Processing apparatus, processing program, processing method, and processing system

Abstract

To provide a processing device, processing program, processing method, and processing system that enable more appropriate recommendations of candidates. [Solution] The system receives subject information from a terminal device, which includes at least selection information indicating the subject's selection for each of the multiple categories indicating the respondent's characteristics or the multiple categories indicating the respondent's desired content. It also receives and stores multiple request information for identifying the content desired by each of the one or more recommended persons from each recommended person's terminal device. Based on the multiple request information, it generates a first index indicating the degree of recommendation for each of the multiple categories indicating the respondent's characteristics or the multiple categories indicating the respondent's desired content. Based on the first index and the subject information, it generates recommendation information indicating the recommended request information from the multiple request information.

Description

Technical Field

[0001] The present disclosure relates to a processing apparatus, a processing program, a processing method, and a processing system capable of processing for recommending a target person to a subject person.

Background Art

[0002] Conventionally, a processing system for recommending a subject person to a target person has been known. For example, Patent Document 1 describes a processing system using "an acquisition unit that acquires information on at least one of a job seeker and a job offerer, an extraction unit that extracts at least one entity of the job seeker and the job offerer from the information on at least one of the job seeker and the job offerer, and a classification unit that classifies at least one entity of the job seeker and the job offerer extracted by the extraction unit into a predetermined category."

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

[0004] However, in the technology of Patent Document 1, feature amounts are extracted from each of the job seeker and the job offerer, and matching is performed based on the degree of coincidence of the feature amounts. Evaluation is not performed using the same evaluation axis, and it is not possible to introduce an appropriate job offerer (or job seeker) to a job seeker (or job offerer).

Summary of the Invention

Problems to be Solved by the Invention

[0005] Therefore, based on the above-described technology, an object of the present disclosure is to provide a processing apparatus, a processing program, a processing method, and a processing system capable of more appropriately recommending a recommended person according to various embodiments.

Means for Solving the Problems

[0006] According to one aspect of the present disclosure, a processing device is provided comprising at least one processor, wherein the at least one processor receives subject information from a terminal device, which includes at least selection information indicating the subject's selection for each of a plurality of categories indicating the subject's characteristics or a plurality of categories indicating the subject's desires; stores a plurality of request information for identifying the content desired by each of one or more recommended persons by receiving each recommended person terminal device; generates a first index indicating the degree of recommendation for each of the plurality of categories indicating the subject's characteristics or a plurality of categories indicating the subject's desires based on each of the plurality of request information; and executes a process to generate recommendation information indicating the recommended request information from the plurality of request information based on the first index and the subject information.

[0007] According to one aspect of the present disclosure, a processing program is provided which, by being executed by at least one processor in a computer, receives subject information from a subject terminal device, which includes at least selection information indicating the respondent's selection for each of a plurality of categories indicating the respondent's characteristics or a plurality of categories indicating the respondent's desires; stores a plurality of request pieces of information for identifying the content desired by each of one or more recommended persons by receiving them from each recommended person terminal device; generates a first index indicating the degree of recommendation in each of the plurality of categories indicating the respondent's characteristics or a plurality of categories indicating the respondent's desires based on each of the plurality of request pieces of information; and causes the at least one processor to function to generate recommendation information indicating the recommended request piece of information based on the first index and the subject information.

[0008] According to one aspect of the present disclosure, a processing method is provided which is performed by at least one processor in a computer, and includes the steps of: receiving subject information from a subject terminal device, which includes at least selection information indicating the respondent's selection for each of a plurality of categories indicating the respondent's characteristics or a plurality of categories indicating the respondent's desires; storing a plurality of request information for identifying the content desired by each of one or more recommended persons by receiving each of the recommended person terminal devices; and for each of the plurality of request information, generating a first index indicating the degree of recommendation in each of the plurality of categories indicating the respondent's characteristics or a plurality of categories indicating the respondent's desires, based on each request information.

[0009] According to one aspect of this disclosure, the system includes "the processing device described above, a target terminal device configured for use by the target and connected to the processing device in a communicative manner, and a recommended terminal device configured for use by the recommended person and connected to the processing device in a communicative manner." [Effects of the Invention]

[0010] This disclosure provides a processing device, processing program, processing method, and processing system that enable more appropriate recommendations of nominees.

[0011] The effects described above are merely illustrative for the sake of explanation and are not limiting. In addition to, or in lieu of, any other effects described herein or that would be obvious to those skilled in the art may be achieved. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a block diagram showing the configuration of a processing system 1 according to one embodiment of the present disclosure. [Figure 2A] Figure 2A is a block diagram showing the configuration of a processing apparatus 100 according to one embodiment of the present disclosure. [Figure 2B]Figure 2B is a block diagram showing the configuration of a terminal device 200 according to one embodiment of the present disclosure. [Figure 3A] Figure 3A is a conceptual diagram showing a job posting management table stored in a processing device 100 according to one embodiment of the present disclosure. [Figure 3B] Figure 3B is a conceptual diagram showing a category management table stored in a processing device 100 according to one embodiment of the present disclosure. [Figure 3C] Figure 3C is a conceptual diagram showing a job seeker management table stored in a processing device 100 according to one embodiment of the present disclosure. [Figure 4] Figure 4 shows a processing sequence executed by a processing system 1 according to one embodiment of the present disclosure. [Figure 5A] Figure 5A is a diagram showing the processing flow performed in the processing apparatus 100 according to one embodiment of the present disclosure. [Figure 5B] Figure 5B is a diagram showing the processing flow performed in the processing apparatus 100 according to one embodiment of the present disclosure. [Figure 6] Figure 6 is a conceptual diagram showing a specific example of a category management table according to one embodiment of this disclosure. [Figure 7A] Figure 7A shows an example of a job application form input screen 30 output in a job applicant terminal device 200-2 according to one embodiment of the present disclosure. [Figure 7B] Figure 7B shows an example of a job seeker information input screen 10 output in a job seeker terminal device 200-1 according to one embodiment of the present disclosure. [Figure 7C] Figure 7C shows an example of a recommendation information screen 20 output in a job seeker terminal device 200-1 according to one embodiment of the present disclosure. [Modes for carrying out the invention]

[0013] Hereinafter, various embodiments of the present disclosure will be described with reference to the accompanying drawings. Note that the same reference numerals are assigned to common components in the drawings. Also, note that components shown in one drawing may be omitted in another drawing for convenience of explanation. Furthermore, note that the attached drawings are not necessarily drawn to an exact scale.

[0014] The various systems, methods, and apparatuses described in the present disclosure should not be construed as being limited in any way. In fact, the present disclosure is directed to any novel features and aspects among each of the disclosed various embodiments, combinations of these various embodiments with each other, and combinations of some of these various embodiments with each other. The various systems, methods, and apparatuses described in the present disclosure are not limited to specific aspects, specific features, or combinations of such specific aspects and specific features, and the things and methods described in the present disclosure do not require the existence of one or more specific effects or the solution of problems. Furthermore, various features or aspects among the various embodiments described in the present disclosure, or some of such features or aspects, can be used in combination with each other.

[0015] Although the operations of some of the various methods disclosed in the present disclosure are described in a specific order for convenience, such a description should be understood to include rearranging the order of the above operations, unless the specific order is required by the following specific text. For example, a plurality of operations described in sequence may, in some cases, be rearranged or executed simultaneously. Furthermore, for the purpose of simplification, the attached drawings do not show various ways in which the various matters and methods described in the present disclosure can be used together with other matters and methods.

[0016] Any operational theories, scientific principles, or other theoretical descriptions presented in this disclosure in relation to the devices or methods of this disclosure are provided for better understanding and are not intended to limit the technical scope. The devices and methods in the appended claims are not limited to devices and methods that operate by the methods described by such operational theories.

[0017] Any of the various methods disclosed in this disclosure are implemented using a plurality of computer-executable instructions stored on one or more computer-readable media and can further be executed on a computer. The above-mentioned one or more media can be non-transitory computer-readable storage media such as, for example, at least one optical disk, a plurality of volatile memory components, or a plurality of non-volatile memory components. Here, the above-mentioned plurality of volatile memory components include, for example, DRAM or SRAM. Also, the above-mentioned plurality of non-volatile memory components include, for example, hard drives and solid state drives (SSDs). Further, the above-mentioned computer includes any computer available on the market, including, for example, smartphones and other mobile devices having hardware for performing calculations.

[0018] Any of the multiple computer-executable instructions for implementing the technology disclosed in this disclosure, along with any data generated and used during implementations of the various embodiments disclosed in this disclosure, may be stored in one or more computer-readable media (e.g., non-temporary computer-readable storage media). Such multiple computer-executable instructions may, for example, be part of a separate software application, or part of a software application accessed or downloaded via a web browser or other software application (such as a remote computing application). Such software may be executed, for example, on a single local computer (as a process run on any suitable computer available on the market), or in a network environment (e.g., the Internet, a wide area network, a local area network, a client-server network (such as a cloud computing network), or other such network) using one or more network computers.

[0019] For clarity, only specific selected aspects of various software-based implementations are described. Other details that are well known in the art are omitted. For example, the technology disclosed in this disclosure is not limited to any particular computer language or program. For example, the technology disclosed in this disclosure may be executed by software written in C, C++, Java®, or any other suitable programming language. Similarly, the technology disclosed in this disclosure is not limited to any particular computer or type of hardware. Specific details of suitable computers and hardware are well known and do not need to be described in detail in this disclosure.

[0020] Furthermore, any of the various embodiments of such software (including, for example, a set of computer-executable instructions for causing a computer to perform any of the various methods disclosed herein) may be uploaded, downloaded, or accessed remotely by preferred means of communication. Such preferred means of communication include, for example, the Internet, the World Wide Web, intranets, software applications, cables (including fiber optic cables), magnetic communications, electromagnetic communications (including RF communications, microwave communications, and infrared communications), electronic communications, or other such means of communication.

[0021] 1. Overview of Processing System 1 The processing system 1 relating to this disclosure is used to match a target person with a recommended person. Typically, such a processing system 1 is used to match a job seeker (target person) who is looking for a job with an employer (recommended person) who is posting a job opening. In this case, the processing system 1 receives target person information from a target person terminal device usable by the target person (e.g., job seeker), which includes at least selection information indicating the target person's (e.g., job seeker's) selection for each of several categories indicating the target person's (e.g., job seeker's) characteristics or preferences. The processing system 1 also receives and stores multiple request information (e.g., job information) from each recommended person terminal device usable by the recommended person (e.g., employer) to identify what each of the recommended person (e.g., employer) is looking for (e.g., job content). The processing system 1 also generates a first index indicating the degree of recommendation for each of the several categories for the multiple request information. The processing system 1 then generates recommendation information indicating recommended request information (e.g., job postings) based on the first indicator in each of the multiple categories and the target person information. It should be noted that matching only requires determining who to recommend to the target person (e.g., recommending a job-seeking person to a recruiting company), and it is not mandatory for the target person to accept the recommended person (e.g., a job-seeking person applying to a recruiting company).

