only the search query as a search condition. In other words, the search unit 114 only searches for job seekers based on the search query and does not compare the search information with the registered information of the job seekers.

[0058] <Database Construction Department 115> The database construction unit 115 is configured to construct a search recommended job seeker database. As shown in Fig. 4, the database construction unit 115 has a candidate selection unit 115A, a characteristic information acquisition unit 115B, an attribute information generation unit 115C, a search information creation unit 115D, and a registration unit 115E.

[0059] <Candidate Selection Unit 115A> The candidate selection unit 115A is configured to select, as a candidate, a searched job seeker who has performed a predetermined number of actions related to the conclusion of a job offer or whose ranking in the number of actions performed among the searched job seekers is within a predetermined range, from among a plurality of searched job seekers searched for using a registration query, which is a search query previously used in searching for job seekers. The candidate selection unit 115A also obtains identification information (ID) of the job seeker selected as a candidate from the search log. The search log is stored, for example, in the storage unit 12.

[0060] Search queries (registration queries) previously used in job seeker searches are stored in a search log along with search results using the search queries (information on previously searched job seekers). The search log also stores the history of actions taken by previously searched job seekers related to the conclusion of job offers.

[0061] The actions related to the conclusion of a job offer include at least one of receiving a scouting document based on a given job offer, replying to a scouting document from a given job offer, passing the screening for a given job offer, and concluding a job offer for a given job offer. This enables employers to preferentially present job seekers who are likely to be the target of sending scouting documents, job seekers who are likely to reply to scouting documents, job seekers who are likely to pass the screening, job seekers who are likely to conclude a job offer, etc. as search results.

[0062] In particular, the candidate selection unit 115A may select as candidates searched job seekers who have received a predetermined number of scouting documents (for example, several tens or several hundreds) or whose ranking in the number of scouting documents received among the searched multiple job seekers is within a predetermined range (for example, within the top 100). This makes it possible to preferentially search for job seekers who are likely to be the recipients of scouting documents, which are the starting point for actions leading up to the conclusion of a job offer.

[0063] The candidate selection unit 115A may select candidates using a candidate extraction model that has been trained to be able to input a registration query and output a candidate. In this case, the candidate selection unit 115A inputs the registration query to the candidate extraction model and causes the candidate extraction model to output a candidate (typically, the candidate's identification number). The candidate extraction model is a learning model that has been trained using as training data a combination of a registration query (search query) for learning and the registration information of job seekers who are searched for by the registration query and who are highly expected to take an action related to the conclusion of a job offer.

[0064] The candidate extraction model may be a generative AI including a large-scale language model. In this case, the candidate selection unit 115A receives the registration query as input, refers to the search log, and inputs a prompt including an instruction to output candidates to the candidate extraction model, causing the candidate extraction model to output the candidates. The candidate selection unit 115A may generate a prompt that instructs the candidate extraction model to extract candidates based on the registration query and input the prompt to the candidate extraction model. Furthermore, the candidate selection unit 115A may input a prompt to the candidate extraction model that includes, as input and output samples, one or more registration query samples and one or more corresponding candidate samples, in addition to the candidate extraction and output instructions and the registration query. The candidate extraction model extracts candidates based on the registration query in accordance with the input prompt.

[0065] <Characteristic information acquisition unit 115B> The characteristic information acquisition unit 115B is configured to acquire characteristic information representing the attributes of the candidates selected by the candidate selection unit 115A from the characteristic information database. When the candidate selection unit 115A selects multiple candidates, the characteristic information acquisition unit 115B acquires the characteristic information of each candidate. In other words, the characteristic information acquisition unit 115B acquires a group of characteristic information corresponding to the group of candidates.

[0066] The characteristic information is text data or feature quantities (vector data) that represent the candidate's experience in the organization, the type of job they have worked in, the skills they possess, the qualifications they possess, their educational background, their annual income, their age, and so on.