[0022] In this way, the processing system 1 allows the target person (e.g., a job seeker) to select their own characteristics and preferences from a range of categories. The processing system 1 also generates a first indicator of recommendation degree for the requested information (e.g., job posting) of the person being recommended (e.g., an employer), using the same categories selected by the target person. In other words, the processing system 1 uses the same categories (i.e., the same evaluation axis) as the categories selected by the target person when generating the recommendation degree for the person being recommended. Therefore, it becomes possible to recommend a person being recommended (or the requested information of the person being recommended) that is a better match for the target person's selection. The recommendation degree is the degree of compatibility between the requested information and the given categories, for example, the degree of compatibility between the job posting and the given categories.

[0023] Such a processing system 1 is preferably used when the target is a job seeker, the person being recommended is an employer (for example, a business that posts job information), the requested information is job information, and the system recommends job information posted by employers to job seekers. However, the processing system 1 is not limited to this case and can also be used in the following cases, for example. • When the target is an employer (e.g., a business that is hiring), the person being recommended is a job seeker, the requested information is the job offer information of the job seeker, and the job offer information submitted by the job seeker is recommended to the employer (e.g., when recommending a specific job seeker to a business that is hiring based on their resume and work history).

[0024] Furthermore, processing system 1 can be used not only for introducing personnel as described above, but also in the following cases. • When recommending service providers (recommended individuals) to individuals (target individuals) who wish to receive various goods or services. • When recommending a business operator (recommended party) who wishes to transfer real estate to a prospective real estate buyer (target party). • When recommending another suitor (the person being recommended) to a suitor (the person being recommended) who desires marriage.

[0025] For the sake of explanation, the following description assumes that Processing System 1 is used when the target is a job seeker, the recommended party is an employer, the requested information is job information, and the system recommends job information issued by an employer to a job seeker. However, this disclosure is not limited to this example alone.

[0026] In this disclosure, the "Subject" can be any person who receives a recommendation from the Recommended Person (or the Information Requested by the Recommended Person). Examples of such Subjects include job seekers, individuals seeking to provide various goods or services, individuals seeking to purchase real estate, and individuals seeking marriage. For the sake of explanation, the following will describe the Subject as a job seeker, but it is not limited to this.

[0027] In this disclosure, the term "recommended person" can refer to any person recommended by the subject. Examples of such recommended persons include employers (e.g., businesses) posting job openings, providers of various goods and services, businesses wishing to transfer real estate, and suitors seeking marriage. For the sake of clarity, the following explanation will focus on the case where the recommended person is an employer, but this is by no means the only example.

[0028] Furthermore, while this disclosure provides examples of the subject and the person being recommended as described above, the roles of subject and person being recommended may be reversed in some cases. For example, as described above, the subject may be the employer (e.g., a business that is hiring) and the person being recommended may be the job seeker. In other words, the terms subject and person being subject are merely designations used to distinguish them, and the roles of the person receiving the recommendation and the person being recommended may be reversed as appropriate.

[0029] Furthermore, while this disclosure uses terms such as "target person," "recommended person," "job seeker," and "employer," as described above, these are merely terms used to distinguish them from one another based on the roles and situations in which they are involved in processing system 1. Therefore, each of the terms "target person," "recommended person," "job seeker," and "employer" may refer to any individual, or to any organization to which each of them belongs (for example, the business to which the recommended person belongs).

[0030] Furthermore, while the terms "First," "Second," etc. may be used in this disclosure, unless otherwise specified, these are merely designations used for explanatory purposes. Therefore, the presence of terms such as "First," "Second," etc., does not mean that the disclosure is limited to only those elements to which these terms are attached. Naturally, it may also include "Third," "Fourth," "Fifth," and more elements.

[0031] 2. Configuration of Processing System 1 (A) Processing system 1 Figure 1 is a block diagram showing the configuration of a processing system 1 according to one embodiment of the present disclosure. According to Figure 1, the processing system 1 includes a processing device 100, a job seeker terminal device 200-1, and a job seeker terminal device 200-2, and each device is connected to communicate via a wired or wireless network.

[0032] In this disclosure, the processing unit 100 can be any device capable of performing the processing that the processing unit 100 executes. In other words, various devices such as on-premise server devices, cloud server devices, smartphones, tablet devices, laptop PCs, and desktop PCs can be used as processing units.

[0033] Furthermore, either of the terminal devices 200, such as the job seeker terminal device 200-1 or the employer terminal device 200-2, can also function as a processing unit. In addition, in this disclosure, the storage and processing performed by the processing unit 100 may be distributed to other terminal devices or other server devices. In other words, the processing unit 100 is not limited to being composed of a single enclosure, and the processing unit 100 also includes various combinations of devices as exemplified above.

[0034] Furthermore, in this disclosure, the job seeker terminal device 200-1 and the employer terminal device 200-2 are distinguished based on whether they are usable by the job seeker (the target) or the employer (the person being recommended). The job seeker terminal device 200-1 can function as a target terminal device usable by the target, and the employer terminal device 200-2 can function as a recommended terminal device usable by the person being recommended. However, this does not mean that the job seeker terminal device 200-1 is only usable by job seekers (and target), and the employer terminal device 200-2 is only usable by employers (and people being recommended). In other words, each terminal device may be used by job seekers (and target), employers (and people being recommended), or other persons, and its name may change depending on who is using it.

[0035] Furthermore, while the job seeker terminal device 200-1 and the employer terminal device 200-2 may be simply referred to as terminal device 200, this means at least one of the terminal devices, including job seeker terminal device 200-1 and employer terminal device 200-2. In addition, although two terminal devices 200, job seeker terminal device 200-1 and employer terminal device 200-2, are given as examples in this disclosure, other terminal devices 200 may also be included.

[0036] (B) Processing apparatus 100 Figure 2A is a block diagram showing the configuration of a processing unit 100 according to one embodiment of the present disclosure. According to Figure 2A, the processing unit 100 includes a processor 111, a memory 112, and a communication interface 113. Each of these components is electrically connected to the others via control lines and data lines. The processing unit 100 does not need to have all of the components shown in Figure 2A; it is possible to omit some components or add other components. For example, it is possible to use an external memory connected via communication as memory, or a database device or server device. Furthermore, it is possible to distribute and execute some processing, such as processing using large language models (LLMs), with other processing units including server devices. In other words, the processing unit 100 is not limited to a single device, but also includes cases where it is distributed across multiple devices depending on the handling of information and the processing load. A large language model is a machine learning model for natural language processing that has learned from a large amount of training data, and is a model that has the function of generating a response including a sentence based on input information. For example, a large language model is a machine learning model based on a Transformer model with a self-attention mechanism. The Transformer model is, for example, a GPT (Generative Pre-trained Transformer) model. In this embodiment, an example of text input to the model is described, but the input may also include image information or audio information.

[0037] The processor 111 functions as a control unit that controls other components of the processing system 1 based on a processing program stored in memory 112. The processor 111 is mainly composed of one or more CPUs, but may be combined with a GPU, FPGA, etc. as appropriate. Based on the processing program stored in memory 112, the processor 111 executes processing to generate recommendation information for recommending job information to job seekers. Specifically, the processor 111 executes the following processes based on a processing program stored in memory 112: "receiving job seeker information (i.e., target information) from a job seeker terminal device (i.e., target terminal device), which includes at least selection information indicating the job seeker's (i.e., target person's) selection for each of multiple categories indicating the characteristics of the respondent or multiple categories indicating the respondent's desired content"; "receiving and storing multiple job information (i.e., request information) for identifying what each of one or more employers (i.e., recommended persons) is looking for, from each employer terminal device (i.e., recommended person terminal device)"; "for each of the multiple job information (i.e., request information), generating a first index indicating the degree of recommendation in each of multiple categories indicating the characteristics of the respondent or multiple categories indicating the respondent's desired content, based on each job information (i.e., request information)"; and "generating recommendation information indicating the recommended job information (i.e., request information) among the job information (i.e., request information), based on the first index and the job seeker information (i.e., target information)."

[0038] Memory 112 consists of RAM, ROM, non-volatile memory, HDD, SSD, etc., and functions as a storage unit. Memory 112 stores instruction commands for various controls of the processing system 1 according to this embodiment as processing programs. Specifically, memory 112 stores a program for processor 111 to execute processes such as: "receiving job seeker information (i.e., target information) from a job seeker terminal device (i.e., target terminal device) which includes at least selection information indicating the job seeker's (i.e., target person's) selection for each of multiple categories indicating the characteristics of the respondent or multiple categories indicating the respondent's desired content"; "receiving and storing multiple job information (i.e., request information) for identifying what each of one or more employers (i.e., recommended persons) is looking for, from each employer terminal device (i.e., recommended person terminal device"; "for each of the multiple job information (i.e., request information), generating a first index indicating the degree of recommendation in each of multiple categories indicating the characteristics of the respondent or multiple categories indicating the respondent's desired content, based on each job information (i.e., request information)"; and "recommendation information indicating recommended job information (i.e., request information) from among the job information (i.e., request information), based on the first index and job seeker information (i.e., target information)". In addition to the program, memory 112 also stores various information stored in the job posting management table, category management table, and job seeker management table. Note that this information does not necessarily need to be constantly stored in memory 112 within the processing unit 100; it may be stored in a remotely located database device. In that case, the database device is also included in memory 112.

[0039] The communication interface 113 functions as a notification unit for sending and receiving various information between the job seeker terminal device 200-1, the employer terminal device 200-2, and other processing devices connected via a wired or wireless network. Examples of the communication interface 113 include wired communication connectors such as USB and SCSI, wireless communication transceivers such as wireless LAN, Bluetooth®, LTE, and infrared, and various connection terminals for printed circuit boards and flexible circuit boards.

[0040] (C) Terminal device 200 (Job seeker terminal device 200-1 and employer terminal device 200-2) Figure 2B is a block diagram showing the configuration of a terminal device 200 according to one embodiment of the present disclosure. Specifically, Figure 2B is a block diagram showing the configuration of a device that can be used as a job seeker terminal device 200-1, which is an example of a target terminal device, or as a recruiter terminal device 200-2, which is an example of a recommended person terminal device.

[0041] The terminal device 200 includes a processor 211, memory 212, input interface 213, output interface 214, and communication interface 215. Each of these components is electrically connected to the others via control lines and data lines. Note that the terminal device 200 does not need to have all of the components shown in Figure 2B; it is possible to omit some components or add other components.