[0067] The characteristic information database may be a database in which attribute information extracted from a job seeker database in which the registration information of job seekers is registered is linked to the job seeker's identification information and registered. In this case, the characteristic information acquisition unit 115B acquires attribute information corresponding to the identification information (candidate ID) acquired by the candidate selection unit 115A from the characteristic information database as characteristic information to be used by the search information creation unit 115D. This makes it possible to extract the candidate's characteristic information without directly referencing the job seeker's registration information contained in the job seeker database. As a result, the actual job seeker's information is prevented from being used in various information processing operations or from being input into a learning model such as a search information creation model used by the search information creation unit 115D. Furthermore, the output accuracy of the search information creation model is improved. The characteristic information database for acquiring attribute information is constructed by the attribute information generation unit 115C. The attribute information may be text data or feature quantities such as vectors.

[0068] Furthermore, the characteristic information database may be, for example, a job seeker database (or its master database). In this case, the characteristic information acquisition unit 115B extracts characteristic information to be used by the search information creation unit 115D from the registration information of the job seeker corresponding to the identification information (candidate ID) acquired by the candidate selection unit 115A. Note that when characteristic information is acquired directly from the job seeker database in this way, the database construction unit 115 does not necessarily have to include the attribute information generation unit 115C described below.

[0069] <Attribute information generation unit 115C> The attribute information generation unit 115C is configured to construct a characteristic information database from the job seeker database. Specifically, the attribute information generation unit 115C registers, in the characteristic information database, information obtained by replacing each piece of registration information of a job seeker included in the job seeker database with a combination of the job seeker's identification information and attribute information extracted from the registration information. In other words, the characteristic information database stores the job seeker's attribute information in association with the job seeker's identification information.

[0070] The attribute information extracted from the job seeker database (i.e., the attribute information registered in the feature information database) should include topics that are generated by inputting the registered information into a topic model and outputting it from the topic model. This increases the accuracy of extracting the attributes of individual job seekers.

[0071] Here, the topic model is a model that assumes that a document (here, registered information) is generated probabilistically from multiple latent topics, and that each word (keyword) contained in a document appears according to the probability distribution of the words in the topic. The topic model outputs topics according to the number of topics specified. The attribute information generation unit 115C inputs registered information from the job seeker database into the topic model. Based on keywords extracted from the registered information, the topic model outputs a probability distribution of topics (the probability of occurrence of each of multiple topics) and a probability distribution of each keyword in each topic (the probability of occurrence of each of multiple keywords). In this way, the characteristics of the registered information are acquired.

[0072] The attribute information generation unit 115C registers the topics output by the topic model in the feature information database as attribute information associated with the identification information included in the registration information input to the topic model. The attribute information includes topic names (attribute names) corresponding to the topics output by the topic model. The topic names may be set manually based on, for example, the distribution of keywords included in each topic. Alternatively, the keyword with the highest frequency of appearance in each topic may be set as the topic name. Furthermore, the topic names may be obtained by inputting the distribution of keywords included in each topic into a topic name determination model and outputting the results. The topic name determination model is a learning model trained to receive keyword distributions as input and output topic names that represent the themes associated with these keyword distributions. Here, the topic name determination model may select and output topic names from a pre-created list of skills, keywords, etc. For example, the topic name determination model may be trained to output topic names from a list of keywords, skills, qualifications, etc. used by a large number of job seekers registered in a job seeker database.

[0073] The attribute information generation unit 115C may convert the topic name into vector data and then register it in the characteristic information database. The attribute information generation unit 115C executes this process for the registered information of all job seekers who are to be registered in the characteristic information database.

[0074] A topic model is a learning model that builds correspondences between multiple topics and keywords contained in the registered information that correspond to the topics, based on the contents of multiple pieces of registered information (typically all of the registered information) registered in a job seeker database. The topic model is trained to output at least one topic that corresponds to the input registered information. Note that analysis methods used in topic models include LDA (Latent Dirichlet Allocation), LSI (Latent Semantic Indexing), and PLSA (Probabilistic Latent Semantic Analysis). Analysis methods other than these may also be used in topic models.

[0075] As described above, the topic names used as attribute information may be, for example, extracted keywords, skills, qualification information, etc. used by many job seekers registered in the job seeker database. In this case, the attribute information generation unit 115C inputs the extracted information such as keywords and skills into a topic model and estimates the topic distribution of this information. For example, it estimates the occurrence probability of each of multiple topics for the skill "Java" (registered trademark). It may calculate the distance (similarity) between the topic distribution of the estimated information such as skills and the topic distribution in the registered information of the job seeker, and use the information such as skills having a topic distribution similar to the topic distribution of the registered information as the attribute information of that registered information.