[0042] The terminal device 200 can be any device capable of performing the processing that the terminal device 200 performs. The terminal device 200 can be a variety of devices, such as a smartphone, tablet, laptop PC, or desktop PC. Furthermore, in this disclosure, the processing performed by the terminal device 200 may be distributed to other terminal devices, processing units, servers, databases, etc. In other words, the terminal device 200 is not limited to a single enclosure, but also includes various combinations of devices as exemplified above. Additionally, the terminal device 200 can be used as a job seeker terminal device 200-1 and a job seeker terminal device 200-2, but when used as job seeker terminal device 200-1 and job seeker terminal device 200-2, they may be different types of terminal devices. For example, job seeker terminal device 200-1 may be a desktop PC, and job seeker terminal device 200-2 may be a smartphone.

[0043] The processor 211 functions as a control unit that controls other components of the terminal device 200 based on a program stored in the memory 212. The processor 211 is mainly composed of one or more CPUs, but may be combined with a GPU, FPGA, etc. as appropriate.

[0044] When the processor 211 functions as a job seeker terminal device 200-1, it performs processes such as selecting job seeker information, which is an example of target information, based on a processing program stored in memory 112. Specifically, the processor 211 executes the following processes based on a program stored in memory 212: "a process to generate job seeker information (i.e., target information) that includes at least selection information indicating the job seeker's (i.e., target person's) selection for each of a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the respondent's desired content"; "a process to transmit the generated job seeker information (i.e., target person's information) to the processing unit 100 via the communication interface 215"; "a process to receive from the processing unit 100 via the communication interface 215 recommendation information indicating the degree of recommendation in each of a plurality of categories generated according to each of a plurality of job information (i.e., request information) for identifying what each of one or more employers (i.e., recommended persons) is looking for, and recommendation information indicating recommended job information (i.e., request information) generated based on the job seeker information (i.e., target person's information)"; and "a process to output the received recommendation information via the communication interface 215".

[0045] Furthermore, when the processor 211 functions as a job seeker terminal device 200-2, it performs processes such as inputting job information, which is an example of request information, based on the processing program stored in the memory 112. Specifically, the processor 211 executes the following processes based on a program stored in memory 212: "a process to receive operational input from the employer (i.e., the person being recommended) via the input interface 213 and input job information (i.e., request information) to identify what the employer (i.e., the person being recommended) is looking for"; "a process to transmit the input job information (i.e., request information) to the processing unit 100 via the communication interface 215"; "a process to receive a notification from the processing unit 100 via the communication interface 215 indicating that the job information (i.e., request information) has been sent to the job seeker (i.e., the person being recommended) as recommendation information, based on a program stored in memory 212: "a process to receive operational input from the employer (i.e., the person being recommended) via the input interface 213 and input job information (i.e., request information) to identify what the employer (i.e., the person being recommended) is looking for"; "a process to transmit the input job information (i.e., request information) to the processing unit 100 via the communication interface 215"; and "a process to output the received notification via the output interface 214".

[0046] Memory 212 consists of RAM, ROM, non-volatile memory, HDD, etc., and functions as a storage unit. Memory 212 stores instruction commands for various controls of the processing system 1 according to this embodiment as programs.

[0047] Specifically, when functioning as a job seeker terminal device 200-1, memory 212 stores programs for processor 211 to execute, including: "a process to generate job seeker information (i.e., target information) that includes at least selection information indicating the job seeker's (i.e., target person's) selection for each of a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the respondent's desired content"; "a process to transmit the generated job seeker information (i.e., target person's information) to the processing unit 100 via the communication interface 215"; "a process to receive from the processing unit 100 via the communication interface 215 recommendation information indicating the degree of recommendation in each of a plurality of categories generated according to each of a plurality of job information (i.e., request information) for identifying what each of one or more employers (i.e., recommended persons) is looking for, and recommendation information indicating recommended job information (i.e., request information) generated based on the job seeker information (i.e., target person's information)"; and "a process to output the received recommendation information via the communication interface 215".

[0048] Furthermore, when functioning as a recruiter terminal device 200-2, memory 212 stores programs for processor 211 to execute, such as: "a process of receiving operational input from a recruiter (i.e., a person being recommended) via the input interface 213 and inputting job information (i.e., request information) to identify what the recruiter (i.e., a person being recommended) is looking for"; "a process of transmitting the input job information (i.e., request information) to the processing unit 100 via the communication interface 215"; "a process of receiving a notification from the processing unit 100 indicating that the job information (i.e., request information) has been sent to the job seeker (i.e., a person being recommended) as recommendation information, based on a first index indicating the degree of recommendation in each of a plurality of categories generated in accordance with the job information (i.e., request information), and job seeker information (i.e., person being information) which includes at least selection information indicating the job seeker's (i.e., target person) selection for each of a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the desired content of the respondent"; and "a process of outputting the received notification via the output interface 214".

[0049] The input interface 213 functions as an input unit that receives user or expert input to the terminal device 200. Examples of input interfaces 213 include physical key buttons and a touch panel having an input coordinate system corresponding to the display coordinate system of the display. In the case of a touch panel, icons are displayed on the display, and the operator makes a selection for each icon by making an input via the touch panel. The method for detecting the user's input via the touch panel can be any method, such as capacitive or resistive. The input interface 213-2 does not always need to be physically provided on the terminal device 200 and may be connected as needed via a wired or wireless network. Therefore, in addition to the above, a mouse or keyboard can also be used as the input interface 213-2.

[0050] The output interface 214 functions as an output unit for outputting various types of information. An example of the output interface 214 is a display, but it is not limited to this and may consist of other liquid crystal panels, organic EL displays, plasma displays, or printers. Furthermore, it is not necessary for a display to be provided; for example, an interface for connecting to a display that can be connected to the processing unit 100 via a wired or wireless network can also function as the output interface 214 for outputting display data to the display.

[0051] The communication interface 215 functions as a communication unit for sending and receiving information with the processing unit 100, other terminal devices 200, and other processing units. Examples of the communication interface 215 include wired communication connectors such as USB and SCSI, wireless communication transceivers such as wireless LAN, Bluetooth®, LTE, and infrared, and various connection terminals for printed circuit boards and flexible circuit boards.

[0052] 3. Various information used in processing in processing system 1 Figures 3A to 3C show various tables containing information stored in the processing unit 100 and transmitted to and received from each terminal device 200 as processing progresses. This information is updated and stored as processing progresses. The information shown in Figures 3A to 3C may be stored in the memory 112 of the processing unit 100, or it may be stored in another database device installed remotely and read out as processing progresses.

[0053] (A) Job posting management table Figure 3A is a conceptual diagram showing a job posting management table stored in a processing device 100 according to one embodiment of the present disclosure. According to Figure 3A, the job posting management table stores content information, employer information, conditions information, and remarks information, associated with job posting ID information. Each of these pieces of information is used as job posting information, which is an example of request information indicating what the employer is looking for.

[0054] "Job posting ID information" is unique information for each job posting and is used to identify each job posting. Job posting ID information is generated each time a new job posting is received from the recruiter terminal device 200-2. As mentioned above, the job posting ID information can be anything as long as it identifies the job posting, and various things can be used, such as the name of the recruiter who created the job posting, the recruiter ID information that identifies the recruiter, or the content information of the job posting.

[0055] A job posting can be any form of information used to communicate job openings to job seekers. Typically, a job posting includes information describing the job, information about the employer, and information about the job requirements. However, a job posting can include any information, and may include information other than those exemplified above. Furthermore, while it is preferable for the job posting to follow the job posting form provided by the administrator managing and operating the processing unit 100, it may, of course, follow any form.

[0056] "Content information" refers to information that describes the content of the job posting. Content information is entered by receiving operational input from the job seeker via the input interface 213 of the job seeker terminal device 200-2. The content information includes, but of course, other types of information are also acceptable. • The nature of the work being advertised (for example, "Projects will be managed by teams of 3-5 people. Work will be carried out in accordance with legal procedures.") • Job title (e.g., "engineer") • Expected annual income (for example, "6 million yen") • Number of holidays (for example, "two days off per week, 5 days of summer vacation and 5 days of year-end / New Year's holiday") • Average overtime hours (e.g., "20 hours") • Work location (e.g., "Head Office")

[0057] "Job seeker information" refers to information about job seekers who are currently hiring. This job seeker information is entered by receiving operational input from the job seeker via the input interface 213 of the job seeker terminal device 200-2. While the following are examples of job seeker information, other types of information may also be included. • Company name (for example, "XXX Company, Ltd.") • Industry (for example, "construction industry") • Number of employees (e.g., "250 people") • Head office location information (e.g., head office address) • Location information of business locations other than the head office (for example, the addresses of business locations other than the head office) • Name of affiliated company (e.g., "XXX Engineering Co., Ltd.") • Location information of affiliated companies (e.g., the address of the affiliated company) • Average length of service (e.g., "15 years") • Employee benefits ·Founding year

[0058] "Condition information" refers to information indicating the requirements for a job posting. Condition information is entered by receiving operational input from the job seeker via the input interface 213 of the job seeker terminal device 200-2. The condition information includes the examples shown below, but other information may also be included. Furthermore, the information shown below does not specifically distinguish between required and desired conditions, but some may be considered required and the rest desired. Is it possible even without prior experience? • Number of companies worked for (e.g., "changed jobs 3 or less") • Language level (e.g., "Business-level Chinese") • Highest level of education (e.g., "University degree in science or engineering or higher")

[0059] "Remarks information" is information provided to employers as reference when posting a job. Remarks information is entered by receiving operational input from the employer via the input interface 213 of the employer terminal device 200-2. Remarks information includes the following as examples, but of course, other information may also be included. • The appeal of the job (for example, "It's work that leaves a mark on the map.") • Training programs (for example, "We will provide training for employees who join the company without prior experience.") • Messages from employees (e.g., "I look forward to working with you as part of the team.") Examples of a typical day for an employee

[0060] (B) Category management table Figure 3B is a conceptual diagram showing a category management table stored in a processing device 100 according to one embodiment of the present disclosure. According to Figure 3B, the category management table is a table provided for each job posting ID information. As an example, Figure 3B shows a category management table stored for a job posting with job posting ID information "A1". The category management table stores category information, first indicator information, reason information, etc., for each job posting ID information. Each of these pieces of information is used, for example, to generate recommendation information.