[0076] In the topic model construction (learning) stage, for example, first, for sentences contained in all or some of the referable registered information registered in the job seeker database (for example, registered information of job seekers with specific attributes such as a specific occupation 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 estimate the probability distribution (probability of appearance) of topics in each registered information and the probability distribution (probability of appearance) of keywords in each topic. The probability distribution of keywords in each topic corresponds to the "correspondence relationship between the topic and the keywords contained in the registered information."

[0077] The topic model constructed in this way receives registered information as input and outputs a probability distribution of topics in the registered information based on keywords included in the registered information.The topic model then outputs, for example, at least one topic from among multiple topics estimated from the registered information whose appearance frequency (probability) is equal to or greater than a predetermined threshold.The topic model may also output, from among the multiple estimated topics, topics whose appearance frequency is equal to or greater than a predetermined value, or topics whose appearance frequency is within a predetermined ranking (for example, within the top three).

[0078] The attribute information generation unit 115C may extract attribute information from registered information using a learning model other than a topic model. For example, the attribute information generation unit 115C may use an attribute information extraction model, which is a generative AI including a large-scale language model. In this case, the attribute information generation unit 115C receives registered information as input, inputs a prompt including an instruction to output attribute information to the attribute information extraction model, and causes the attribute information extraction model to output the attribute information. The attribute information generation unit 115C may generate a prompt that instructs the attribute information extraction model to extract attribute information from registered information and input the prompt to the attribute information extraction model. Furthermore, the attribute information generation unit 115C may input a prompt that includes, as input and output samples, one or more registered information samples and one or more corresponding attribute information samples, in addition to the instruction to output the attribute information and the registered information, to the attribute information extraction model. The attribute information extraction model extracts attribute information from registered information according to the input prompt.

[0079] The attribute information extracted from the job seeker database may include skills or qualifications extracted from the registration information, thereby assisting in narrowing down the search results for job seekers based on the job seeker's skills or qualifications.

[0080] The attribute information generation unit 115C extracts text data such as keywords and sentences indicating skills or qualifications from the registered information using, for example, natural language processing such as morphological analysis, a learning model (large-scale language model), etc. The attribute information generation unit 115C may also extract skills or qualifications by referring to tags indicating attributes such as skills and qualifications that are previously assigned to the registered information. The attribute information generation unit 115C registers the extracted text data or tags in the characteristic information database as attribute information associated with identification information included in the registered information from which the skills or qualifications were extracted. The attribute information generation unit 115C may also convert the extracted text data or tags into vector data before registering it in the characteristic information database. The attribute information generation unit 115C performs this process for the registered information of all job seekers to be registered in the characteristic information database.

[0081] Furthermore, the attribute information may include the degree of experience with a skill. The degree of experience with a skill may be expressed, for example, by the period of skill use (years of work experience), the number of occupations or industries in which the skill has been used, or the like. For example, when extracting a skill, the attribute information generation unit 115C further extracts text data such as keywords and sentences that indicate the degree of experience with the skill from the registered information, and converts the extracted text data into feature amounts (vector data) as necessary. The attribute information generation unit 115C registers the extracted skill and the degree of experience with the skill in the feature information database as attribute information.

[0082] When the attribute information in the characteristic information database includes the degree of skill experience, the characteristic information acquisition unit 115B may select skills to acquire as characteristic information based on the degree of the candidate's skill experience. As a result, the skills used to generate recommended job seekers for search (search information) in the search information creation unit 115D are carefully selected, and therefore search information that can present more accurate search results is generated in the database construction unit 115.

[0083] <Search information creation unit 115D> The search information creation unit 115D is configured to input at least one piece of characteristic information acquired by the characteristic information acquisition unit 115B into the search information creation model and cause the search information creation model to output search information. The search information creation model is a learning model that has been trained so as to be able to input characteristic information and output search information that identifies the attributes of recommended job seekers for search purposes. In other words, the search information creation model inputs characteristic information that is information about actual job seekers, and creates and outputs information about virtual job seekers.

[0084] When the characteristic information acquisition unit 115B acquires characteristic information (a group of characteristic information) of multiple candidates, the search information creation unit 115D inputs the group of characteristic information into the search information creation model. The search information creation model creates one piece of search information from the group of characteristic information.

[0085] The search information creation model is a learning model that is trained using a combination of learning feature information and search information (e.g., resumes and / or curriculum vitae) that identifies the attributes of recommended job seekers for search purposes characterized by the feature information as training data.

[0086] The search information creation model may be a generative AI including a large-scale language model. In this case, the search information creation unit 115D inputs at least one piece of feature information, inputs a prompt including an instruction to output search information to the search information creation model, and causes the search information creation model to output the search information. The search information creation unit 115D may generate a prompt that instructs the search information creation model to create search information from the feature information, and input the prompt to the search information creation model. Furthermore, the search information creation unit 115D may input a prompt that inserts, for example, one or more sample feature information and one or more corresponding sample search information as input and output samples to the search information creation model, in addition to the instruction to create and output search information and the feature information. The search information creation model creates search information from the feature information according to the input prompt.

[0087] The search information creation unit 115D may cause the search information creation model to output the search information as text data and convert the text data into vector data. This makes it possible to make the attributes and descriptions contained in the search information closer to the attributes of actual job seekers and the descriptions in the registration information created by job seekers, while reducing the calculation load and calculation time required when comparing the search information with the information registered in the job seeker database. As a result, it is possible to improve the accuracy of job seeker searches using the search information and also improve the processing efficiency of the search.

[0088] Specifically, the search information creation unit 115D may insert, into a prompt to be input into the search information creation model, which is a large-scale language model, instructions to create text data (numerical values, keywords, or sentences) corresponding to each item included in the job seeker's registered information (resume and / or curriculum vitae), and instructions to output search information in the same format as the registered information. Items included in the registered information may include, for example, age, gender, address, current occupation, current industry, current annual salary, desired annual salary, desired work location, desired occupation, desired industry, job summary, work history (including job title, company name, department / position, length of employment, detailed description, etc.), experience, skills, abilities, self-promotion, qualifications, awards, and educational background. In addition to instructions to create text data corresponding to each item included in the registered information, the search information creation unit 115D may also insert samples of each item into the prompt to the search information creation model. Furthermore, the search information creating unit 115D may attach to the prompt a sample or format data of the job seeker's registration information, candidates for each item (for example, a list of job titles), and the like.

[0089] The search information created as text data may be vector-converted for each item (attribute), or the entire search information may be vector-converted. When the search information is vector-converted for each item, the search information converted into vector data is composed of a set of multiple vectors or a composite value of multiple vectors.

[0090] The search information creation unit 115D may input all of the feature information acquired by the feature information acquisition unit 115B from the feature information database into the search information creation model, or may input only some of the feature information into the search information creation model. For example, the search information creation unit 115D may input, among the feature information acquired by the feature information acquisition unit 115B, feature information in which the number of candidates who share the feature information (i.e., shared candidates having the same feature information) is greater than or equal to a predetermined number, or the acquired feature information is ranked within a predetermined number of shared candidates, into the search information creation model. Specifically, when the feature information acquisition unit 115B acquires multiple skills as a feature information group from the feature information database, skills among these skills that are possessed by a predetermined number of candidates or more are input as feature information into the search information creation model.

[0091] <Registration Section 115E> The registration unit 115E is configured to link the registration query used in the candidate selection by the candidate selection unit 115A with the search information created by the search information creation unit 115D, and register them in the recommended job seeker for search database. If the search information has been converted into vector data by the search information creation unit 115D, the registration unit 115E registers the search information converted into vector data in the recommended job seeker for search database. As a result, the recommended job seeker for search database from which the acquisition unit 113 acquires search information is constructed based on the actual search results and the registration information of actual job seekers.

[0092] FIG. 5 is a schematic diagram showing the procedure for constructing a database of recommended job seekers for search. As shown in FIG. 5, the database construction unit 115 acquires, from a search log database DB1 in which search logs are recorded, search queries (registration queries) used in past searches and a group of identification information for candidates corresponding to the search queries. The database construction unit 115 uses this group of identification information to extract a group of characteristic information from a characteristic information database DB2. The characteristic information database DB2 is constructed based on a job seeker database DB3. The database construction unit 115 inputs the extracted group of characteristic information into a search information creation model LM to acquire search information (information identifying recommended job seekers for search). The database construction unit 115 associates each acquired piece of search information with one search query acquired from the search log database DB1 and registers it in a database of recommended job seekers for search DB4.