[0061] "Category information" is information that identifies each category. Examples of category information include the category ID information and category name assigned to each category. Categories are information that is set in advance by the administrator who manages and operates the processing device 100. Categories include at least characteristic categories that show the characteristics of the job seeker who is the respondent, and desire categories that show what the job seeker desires from the job posting. Job seeker characteristics include, for example, the job seeker's personality tendencies and preferences. Desired items for job postings include, for example, the items that the respondent tends to value when changing jobs. Among such categories, specific categories include, for example, those exemplified below, but of course, other information may also be included. • Work steadily and diligently • Think logically • Can play as a team • Make a plan • Make friends with people • Explain to someone I am good at writing. I like challenges • High language proficiency • Qualifications held • Special skills and strengths • Areas of weakness or shortcomings • Experience as a captain or leader

[0062] Furthermore, while the following are examples of desired categories, other types of information are also acceptable. I want a raise. I want to cherish my private life. I want to acquire a variety of skills. I want to work at the same company for a long time. I want to work for a stable company. I want to work overseas in the future. I would prefer not to be transferred to another location.

[0063] "First Indicator Information" is information indicating the degree of recommendation for each category for a job posting identified by the job posting ID information. The first indicator is information generated by the processing device 100 when category determination processing is performed. Examples of the first indicator information include a numerical value indicating the degree of recommendation, a classification or category assigned based on that numerical value, or a combination thereof.

[0064] "Reason information" is information indicating the reason why the first indicator information was assigned to each category of the job posting identified by the job posting ID information. Reason information is generated when the processing device 100 performs category determination processing, etc. An example of reason information is the information shown in Figure 6.

[0065] (C) Job seeker management table Figure 3C is a conceptual diagram showing a job seeker management table stored in a processing device 100 according to one embodiment of the present disclosure. According to Figure 3C, the job seeker management table stores attribute information, selection information, recommended job information, and recommendation letter information, etc., associated with job seeker ID information.

[0066] "Job seeker ID information" is unique information for each job seeker and is used to identify each individual. Job seeker ID information is generated each time new job seeker information is received from the job seeker terminal device 200-1. As mentioned above, the job seeker ID information can be anything that identifies the job seeker, and various things can be used, such as the job seeker's name, SNS account, email address, or phone number.

[0067] "Attribute information" is information that indicates the attributes of a job seeker and can be used as job seeker information, which is an example of target information. Attribute information is input by receiving operation input from the job seeker via the input interface 213 of the job seeker terminal device 200-1. Attribute information includes the following as examples, but of course, other information may also be included. • Name of job seeker • Contact information of the job seeker (e.g., email address, phone number, social media accounts, etc.) • Job seeker's location information (e.g., job seeker's current address) • Location information of the job seeker (for example, preferred work location) • Desired job type • When you are available to start work (for example, "from October 1, 2024") • Current employment status (e.g., "Scheduled to resign on July 31, 2024") • Highest level of education • Qualifications held • The time you would like to receive recommendation information (for example, "around 9 AM every morning")

[0068] "Selection information" is information that shows the results of the selection made by the job seeker for each of several pre-set categories, and is information that can be used as job seeker information, which is an example of target information. The categories are, for example, information that is set in advance by the administrator who manages and operates the processing device 100. The categories include at least characteristic categories that show the characteristics of the job seeker and desired categories that show what the job seeker desires in a job. Examples of characteristic categories and desired categories are as explained in Figure 3B. As an example of selection information, the selection information may include information that shows whether or not each of the categories exemplified in Figure 3B was selected (i.e., if selected, it shows that the job seeker possesses or desires that characteristic) and information that shows the degree of each category (for example, for the category "works steadily," it may show that one of the following was selected: "good," "somewhat good," "average," "somewhat not good," and "not good").

[0069] "Recommended job information" is information that indicates recommended job information and is information that can be used as recommendation information. Recommended job information is information generated in the processing device 100 by performing matching processing etc. based on at least the first indicator information and job seeker information, which is an example of target information. Recommended job information is not limited to what is exemplified below, but as an example, it may include job posting ID information, content information associated with the job posting ID information, employer information, conditions information, remarks information, first indicator information, reason information, second indicator information, or a combination thereof.

[0070] "Recommendation letter information" is information that represents a text used to recommend a job posting that is the target of recommendation, among the job postings which are an example of the requested information, and is information that can be used as recommendation information. Recommendation letter information is, for example, information generated by the processing unit 100 using a large-scale language model. Such recommendation letter information is not limited to what is exemplified below, but one example is the information shown in Figure 7B.

[0071] 4. Processing sequence executed by processing system 1 Figure 4 is a diagram showing a processing sequence executed in a processing system 1 according to one embodiment of the present disclosure. Specifically, Figure 4 shows a processing sequence executed between the processing unit 100, a job seeker terminal device 200-1 which is one of the target terminal devices, and a job offerer terminal device 200-2 which is one of the recommended person terminal devices. Of these, S11 to S16 show the category determination process of the job information which is the requested information, and S31 to S38 show the generation process of recommendation information for the job seeker which is the target. Each of the processes shown in these processing sequences is mainly executed by the processor processing a program stored in memory in each device.

[0072] (A) Processing to determine the category of job postings First, the process for determining the category of the requested job information will be explained. As shown in Figure 4, when the processor 211 of the recruiter terminal device 200-2 receives operation input from the recruiter via the input interface 213, it starts the application program for creating job postings (S11). After starting the application program, the processor 211 of the recruiter terminal device 200-2 displays a login screen for the service provided by the processing system 1 via the output interface 214. Then, the processor 211 of the recruiter terminal device 200-2 receives operation input from the recruiter via the input interface 213 and inputs the recruiter ID information and password (S12). The processor 211 of the recruiter terminal device 200-2 sends the received recruiter ID information and password to the processing unit 100 via the communication interface 215, along with a form request (T11) to request a job posting form for creating job postings.

[0073] Although not shown in the diagram, it is assumed that the job seeker has already been registered as a user of the service provided by processing system 1, and that the job seeker ID information and password have been stored in the job seeker management table (not shown) by processing device 100.

[0074] When the processor 111 of the processing unit 100 receives a form request via the communication interface 113, it authenticates whether the recruiter is a pre-registered person based on the recruiter ID information and password received along with the request (S13). If the recruiter ID information is stored in the recruiter management table, the processor 111 of the processing unit 100 checks whether the password associated with the recruiter ID information matches the received password. If the passwords match, the processor 111 of the processing unit 100 authenticates that the recruiter who sent the login request is a legitimate recruiter.

[0075] When the processor 111 of the processing unit 100 authenticates that the job seeker is a legitimate job seeker, it reads job application form information for creating a job application from memory 112 (S14). Although not specifically shown in the diagram, the job application form information includes at least items (content information, job seeker information, conditions information, and remarks information) and their input fields to prompt the job seeker to fill out as a job application. The processor 111 of the processing unit 100 then transmits the read job application form information (T12) to the job seeker terminal device 200-2 that sent the form request via the communication interface 113.

[0076] The processor 211 of the recruiter terminal device 200-2 receives job application form information via the communication interface 215 and outputs the received job application form information via the output interface 214. Then, the processor 211 of the recruiter terminal device 200-2 receives operation input from the recruiter via the input interface 213 and generates job information, which is one of the requested pieces of information (S15).

[0077] Here, Figure 7A shows an example of a job posting input screen 30 output in a recruiter terminal device 200-2 according to one embodiment of the present disclosure. Specifically, Figure 7A shows an example of a job posting input screen 30 output in S15 of Figure 4. According to Figure 7A, the job posting input screen 30 includes, as recruiter information, which is one of the items to prompt the recruiter to input according to the job posting form information received in S15 of Figure 4, the company name, industry, head office address, other business office addresses, affiliated company names, and average length of service, each item 31 and an input field 35 provided corresponding to each item. Similarly, the job posting input screen 30 also includes, as content information, the job description, industry, work location, and expected annual salary, each item 32 and an input field 36 provided corresponding to each item. Similarly, the job posting input screen 30 also includes, as condition information, the condition whether or not inexperienced applicants are accepted, and the language level, each item 33 and an input field 37 provided corresponding to each item. Similarly, the job posting input screen 30 includes, as remarks information, items 34 for training programs, the appeal of the job, employee messages, and how employees spend their days, as well as input fields 38 corresponding to each item.

[0078] The processor 211 of the recruiter terminal device 200-2 receives operational input from the recruiter to each input field via the input interface 213, thereby inputting information for each item of the job information. This input may be performed by the recruiter in free text format, or by the recruiter selecting a desired option from a pre-prepared set of choices.

[0079] Note that the job posting form information shown on the job posting input screen 30 in Figure 7A is merely an example. Therefore, it may include items other than those exemplified in Figure 7A. Furthermore, instead of using the job posting form information provided by the processing device 100, as exemplified in Figure 7A, it is also possible for job seekers to freely input job information using arbitrary form information or in free text format.

[0080] Returning to Figure 4, the processor 211 of the job seeker terminal device 200-2 generates job information when information for each item is entered on the job application input screen 30 illustrated in Figure 7A, and then transmits the generated job information (T15) as request information to the processing unit 100 via the communication interface 215.

[0081] When the processor 111 of the processing unit 100 receives job information, which is one of the requested pieces of information, via the communication interface 113, it generates new job posting ID information and stores the received job information (content information, recruiter information, conditions information, and remarks information) in the job posting management table in association with the job posting ID information (S16). Here, we are describing the case where job information is received from the recruiter terminal device 200-2, but other job information may be newly received from the recruiter terminal device 200-2, or even more job information may be newly received from other recruiter terminal devices. For each piece of received job information, new job posting ID information is generated for each piece of information, similar to the job information stored in S16, and it is stored in the job posting management table in association with each job posting ID information.

[0082] Next, the processor 111 of the processing unit 100 performs a category determination process (S17) for the job information received as one of the request information in S16, in order to generate a first index that indicates the degree of recommendation in each of a plurality of pre-set categories based on the job information. This category determination process is performed not only for the job information received in S16, but also for other job information, each time other job information is received.

[0083] Here, Figure 5A is a diagram showing the processing flow executed in a processing apparatus 100 according to one embodiment of the present disclosure. Specifically, Figure 5A shows the processing flow executed in the processing apparatus 100 in S16 and S17 of the processing sequence in Figure 4. This processing flow is mainly performed by the processor 111 of the processing apparatus 100 reading and executing a program stored in the memory 112.

[0084] According to Figure 5A, the processor 111 starts the processing flow by receiving job information, which is one of the requested pieces of information, from the recruiter terminal device 200-2 via the communication interface 113 (S111). Upon receiving the job information, the processor 111 generates new job posting ID information (S112). After generating the job posting ID information, the processor 111 stores the received job information, including content information, recruiter information, conditions information, and remarks information, in the job posting management table in association with the job posting ID information (S113).

[0085] In S113, the case of storing job information generated according to the job application form information provided by the processing unit 100 was described. However, as mentioned above, it is also possible to process job information that follows arbitrary form information or that is freely entered by the job seeker in free text format. In such cases, the processor 111 converts the received job information into a format that conforms to the job application form information provided by the processing unit 100, for example, using language analysis processing or a pre-trained generative model such as a large-scale language model. Then, similar to S113, the processor 111 stores each piece of the converted job information in the job application management table, associating it with the job application ID information.