[0093] At a predetermined timing, the database construction unit 115 updates the characteristic information database DB2 by the attribute information generation unit 115C, and updates the search recommended job seeker database DB4 by the candidate selection unit 115A, the characteristic information acquisition unit 115B, the search information creation unit 115D, and the registration unit 115E. The predetermined timing may be, for example, daily.

[0094] 6 is a flow diagram showing an example of a process for constructing a recommended job seeker database for search, which is executed by the database construction unit 115. The database construction unit 115 first refers to a search log in which past search results are recorded (Step S110). Next, the database construction unit 115 extracts search queries that are not registered in the recommended job seeker database for search (unregistered queries) from the search log (Step S120). Next, the database construction unit 115 acquires a group of identification information of candidates (searched job seekers with a high expected value for the occurrence of an action) for each unregistered query (Step S130).

[0095] After acquiring the identification information group, the database construction unit 115 refers to the characteristic information database (Step S140). Next, the database construction unit 115 acquires a characteristic information group corresponding to the identification information group from the characteristic information database (Step S150). After acquiring the characteristic information group, the database construction unit 115 inputs the characteristic information group to a search information creation model and acquires search information output from the search information creation model (Step S160). Furthermore, the database construction unit 115 converts the search information into vector data as needed. After acquiring the search information, the database construction unit 115 associates the search information with an unregistered query and registers it in a recommended job seeker for search database (Step S170).

[0096] After registering the search information, the database construction unit 115 determines whether or not there is an unregistered query in the search log (step S180). If there is an unregistered query in the search log (S180: YES), the database construction unit 115 repeats the process from obtaining the identification information group in step S130. On the other hand, if there is no unregistered query in the search log (S180: NO), the database construction unit 115 ends the process of constructing the recommended job seeker database for search.

[0097] <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.

[0098] 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.

[0099] 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.

[0100] 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.

[0101] The artificial intelligence unit 120 may be a general-purpose natural language processing learning model, such as a large-scale language model (LLM), trained on a huge amount of data. LLMs are learning models that have previously trained on 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). They can execute various language processing tasks when given tasks, and can perform a wide range of natural language processing tasks, such as grasping sentence patterns and contexts, answering questions, and generating sentences, according to given prompts. Such general-purpose learning models include language models that can handle various tasks without fine-tuning, such as through one-shot learning or few-shot learning. Furthermore, general-purpose learning models may also be configured to handle various tasks through 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.

[0102] The learning models included in the artificial intelligence unit 120 (learning models used in each functional unit, such as the search information creation model) can undergo additional learning as transfer learning or fine tuning. For example, each time new job seeker registration information, job posting registration, etc. is generated, the artificial intelligence unit 120 may perform additional learning and fine tuning using this as new training data. This improves the accuracy of the information output from the learning model.

[0103] 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 reduce the output loss (Soft Target Loss) of the student model (distilled model) relative to the output (Soft Target) of the teacher model, thereby learning the student model, which becomes the distilled model. Alternatively, the student model may be learned to reduce 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 search support system 1.

[0104] For example, the search information creation model etc. may be a distilled model trained using a combination of input data and output data in a large-scale language model as training data. Furthermore, when the search support system 1 is introduced, a large-scale language model may be used as the search information creation model etc., and once training data from the large-scale language model has been accumulated, a distilled model obtained by knowledge distillation using the training data may be used as the search information creation model etc.

[0105] <Display> 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.

[0106] <Operation acquisition part> 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.

[0107] 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.

[0108] This information processing includes a receiving step, an acquiring step, and a searching step. In the receiving step, input of a search query including at least one search keyword is accepted. In the acquiring step, search information corresponding to the search query accepted in the receiving step is acquired from a database of recommended job seekers for search. In the searching step, a job seeker is searched for using the search query and search information accepted in the receiving step as search conditions.

[0109] 7 is an activity diagram showing an example of the flow of information processing (job seeker search processing) executed by the search support system 1. Below, the information processing will be explained along with each activity in this activity diagram.

[0110] The job seeker search process begins with the input of a search query by the user, the recruiter. The recruiter inputs the search query at the recruiter terminal 20 (activity A101). The server device 10 accepts the input of the search query from the recruiter terminal 20 (activity A102). Next, the server device 10 acquires search information corresponding to the search query (activity A103).