[0086] Next, the processor 111 performs a category determination process to generate a first index indicating the degree of recommendation for each of a set of predefined categories based on the stored job information. This determination process can be performed, for example, by using a pre-trained generative model (e.g., a large-scale language model), a pre-trained determination model using machine learning, a rule-based process, or a combination thereof. Below, we will mainly describe an example of a process using a pre-trained generative model (e.g., a large-scale language model), but it is not limited to this. Furthermore, the degree of recommendation may be determined using a pre-trained generative model for characteristic categories and the degree of recommendation may be determined using a rule-based model for desired categories. This allows for reliable determination of the factors that job seekers prioritize when changing jobs, as well as determination of personality traits and preferences that are difficult to determine using rule-based methods.

[0087] (a1) Processing using a pre-trained generative model (e.g., a large-scale language model) Generative pre-trained models are models known as generative AI, which generate content such as text and images through deep learning or machine learning. Among such generative pre-trained models, large-scale language models are preferred. Even more preferred examples of such pre-trained decision models include BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer), with ChatGPT® or GPT-4® being particularly preferred examples.

[0088] When using a large-scale language model as described above, the processor 111 generates a prompt for input to the large-scale language model (S114). Specifically, the processing unit 100 generates a prompt by embedding each piece of information received as job information into form information prepared in advance for prompt generation. The prompt includes multiple categories, job information, and instructions to generate a first index representing the compatibility between the job information in each category. An example of such a prompt is as follows. <Example of a prompt> You are a professional HR representative. Below you will find job postings and categories. For each user category, indicate how well the job posting fits, i.e., rate its recommendation level on a scale of 0.1 to 1.0. Also, explain which parts of the job posting you focused on when assigning the rating. [Job postings] • Each piece of information entered into the job posting input screen 30 in Figure 7A (i.e., the information stored in the job posting management table is listed) [User Category] • Work steadily and diligently • Think logically • Can play as a team I want to acquire a variety of skills. I want to work at the same company for a long time.

[0089] The above prompt is merely an example, and of course, it may take other forms. Furthermore, the prompt is not limited to being generated by the processing unit 100 as described above; it may also be generated by an administrator on an administrator terminal device that can be used by an administrator who manages and operates the processing unit 100.

[0090] Next, when a prompt is generated, the processor 111 inputs the generated prompt into a generative system trained model such as a large-scale language model. The processor 111 then obtains a numerical value and its reason from the large-scale language model as output for each category entered in the prompt (S116). The processor 111 stores the obtained numerical value and its reason in the category management table as first indicator information and reason information (S117).

[0091] Here, Figure 6 is a diagram conceptually showing a specific example of a category management table according to one embodiment of the present disclosure. Specifically, Figure 6 is a diagram showing specific examples of first indicator information and reason information acquired in S116 and stored in the category management table in S117. According to Figure 6, the category information includes category ID information and category name assigned to each category. Specifically, it includes category ID information "G1" and its category name "Works steadily", category ID information "G2" and its category name "Thinks logically", category ID information "G3" and its category name "Can play in a team", category ID information "G4" and its category name "Wants to acquire various skills", and category ID information "G5" and its category name "Wants to work in the same workplace for a long time".

[0092] Furthermore, as shown in Figure 6, each category of information includes both primary indicator information and reason information. Specifically, for the category name "Working steadily," the numerical value (primary indicator information) assigned based on the job posting is "0.5," and the reason (reason information) includes the information, "From expressions such as 'We will conduct training for employees who joined without prior experience,' it is clear that perseverance is required." Thus, from the reason information, it can be inferred that in the judgment by the large-scale language model, the numerical value of 0.5 (primary indicator information) is determined based on the training system information entered as remarks information in the job posting. Similarly, as shown in Figure 6, other category names also include numerical values ​​(primary indicator information) and their reasons (reason information).

[0093] Returning to Figure 5A, the processing flow is terminated by storing the outputs from the large-scale language model in the category management tables, as described above.

[0094] As shown in Figure 5A, the first indicator information and reason information are generated based on job postings for each pre-defined category, and these categories are the same as those selected by job seekers. Therefore, since job seekers and employers (or job postings) can be evaluated using the same evaluation axis, it becomes possible to recommend employers (or job postings) more appropriately.

[0095] In Figure 5A, the process of S115 and S116 is described in which the processor 111 processes based on the large-scale language model stored in the memory 112 of the processing unit 100. However, it is also possible for another processing unit (for example, a large-scale language model server device) that stores the large-scale language model to process it. In this case, the processor 111 of the processing unit 100 inputs a prompt to the large-scale language model by sending the prompt generated in S114 to the large-scale language model server device via the communication interface 113. The processor 111 of the processing unit 100 then receives a numerical value and its reason as the output of the large-scale language model from the large-scale language model server device via the communication interface 113.

[0096] Furthermore, in Figure 5A, no particular distinction is made to the job postings referenced in the generation of the first indicator information and reason information for each of the multiple categories in the processing from S114 to S116. However, the job postings referenced in the generation of the first indicator information and reason information for each of the multiple categories in this processing may be of different types. Specifically, in the prompt generated in S114, the processor 111 specifies that for the characteristic categories of "works steadily," "thinks logically," and "can work in a team," it should refer to the content information, employer information, and remarks information from the job postings, and for the desired categories of "want to acquire various skills" and "want to work at the same workplace for a long time," it should refer to the employer information and condition information from the job postings to generate the first indicator information and reason information, respectively. Then, the processor 111 obtains, from the large-scale language model in which the prompt was input, first indicator information and reason information generated based on content information, recruiter information and remarks information for characteristic categories, and first indicator information and reason information generated based on recruiter information and condition information for desired categories.

[0097] (a2) Processing using a pre-trained decision model based on machine learning In Figure 5A, as described above, the first indicator information and reason information were generated using a pre-trained generative model such as a large-scale language model. However, it is also possible to generate the first indicator information and reason information using a pre-trained machine learning model instead, or in combination with this. Specifically, the processor 111 first obtains job information for training from the job seeker terminal device 200-2. Next, the processor 111 refers to the received job information for training and assigns numerical values ​​as correct label information for each pre-set category. These numerical values ​​may be assigned, for example, by an administrator via an administrator terminal device. The processor 111 then stores each job information for training and the correct label information assigned to each job information for training in association with each other.

[0098] Once the job information for training and the corresponding ground truth label information are obtained, the processor 111 performs a step of machine learning to create numerical assignment patterns using these. This machine learning is performed, for example, by providing these sets of information to a neural network made up of neurons and repeatedly training while adjusting the parameters of each neuron so that the output from the neural network is the same as the ground truth label information. As a result, the processor 111 obtains a trained judgment model for assigning numerical values ​​to each category based on job seeker information.

[0099] Furthermore, trained decision models can also be generated using methods employing neural networks such as neural networks, convolutional neural networks, multilayer herceptons (MLP), LSTM (Long Short-Term Memory), GRU (Gated Recurrent Unit), GNN (Graph Neural Network), and Transformers; methods employing gradient boosting decision trees (GBDT) such as LightGBM (Light Gradient Boosting Machine), XGBoost, and CatBoost; and machine learning methods such as ridge regression, logistic regression, support vector regression (SVR), nearest neighbor, decision trees, regression trees, and random forests.

[0100] Furthermore, the trained decision model is not limited to being generated by processing by the processor 111 of the processing unit 100; it may also be generated by a model generation device that is connected via communication.

[0101] As described above, once the trained judgment model is generated, the processor 111 inputs the job information stored in S113 into the trained judgment model. When the processor 111 obtains numerical values ​​for each category (first indicator information) as output from the trained judgment model, it stores them in the first indicator information.

[0102] Furthermore, the processor 111 obtains reason information by inputting the job information and category-specific numerical values ​​(first indicator information) stored in S113 into an explanatory analysis model for the inference of a trained decision model, such as SHAP Deep Explainer, Local Interpretable Model-agnostic Explanations, or a combination thereof. The processor 111 then stores the obtained reason information in association with the input first indicator information.

[0103] Thus, processing using a pre-trained judgment model based on machine learning also makes it possible to obtain the first indicator information and reason information, similar to Figure 5A.

[0104] The above description of the determination process describes the case where the processor 111 processes the data based on a trained determination model stored in the memory 112 of the processing unit 100. However, it is also possible for another processing unit (for example, a server device that stores the trained determination model) to perform the processing. In this case, the processor 111 of the processing unit 100 inputs the job information stored in S113 to the server device via the communication interface 113. The processor 111 of the processing unit 100 then receives the numerical value and its reason from the server device via the communication interface 113 as the output of the trained determination model.

[0105] (a3) Rule-based processing In Figure 5A, as described above, the first indicator information and reason information were generated using a pre-trained generative model such as a large-scale language model. However, it is also possible to generate the first indicator information and reason information by rule-based processing, either as an alternative or in combination with this method. Specifically, the processor 111 stores in advance a judgment table for each category, which associates the job information referenced as the target of judgment with the judgment criteria. For example, in the judgment table, for the category "I want a higher salary," the expected annual income from the content information is set as the referenced job information, and the judgment criteria are set as follows: 0.2 for 2 million yen or less, 0.4 for more than 2 million yen and 3 million yen or less, 0.6 for more than 3 million yen and 4 million yen or less, 0.8 for more than 4 million yen and 5 million yen or less, and 1.0 for more than 5 million yen. Furthermore, the judgment table also includes, as another example, the category "I want a raise," where the expected annual salary is set as the content information of the job information referenced, and the judgment criteria are set to evaluate it relatively against the desired annual salary or current annual salary, with values ​​of -1.0 for a pay cut, -0.5 for a pay increase or no pay cut, 0.5 for a 500,000 yen pay increase, 1.0 for a 1,000,000 yen pay increase, and 2.0 for any increase above that. In addition, the judgment table associates reason information for each judgment criterion, such as "This evaluation is given because the expected annual salary is 2,000,000 yen or less." When processor 111 stores the job information in S113, it reads the stored job information and refers to the judgment table exemplified above to determine the numerical value (first indicator information) and the reason (reason information) for each category. Then, processor 111 stores the determined numerical value (first indicator information) and the reason (reason information) in the category management table, associating them with the category information.

[0106] Thus, rule-based processing also makes it possible to obtain the first indicator information and reason information, similar to Figure 5A.

[0107] The above explanation focused on the "want a higher salary" category, but other categories can be processed similarly by using a judgment table that associates the referenced job information, evaluation criteria, and reason information. For example, for the "want to acquire various skills" category, the number of offices and affiliated companies can be set; for the "want to value private life" category, the number of holidays and average overtime hours can be set; for the "want to work for a stable company" category, the founding year and average length of service can be set; and for the "want to work close by" category, the work location can be set to create a judgment table. For example, in the judgment table for the "want to work close by" category, the work location is set as the referenced job information, and the evaluation criteria are set to evaluate the work location relatively from the desired work location, with values ​​of 2.0 for within a 100m radius, 1.0 for within a 500m radius, 0.5 for within a 5km radius, and 0 for beyond that.