[0111] After acquiring the search information, the server device 10 executes a job seeker search based on the input search query and the acquired search information (activity A104). Subsequently, the server device 10 outputs the job seeker search results to the recruiting party terminal 20 (activity A105). As a result, the job seeker search results are displayed on the recruiting party terminal 20 (activity A106).

[0112] 4. Effect The operation of this embodiment can be summarized as follows: When a recruiter inputs a search query to search for a job seeker, information about recommended job seekers for search (search information) who are expected to take action related to the conclusion of a job offer can be used to assist in narrowing down the job seekers. This reduces the cost required for searches by recruiters.

[0113] 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.

[0114] 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.

[0115] The database construction unit 115 may also be configured externally to the server device 10. Furthermore, the search recommended job seeker database may be stored in a server external to the server device 10 and configured to be referenced by each functional unit of the server device 10.

[0116] 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.

[0117] It may be provided in the following manner.

[0118] (1) A search support system comprising at least one processor, the processor being configured to execute the following steps by reading a program: in a receiving step, input of a search query including at least one search keyword is received; in an acquiring step, search information corresponding to the search query received in the receiving step is obtained from a database of recommended job seekers for search; the search information is information specifying attributes of recommended job seekers for search that are associated with a search query including at least one keyword used in a job seeker search and registered in the database of recommended job seekers for search; the recommended job seekers for search are job seekers who are assumed to have a high expected value of taking an action related to the conclusion of a job offer based on past search results based on the search query; and in a searching step, job seekers are searched for using the search query received in the receiving step and the search information as search conditions.

[0119] (2) In the search support system described in (1) above, in the search step, the search query received in the reception step is matched with keywords contained in the registration information of the job seeker registered in the job seeker database, and the job seeker is searched for based on a comparison between the search information and the registration information.

[0120] (3) In the search support system described in (2) above, the search information is vector data, and in the search step, job seekers whose registered information contains keywords that are identical or similar to the search keywords contained in the search query received in the reception step are extracted as search targets, and the display of the job seekers extracted as search targets is adjusted based on a similarity score that indicates the degree of similarity between the search information and the registered information, wherein the similarity score is obtained by comparing the search information with vector data obtained by vectorizing the registered information.

[0121] (4) In the search support system described in (3) above, in the searching step, a display order of the extracted job seekers is determined based on the magnitude of the similarity score.

[0122] (5) In the search support system described in any one of (1) to (4) above, the search recommended job seeker database is constructed by the following steps: in the candidate selection step, from among a plurality of searched job seekers searched by a registration query, which is a search query previously used in searching for job seekers, the searched job seeker who has a predetermined number of occurrences of actions related to job offer conclusions or more, or whose ranking in the number of occurrences among the plurality of searched job seekers is within a predetermined range, is selected as a candidate; in the characteristic information acquisition step, characteristic information representing the attributes of the candidate is acquired from the characteristic information database; in the search information creation step, the characteristic information is input into a search information creation model and the search information creation model is caused to output the search information, wherein the search information creation model is a learning model that has been trained so as to be able to input the characteristic information and output the search information; and in the registration step, the registration query and the search information are linked and registered in the search recommended job seeker database.

[0123] (6) In the search support system described in (5) above, in the search information creation step, the search information creation model is caused to output the search information as text data, and the text data is converted into vector data, and in the registration step, the search information converted into vector data is registered in the search recommended job seeker database.

[0124] (7) In the search support system described in (5) or (6) above, the characteristic information database is a database in which attribute information extracted from a job seeker database in which the registration information of job seekers is registered is linked to the identification information of the job seekers.

[0125] (8) In the search support system described in (7) above, the attribute information includes topics that are generated by inputting the registration information into a topic model and outputting the topic model.

[0126] (9) In the search support system described in (7) or (8) above, the attribute information includes skills or qualifications extracted from the registration information.

[0127] (10) In the search support system described in (9) above, the attribute information includes the degree of experience of the skill, and in the characteristic information acquisition step, the skill to be acquired as the characteristic information is selected based on the magnitude of the degree of experience.

[0128] (11) In the search support system described in any one of (5) to (9) above, the action includes at least one of receiving a scouting document based on any job offer, replying to a scouting document from any job offer, passing the screening for any job offer, and concluding a contract for any job offer.

[0129] (12) A search support method comprising the steps executed by the search support system according to any one of (1) to (11) above.

[0130] (13) A program for causing a computer to execute each step of the search support system described in any one of (1) to (11) above. Of course, this is not the case.

[0131] 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]

[0132] 1: 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: Reception 113: Acquisition Department 114: Search section 115: Database Construction Department 115A: Candidate Selection Division 115B: characteristic information acquisition unit 115C: Attribute information generation section 115D: Search information creation section 115E: Registration Department 120: Artificial Intelligence Department 211:Display section 212: Operation acquisition section 311: Display section 312: Operation reception section

Claims

1. A search support system, at least one processor; The processor is configured to execute the following steps by reading the program: In the receiving step, an input of a search query including at least one search keyword is received, In the acquisition step, search information corresponding to the search query received in the reception step is acquired from a database of recommended job seekers for search, wherein the search information is information specifying attributes of recommended job seekers for search that are associated with a search query containing at least one keyword used in a job seeker search and registered in the database of recommended job seekers for search, and the recommended job seekers for search are job seekers who are assumed to have a high expected value for the occurrence of an action related to the conclusion of a job offer based on past search results based on the search query, In the search step, the search support system searches for job seekers using the search query received in the reception step and the search information as search conditions.

2. 2. The search support system according to claim 1, In the search step, the search support system searches for job seekers based on matching the search query received in the reception step with keywords contained in the registration information of job seekers registered in a job seeker database, and comparing the search information with the registration information.

3. 3. The search support system according to claim 2, the search information is vector data, In the search step, a search support system extracts job seekers whose registered information contains keywords that are identical or similar to the search keywords included in the search query received in the reception step, and adjusts the display of the job seekers extracted as search targets based on a similarity score that indicates the degree of similarity between the search information and the registered information, wherein the similarity score is obtained by comparing the search information with vector data obtained by vectorizing the registered information.

4. 4. The search support system according to claim 3, In the search step, the search support system determines the display order of the extracted job seekers based on the magnitude of the similarity score.

5. 2. The search support system according to claim 1, The search recommended job seeker database is constructed by the following steps: In the candidate selection step, from among a plurality of searched job seekers searched by a registration query, which is a search query previously used in searching for job seekers, a searched job seeker who has a predetermined number of occurrences of an action related to the conclusion of a job offer or whose ranking in the number of occurrences among the plurality of searched job seekers is within a predetermined range is selected as a candidate; In the characteristic information acquisition step, characteristic information representing attributes of the candidate is acquired from a characteristic information database; In the search information creation step, the feature information is input to a search information creation model, and the search information creation model is caused to output the search information, wherein the search information creation model is a learning model that has been trained so as to be able to input the feature information and output the search information, In the registration step, the search support system links the registration query with the search information and registers them in the search recommended job seeker database.

6. 6. The search support system according to claim 5, In the search information creation step, the search information creation model is caused to output the search information as text data, and the text data is converted into vector data; In the registration step, the search information converted into vector data is registered in the search recommended job seeker database.

7. 6. The search support system according to claim 5, A search support system, wherein the characteristic information database is a database in which attribute information extracted from a job seeker database in which registration information of job seekers is registered is linked to the identification information of the job seekers.

8. 8. The search support system according to claim 7, A search support system, wherein the attribute information includes topics that are generated by inputting the registration information into a topic model and outputting the topic model.

9. 8. The search support system according to claim 7, A search support system, wherein the attribute information includes skills or qualifications extracted from the registration information.

10. 10. The search support system according to claim 9, The attribute information includes an experience level of the skill, In the characteristic information acquisition step, the skill to be acquired as the characteristic information is selected based on the level of the experience level.

11. 6. The search support system according to claim 5, In the search support system, the actions include at least one of receiving a scout document based on the given job offer, replying to the scout document from the given job offer, passing the screening for the given job offer, and closing the deal for the given job offer.

12. A search support method, comprising: A search support method comprising the steps executed by the search support system according to any one of claims 1 to 11.

13. A program, A program for causing a computer to execute each step of the search support system according to any one of claims 1 to 11.

Images