[0108] Furthermore, while the above description uses job information stored in the job posting management table as the job information referenced when making a determination using the determination table, the processor 111 can also obtain job information referenced from websites containing job information or external database devices, for example, by using a large-scale language model.

[0109] Returning to Figure 4, once the first indicator information and reason information are stored in the category management table by the category determination process shown in Figure 5A, the processor 111 of the processing unit 100 generates a registration completion notification indicating that the process is complete and the job information has been registered in the job posting management table. The processor 111 of the processing unit 100 then sends the generated registration completion notification to the job seeker terminal device 200-2 that sent the job information via the communication interface 113. This completes the processing sequence.

[0110] As shown in Figure 4, the first indicator information and reason information are generated based on job postings for each pre-defined category, and these categories are the same as those selected by job seekers. Therefore, job seekers and employers (or job postings) can be evaluated using the same evaluation axis, making it possible to recommend employers (or job postings) more appropriately. Furthermore, categories are managed by dividing them into trait categories and desired categories, and the job postings referenced differ depending on the category. This makes it possible to generate the first indicator information and reason information more efficiently. In addition, as mentioned above, for relative categories that can be evaluated relatively to current or desired content, such as "I want a higher salary" or "I want to work somewhere close," it is preferable to generate the first indicator information and reason information by rule-based processing, and for qualitative categories where qualitative evaluation is preferable, such as "I work diligently" or "I think logically," it is preferable to generate the first indicator information and reason information by processing using a pre-trained generative model or a pre-trained judgment model using machine learning. This makes it possible to generate the first indicator information and reason information even more efficiently. Furthermore, for hypothetical categories that can be evaluated relatively, it is possible to generate primary indicator information and reasoning information not only through the methods mentioned above, but also through online judgment and processing.

[0111] (B) Recommendation information generation process Next, the process for generating recommendation information for the target job seeker will be explained. As shown in Figure 4, when the processor 211 of the job seeker terminal device 200-1 receives operation input from the job seeker via the input interface 213, it starts the application program for creating job seeker information (S31). After starting the application program, the processor 211 of the job seeker terminal device 200-1 displays a login screen for the service provided by the processing system 1 via the output interface 214. Then, the processor 211 of the job seeker terminal device 200-1 receives operation input from the job seeker via the input interface 213 and inputs the job seeker ID information and password (S32). The processor 211 of the job seeker terminal device 200-1 transmits the received job seeker ID information and password to the processing device 100 via the communication interface 215, along with a form request (T31) to request a job seeker information form for creating job seeker information.

[0112] Although not shown in the diagram, it is assumed that the job seeker has been registered in advance as a user of the service provided by processing system 1, and that the processing device 100 has stored the job seeker ID information and password in the job seeker management table (not shown in Figure 3C).

[0113] When the processor 111 of the processing unit 100 receives a form request via the communication interface 113, it authenticates whether the job seeker is a pre-registered person based on the job seeker ID information and password received along with the request (S13). If the job seeker ID information is stored in the job seeker management table, the processor 111 of the processing unit 100 checks whether the password associated with the job seeker ID information matches the received password. If the passwords match, the processor 111 of the processing unit 100 authenticates that the job seeker who sent the login request is a legitimate job seeker.

[0114] When the processor 111 of the processing unit 100 authenticates that the job seeker is a legitimate job seeker, it reads job seeker form information for creating the job seeker from the memory 112 (S34). Although not specifically shown in the diagram, the job seeker form information includes at least items (attribute information and selection information) and their input fields to prompt the job seeker to enter information. The processor 111 of the processing unit 100 then transmits the read job seeker form information (T32), along with the job seeker ID information, to the job seeker terminal device 200-1 that sent the form request via the communication interface 113.

[0115] When the processor 211 of the job seeker terminal device 200-1 receives job seeker form information via the communication interface 215, it outputs the received job seeker form information via the output interface 214. Then, the processor 211 of the job seeker terminal device 200-1 receives operation input from the job seeker via the input interface 213 and generates job seeker information, which is one of the target person information (S35).

[0116] Here, Figure 7B shows an example of a job seeker information input screen 10 output in a job seeker terminal device 200-1 according to one embodiment of the present disclosure. Specifically, Figure 7B shows an example of a job seeker information input screen 10 output in S35 of Figure 4. According to Figure 7B, the job seeker information input screen 10 includes a characteristic category selection area 11 and a desired category selection area 12 as selection information, which are one of the items to prompt the employer to input according to the job seeker form information received in S35 of Figure 4.

[0117] The characteristic category selection area 11 contains category names and other information for the same categories that are assigned to the job posting. Specifically, the characteristic category selection area 11 displays category names from among several pre-set categories, such as "works diligently," "thinks logically," "can work in a team," "plans," "gets along with people," "can explain things to people," "good at writing," and "likes challenges," along with checkboxes corresponding to each category name for selecting the desired category. In Figure 7B, up to three categories that are thought to correspond to the job seeker's own characteristics can be selected from the characteristic categories displayed in the characteristic category selection area 11. In other words, the processor 211 of the job seeker terminal device 200-1 accepts operation input from the job seeker to the desired checkboxes via the input interface 213, thereby allowing the job seeker to select the desired characteristic categories.

[0118] The desired category selection area 12 contains category names and other information for the same categories assigned to the job posting. Specifically, the desired category selection area 12 displays several pre-set categories as desired categories, such as "I want to acquire various skills," "I want to work at the same company for as long as possible," "I want to enrich my private life as well as my work," "I want to work for a stable company," and "I want to work overseas in the future," along with checkboxes corresponding to each category name for selecting the desired category. In Figure 7B, the job seeker can select up to three desired categories from those displayed in the desired category selection area 12. That is, the processor 211 of the job seeker terminal device 200-1 selects the desired categories by receiving input from the job seeker via the input interface 213 to select the desired categories.

[0119] Furthermore, the job seeker information input screen 10 includes a send icon. That is, when the processor 211 of the job seeker terminal device 200-1 receives an operation input from the job seeker to the send icon via the input interface 213, it stores the information of the selected characteristic category and desired category as selected information.

[0120] The processor 211 of the job seeker terminal device 200-1 accepts operational input from the job seeker to each checkbox via the input interface 213, thereby inputting selection information for each category. However, instead, the job seeker information input screen 10 displays the category name for each category and information indicating the degree of each category (for example, for the category "works steadily," the options are "good," "somewhat good," "average," "somewhat not good," and "not good"). The processor 211 of the job seeker terminal device 200-1 then accepts operational input from the job seeker to the options via the input interface 213 and selects the desired degree for each category.

[0121] Furthermore, although not specifically illustrated in Figure 7B, the job seeker information input screen 10 outputs various items of attribute information for each item in a series of transitioning screens, including the job seeker's name, contact information (e.g., email address, phone number, SNS account, etc.), location information (e.g., current address, etc.), desired location information (e.g., desired work location), desired job type, available start date (e.g., "from October 1, 2024"), current employment status (e.g., "scheduled to resign on July 31, 2024"), highest level of education, qualifications held, and the desired time for sending recommendation information (e.g., "around 9:00 every morning"), along with input fields provided for each item.The processor 211 of the job seeker terminal device 200-1 then receives operational input from the job seeker into each input field via the input interface 213, thereby inputting information for each item of attribute information. The input may be made by the job seeker entering information in free text format, or by the job seeker selecting a desired option from a pre-prepared list of choices. The processor 211 of the job seeker terminal device 200-1 stores each piece of input information as attribute information.

[0122] Furthermore, the job seeker form information shown on the job seeker information input screen 10 in Figure 7B is merely an example. Therefore, it may include items other than those exemplified in Figure 7B. Also, instead of using the job seeker form information provided by the processing device 100, as exemplified in Figure 7B, it is possible for job seekers to follow arbitrary form information or to freely input job seeker information in free text format. In addition, although not illustrated, the system may automatically generate information not only from the job seeker's input but also from the job seeker's attribute information, such as their viewing history or browsing history of specific content (for example, if they have a viewing history of self-improvement content on logical thinking methods, this will be automatically entered into "Think Logically").

[0123] Returning to Figure 4, the processor 211 of the job seeker terminal device 200-1 generates job seeker information when attribute information and selection information are entered on the job seeker information input screen 10 illustrated in Figure 7B. It then transmits the generated job seeker information (T33) along with the job seeker ID information to the processing unit 100 via the communication interface 215 as target information.

[0124] When the processor 111 of the processing unit 100 receives job seeker information, which is one of the target person information, via the communication interface 113, it stores the received job seeker information (selection information and attribute information) in the job seeker management table, associating it with the received job seeker ID information (S36). Here, we are describing the case where job seeker information is received from the job seeker terminal device 200-1, but additional job seeker information may be newly received from other job seeker terminal devices. Each received job seeker information is stored in the job posting management table, associating it with each job seeker ID information, similar to the job seeker information stored in S36.

[0125] Next, the processor 111 of the processing unit 100 performs a matching process with each job posting information based on the job seeker information received as one of the target information in S36 and the first index information associated with each job posting ID information stored in S17 (S37).

[0126] Here, Figure 5B is a diagram showing the processing flow executed in a processing apparatus 100 according to one embodiment of the present disclosure. Specifically, Figure 5B shows the processing flow executed in the processing apparatus 100 in S36 and S37 of the processing sequence in Figure 4. This processing flow is mainly performed by the processor 111 of the processing apparatus 100 reading and executing a program stored in the memory 112.

[0127] According to Figure 5B, the processor 111 starts the processing flow by receiving job seeker information, which is one of the target person information, from the job seeker terminal device 200-1 via the communication interface 113 (S211). When the processor 111 receives the job seeker information, it stores the attribute information and selection information, which are the received job seeker information, in the job seeker management table, in association with the job seeker ID information that was also received (S212).

[0128] In S212, the case of storing job seeker information generated according to the job seeker form information provided by the processing unit 100 was described. However, as mentioned above, it is also possible to process job seeker information according to arbitrary form information or freely entered by the job seeker in free text format. In such cases, the processor 111 converts the received job seeker information into a format according to the job seeker form information provided by the processing unit 100, for example, using language analysis processing or a pre-trained generative model such as a large-scale language model. Then, similar to S212, the processor 111 stores each piece of the converted job seeker information in the job seeker management table, associating it with the job seeker ID information.

[0129] Next, the processor 111 performs a matching process based on the first indicator information assigned to each job posting ID information for each category, and the job seeker information stored in S212. This matching process may be performed by, for example, calculating an accumulated value for multiple categories, weighting, filtering based on the job seeker's attribute information, or a combination of these. Below, we will mainly describe the process of calculating an accumulated value for multiple categories, but of course, it is not limited to this. For example, the second indicator can be calculated using the recommendation level of the category selected by the job seeker, and the algorithm is not limited.

[0130] (b1) Process to calculate cumulative values ​​for each of the multiple categories Processor 111 refers to the selection information contained in the job seeker information and assigns the value "1" to the category selected by the job seeker and the value "0" to the category that was not selected. Then, Processor 111 takes the product of the value assigned to each category of the selection information as described above and the value of the first indicator information for the category that corresponds to the category of the selection information among the categories assigned to the job information. More specifically, if "works steadily" is selected in the selection information, Processor 111 assigns "1" to that category. Also, Processor 111 refers to the job posting management table and the category management table and reads the first indicator information associated with the category "works steadily" for each job information. Then, Processor 111 calculates the product of the "1" assigned to the selection information and the "0.5" assigned to the job information (see, for example, Figure 6). Processor 111 similarly calculates the product of the value assigned to the selection information and the value assigned to the job information for all categories. The processor 111 then accumulates the calculated values ​​and stores the accumulated value as second indicator information, associating it with each job posting ID information (S213). Alternatively, the system may read only the values ​​of the categories selected by the job seeker from the first indicator information. Alternatively, each category may be treated as a vector, and the values ​​"1" for the categories selected by the job seeker and "0" for the categories not selected may be assigned to generate vector information corresponding to the job seeker. Vector information may then be generated from the first indicator information for each job posting, and the similarity (e.g., cosine similarity or dot product comparison) between the job seeker's vector information and the vector information of each job posting may be calculated, with the calculated similarity being used as the second indicator information.

[0131] Next, the processor 111 reads the second indicator information stored in association with each job posting ID information and selects the job posting ID information for which the accumulated value exceeds a pre-set threshold as a recommendation target (S214).

[0132] (b2) Weighting process As described above, in S213 and S214, a cumulative value was calculated as the second indicator information, and job posting ID information was selected to identify job postings to be recommended based on this cumulative value. However, it is also possible to weight each category using predetermined weighting coefficients when calculating the cumulative value. Furthermore, when calculating the similarity between the vector information of job seekers and the vector information of each job posting, a configuration in which each vector is weighted is also possible.

[0133] For example, as explained in Figure 7B, when inputting selection information on the job seeker information input screen 10, it is possible to select information indicating the degree of each category (for example, for the category "works steadily," the options are "good," "somewhat good," "average," "somewhat not good," and "not good"). In other words, the processor 111 can store this degree-indicating information for each category as selection information and use this information to weight the results. For example, if "good" is selected, the processor 111 multiplies the product of the number assigned to the selection information and the number assigned to the job information by "×2," if "somewhat good" is selected, by "×1.5," if "average" is selected, by "×1," if "somewhat not good" is selected, by "×0.7," and if "not good" is selected, by "×0.5." The processor 111 then obtains the second indicator information by summing the weighted values ​​obtained for each category. In this way, the processor 111 can select job information that better reflects the job seeker's preferences by weighting the results according to the selection information selected by the job seeker.

[0134] Furthermore, in Figure 7B, it is possible to set priorities within the selected categories by further selecting categories that the job seeker considers important for each category selected on the job seeker information input screen 10. In other words, the processor 111 further selects categories that indicate a high priority from among the categories selected by the job seeker as selection information. Then, for categories with a high priority, a value such as "×1.2" is multiplied by the product of the value assigned to the selection information and the value assigned to the job information. The processor 111 then obtains the second indicator information by summing the weighted values ​​obtained for each category. In this way, the processor 111 can select job information that better reflects the job seeker's preferences by weighting the categories according to their priority.

[0135] Furthermore, as shown in Figure 7A, the job posting input screen 30 allows employers to select categories they prioritize in advance, thereby reflecting those categories. Specifically, the processor 111 acquires the information entered on the job posting input screen 30 as job information, as well as the selection of categories that employers prioritize. The processor 111 then multiplies the product of the values ​​assigned to the selection information and the values ​​assigned to the job information by a value such as "×1.2" for the selected categories. The processor 111 then acquires the second indicator information by summing the weighted values ​​obtained for each category. In this way, the processor 111 can select job information that better reflects the employer's preferences by weighting the information according to the categories selected by the employer.

[0136] Next, the processor 111 reads the second indicator information stored in association with each job posting ID information and selects the job posting ID information for which the accumulated value exceeds a pre-set threshold as a recommendation target.

[0137] (b3) A process to filter based on the attribute information of job seekers. It is also possible to further filter the job postings selected by the processes in b1 or b2 above based on attribute information from the job seeker information. For example, the job seeker's desired location information (e.g., desired work location) is entered as attribute information on the job seeker information input screen 10. The processor 111 then refers to the location information of the head office from the job postings associated with the job posting ID information selected by the processes in b1 or b2 above, and filters the job postings to be selected as recommendations based on each location information (for example, it is possible to filter if the work location desired by the job seeker is more than a predetermined distance from the location information of the job posting, or to sort in order of proximity to the location information of the job posting). The timing of the filtering can be arbitrary, and for example, to reduce the difference in calculating the second indicator for each job seeker, filtering may be performed using the job seeker's attribute information to narrow down the job postings that are subject to calculation of the second indicator. This eliminates the need to calculate the second indicator for all job postings, and is expected to reduce the processing load.

[0138] Furthermore, while the above example filtered based on the desired location information of job seekers and the location information of employers (for example, the location of the head office), the processor 111 can, of course, also filter using other attributes such as desired job type.

[0139] When the processor 111 selects the job posting ID information of a job posting to be recommended through any of the processes b1 to b3 above, it generates recommendation information based on at least one of the job posting information, first indicator information, reason information, and second indicator information associated with the job posting ID information (S215). Such recommendation information may include, for example, recommended job posting information and recommendation text information. The processor 111 reads pre-set information from the job posting information, first indicator information, reason information, and second indicator information to generate recommended job posting information. The processor 111 stores the generated recommended job posting information as recommendation information, associated with the job seeker ID information. An example of the generated recommended job posting information is explained in Figure 7C.

[0140] Furthermore, the processor 111 generates recommendation information based on at least one of the job information, first indicator information, reason information, and second indicator information associated with the job posting ID information. For example, the processor 111 generates recommendation information using a generative pre-trained model (e.g., a large-scale language model). As mentioned above, preferred examples of such generative pre-trained models include BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer), with ChatGPT or GPT-4 being particularly preferred examples among these.

[0141] The processor 111 first reads the job information, first indicator information, reason information, and second indicator information associated with the job posting ID information in order to generate recommendation information, and generates a prompt. That is, the prompt includes an instruction to generate a recommendation statement for recommending the job information based on the job information, first indicator information, reason information, and second indicator information. An example of such a prompt is: "You are a professional HR person. Write a recommendation statement for the job information based on the reason information of the two categories with the highest numerical values ​​from the first indicator information associated with the job posting ID information, the reason information of the category with the lowest numerical value, and the job information associated with the job posting ID information."

[0142] The above prompt is merely an example, and of course, it may take other forms. Furthermore, the prompt is not limited to being generated by the processing unit 100 as described above; it may also be generated by an administrator on an administrator terminal device that can be used by an administrator who manages and operates the processing unit 100.

[0143] Next, when a prompt is generated, the processor 111 inputs the generated prompt into a pre-trained generative model, such as a large-scale language model. The processor 111 then obtains recommendation information as output from the large-scale language model. The processor 111 stores the obtained recommendation information as recommendation information in the job seeker management table, associating it with the job seeker ID information. An example of the generated recommendation information is explained in Figure 7C.

[0144] When the processor 111 generates recommendation information, it transmits the generated recommendation information to the job seeker terminal device 200-1 that sent the job seeker information via the communication interface 113 (S216). This completes the processing flow.

[0145] As shown in Figure 5B, efficient processing is possible by matching the first indicator information with the job seeker information. In particular, the same categories are set for the categories to which the first indicator information is assigned and the categories that job seekers can select. Therefore, since job seekers and employers (or job postings) can be evaluated using the same evaluation axis, it becomes possible to recommend employers (or job postings) more appropriately.

[0146] In Figure 5B, the process of S215 is described in which the processor 111 processes the process based on a large-scale language model stored in the memory 112 of the processing unit 100. However, it is also possible for another processing unit (for example, a large-scale language model server unit) that stores the large-scale language model to process it. In this case, the processor 111 of the processing unit 100 inputs a prompt to the large-scale language model by sending the prompt generated in S215 to the large-scale language model server unit via the communication interface 113. The processor 111 of the processing unit 100 then receives the recommendation text information from the large-scale language model server unit as output of the large-scale language model via the communication interface 113.

[0147] Returning to Figure 4, when recommendation information is generated by the matching process, the processor 111 of the processing unit 100 transmits the generated recommendation information (T34) to the job seeker terminal device 200-1 that sent the job seeker information via the communication interface 113.

[0148] When the processor 211 of the job seeker terminal device 200-1 receives recommendation information via the communication interface 215, it outputs the received recommendation information via the output interface 214 (S38).

[0149] Here, Figure 7C shows an example of a recommendation information screen 20 output in a job seeker terminal device 200-1 according to one embodiment of the present disclosure. Specifically, Figure 7C shows an example of a recommendation information screen 20 output in S38 of Figure 4. According to Figure 7C, the recommendation information screen 20 includes a recruiting company display area 28, a recommendation text display area 21, a desired category display area 22, a characteristic category display area 23, a matching degree display area 26, and a reaction input area 25. The information displayed in each area includes recommendation information and recommendation text information generated using the job seeker information of each target job seeker. Therefore, different content may be displayed on the recommendation information screen 20 for each job seeker. Specifically, it is as follows.

[0150] The recruiting company display area 28 is information output by the processor 111 when the company name is read from the recruiter information of the job information for which recommendation information has been generated. In other words, by the information displayed in the recruiting company display area 28, job seekers can identify the company that is posting the job information displayed on the recommendation information screen 20.

[0151] The recommendation display area 21 is the area where the recommendation information generated by the processor 111 is displayed. In the example in Figure 7C, based on the job information, first indicator information, reason information, and second indicator information, the recommendation information displayed is: "After reviewing your resume, I strongly feel that you have the potential to fully utilize your strengths and thrive at XXX Company, Ltd., which places great importance on teamwork."

[0152] The desired category display area 22 displays classification information and recommendation information generated based on the first indicator and reason information managed in the category management table in Figure 6, for the desired category selected by the job seeker from the desired category selection area 12 of the job seeker information input screen 10 in Figure 7B. In the example in Figure 7C, the display shows the case when three desired categories are selected in the job seeker information input screen 10 in Figure 7B: "I want to acquire various skills," "I want to work at the same workplace for as long as possible," and "I want to enrich my private life as well as my work." According to Figure 7C, the desired category display area 22 is • Title information for each desired category selected by the job seeker (e.g., "Skills" indicating "I want to acquire various skills," "Long-term employment" indicating "I want to work at the same company for as long as possible," and "Private life" indicating "I want to have a fulfilling private life as well as a fulfilling work life"). • In the category management table for job postings at XXX Company, Ltd., classification information indicating the primary indicator information associated with the desired category selected by the job seeker (e.g., "<Low>" indicating the primary indicator information for "I want to acquire various skills", "<High>" indicating the primary indicator information for "I want to work at the same company for as long as possible", and "<High>" indicating the primary indicator information for "I want to enrich not only my work life but also my private life"). Includes tabs containing [this].

[0153] Furthermore, each tab in the desired category display area 22 includes an area for displaying recommendation information generated based on reason information. In the example in Figure 7C, in the category management table for job postings of XXX Company, the recommendation information displayed is "The average length of service is 15 years, and employee benefits are excellent. Although there are few opportunities for transfers between departments, you can work in a fulfilling environment for a long time," which was generated based on the reason information associated with the desired category "I want to work at the same company for as long as possible," selected by the job seeker. It is also possible to display the reason information associated with each desired category by accepting input from the job seeker for the tab they wish to refer to. That is, when the job seeker accepts input for the "Skills" tab, the recommendation information generated based on the reason information associated with the desired category "I want to acquire various skills" is displayed, and when the job seeker accepts input for the "Private Life" tab, the recommendation information generated based on the reason information associated with "I want to have a fulfilling private life as well as a fulfilling work life" is displayed.

[0154] The characteristic category display area 23 displays recommendation information generated based on the characteristic category selected by the job seeker from the characteristic category selection area 11 of the job seeker information input screen 10 in Figure 7B. In the example in Figure 7C, the display shows the case where two characteristic categories, "Ability to work in a team" and "Ability to think logically," are selected as characteristic categories in the job seeker information input screen 10 in Figure 7B. According to Figure 7C, the characteristic category display area 23 contains the category name of the category with the highest first indicator information from each characteristic category selected by the job seeker in the category management table of the job information of XXX Company, Ltd. ("<Ability to work in a team>") and recommendation information generated based on the reason information ("Each project is carried out in teams of 3 to 5 people, and you can work while discussing with everyone as a member of the team."). In other words, the characteristic category display area 23 displays the characteristics (features) of job information that are likely to suit the job seeker based on the characteristic category selected by the job seeker.

[0155] In this way, the desired category display area 22 and the characteristic category display area 23 will display information corresponding to each category selected by the job seeker. This will enable job seekers to refer to job postings more efficiently. Furthermore, by displaying classification information indicating the first indicator information in the desired category display area 22, it will be possible to appropriately understand how highly the displayed category was rated.

[0156] The matching degree display area 26 displays the numerical value (cumulative value) of the second indicator information calculated for the displayed job information. Here, the recommendation information screen 20 exemplified in Figure 7C is generated for each job ID information in which the numerical value of the second indicator information exceeds a pre-set threshold. Therefore, by displaying the numerical value (cumulative value) of the second indicator information for each job information displayed on the recommendation information screen 20, job seekers can appropriately identify job information with a high degree of matching.

[0157] In addition, while the desired category display area 22 displays classification information based on the first indicator information, the numerical value of the first indicator information may also be displayed. Furthermore, while the characteristic category display area 23 displays information for one characteristic category based on the first indicator information, information for multiple characteristic categories selected by the job seeker may also be displayed along with the first indicator information. In addition, the numerical values ​​displayed in the matching degree display area 26 may be the stored numerical values ​​themselves, or the classifications or categories assigned based on each numerical value may be displayed.

[0158] The reaction input area 25 is an area for inputting reaction information that indicates the job seeker's reaction to each recommendation displayed on the recommendation information screen 20. The processor 211 of the job seeker terminal device 200-1 receives operation input from the job seeker to the reaction input area 25 via the input interface 213 and selects reaction information that indicates the job seeker's reaction. In the example in Figure 7C, when the job seeker makes an operation input to the reaction input area 25, the "Like!" display is highlighted, and reaction information indicating that the job seeker liked the job information identified by that recommendation information has been selected.

[0159] Although not specifically illustrated in Figure 4, the processor 111 of the processing unit 100 stores the job posting ID information for which reaction information has been selected, in association with each job seeker ID information. When the processor 111 of the processing unit 100 receives a filtering request based on reaction information from the job seeker terminal device 200-1, it reads the job posting ID information for which reaction information has been selected and filters it to show only the recommendation information generated for that job posting ID information. The processor 111 of the processing unit 100 then sends only this recommendation information to the job seeker terminal device 200-1, allowing the job seeker to efficiently refer only to the recommendation information they like.

[0160] Furthermore, the recommendation information screen 20 includes an application icon 24. That is, when the processor 211 of the job seeker terminal device 200-1 receives an operation input from the job seeker to the application icon 24 via the input interface 213, it can execute the process of applying to the job information included in the recommendation information. For example, when the processor 211 of the job seeker terminal device 200-1 receives an operation input to the application icon 24, it transmits the job information (job posting ID information) and job seeker ID information included in the recommendation information to the processing device 100 via the communication interface 215. When the processor 111 of the processing device 100 receives the job posting ID information, it refers to the job seeker management table and reads the selection information and attribute information associated with the job seeker ID information. Then, the processor 111 of the processing device 100 transmits the read selection information and attribute information to the job seeker's job seeker terminal device 200-2 associated with the job posting ID information via the communication interface 113. Generally, when an application is submitted for a job, it goes through a document screening process followed by interviews. However, in this case, for example, it is possible to skip the document screening and start directly with the interview process. In this way, by using the application icon 24, it is possible to proceed with the application process more efficiently.

[0161] Returning to Figure 4, the processor 111 generates a recommendation completion notification to the employer's terminal device 200-2, which is the employer of the job posting for which the recommendation information was generated, informing them that the job posting has been recommended. The processor 111 then transmits the generated recommendation completion notification to the employer's terminal device 200-2 via the communication interface 113. This completes the processing sequence.

[0162] As shown in Figure 4, efficient processing is possible by matching the first indicator information with the job seeker information. In particular, the same categories are set for the categories to which the first indicator information is assigned and the categories that job seekers can select. Therefore, since job seekers and employers (or job postings) can be evaluated using the same evaluation axis, it becomes possible to recommend employers (or job postings) more appropriately. Furthermore, by using processes such as weighting and filtering in the generation of recommendation information, it is possible to process even more efficiently.

[0163] As shown in Figures 4 and 7C, the processor 111 of the processing unit 100 stores the job posting ID information for which reaction information has been selected, in association with each job seeker ID information. Conversely, this means that for job postings included in recommendation information for which no reaction information has been selected, the matching process in S37 of Figure 4 may not have been performed correctly. Therefore, the processor 111 of the processing unit 100 can also change the weighting coefficients used in the matching process. For example, the processor 111 changes the weighting coefficients by updating the weighting coefficients assigned to each category of the above job posting information to lower values.

[0164] Furthermore, the processor 111 of the processing unit 100 can also change the weighting coefficients for categories of other job postings that are similar to the job postings of the job posting ID information for which reaction information has been selected. For example, the processor 111 of the processing unit 100 can refer to the job postings for which reaction information has been selected, read out common job postings (for example, all of which have an estimated annual salary of 6 million yen or more), and change the weighting coefficients by updating the weighting coefficients for categories related to the common job postings to higher values.

[0165] Furthermore, the processor 111 of the processing unit 100 can also send a feedback request along with the recommendation information as feedback information for the pre-generated recommendation information, asking, "Of the categories output as recommendation information, which categories did not match?" The processor 211 of the job seeker terminal device 200-1 accepts the selection of feedback information from the job seeker by outputting the feedback request together with the recommendation information screen 20 via the output interface 214. The processor 211 of the job seeker terminal device 200-1 then transmits the received feedback information to the processing unit 100 via the communication interface 215. Upon receiving the feedback information, the processor 111 of the processing unit 100 modifies the weighting coefficients set for the categories that received feedback indicating they did not match, for example, by updating them to lower values.

[0166] In summary, this disclosure provides a processing device, processing program, processing method, and processing system that enable more appropriate recommendations of nominees. [Explanation of Symbols]

[0167] 1. Processing System 100 Processing Units 200-1 Job seeker terminal device 200-2 Job seeker terminal device

Claims

1. A processing unit comprising at least one processor, The at least one processor, The system receives subject information from the subject terminal device, which includes at least selection information indicating the subject's choices for each of the multiple categories representing the characteristics of the respondent or the multiple categories representing the desired content of the respondent. Multiple request information items, each specifying the content requested by one or more nominees, are received and stored from each nominee's terminal device. Based on the aforementioned multiple request information, a first index is generated that indicates the degree of recommendation in each of the multiple categories representing the characteristics of the respondent or the multiple categories representing the desired content of the respondent, Based on the first indicator and the target information, recommendation information is generated that indicates the recommended request information from among the multiple request information. A processing unit configured to perform a process for that purpose.

2. The aforementioned individuals are job seekers, The aforementioned person being recommended is the employer. The apparatus according to claim 1.

3. The requested information is job information. The aforementioned recommendation information is a recommended job posting from among multiple job postings, generated based on a first indicator for each job posting in the category selected by the job seeker. The apparatus according to claim 2.

4. The processing apparatus according to claim 1, wherein the first index is generated by inputting a prompt to a large-scale language model, the prompt comprising: a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the desires of the respondent; a single request information included in the plurality of request information; and a command to generate an index as a compatibility between the single request information in each of the plurality of categories indicating the characteristics of the respondent or the plurality of categories indicating the desires of the respondent.

5. The apparatus according to claim 1, wherein the recommendation information includes reason information indicating the reason why the first indicator was generated.

6. The processing apparatus according to claim 5, wherein the first indicator and the reason information are generated by a large-scale language model, and the prompt includes a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the desires of the respondent, one request information included in the plurality of request information, and an indicator as the compatibility between the one request information in each of the plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the desires of the respondent, and an instruction to generate an indicator as the compatibility and the reason for determining the compatibility.

7. The apparatus according to claim 1, wherein the recommendation information includes recommendation statement information for recommending the request information, which is generated by inputting a prompt to a large-scale language model that includes an instruction to generate a recommendation statement for recommending the request information based on the first indicator and the target information.

8. The apparatus according to claim 1, wherein the multiple categories indicating the characteristics of the respondent are characteristic categories indicating the personality traits of the job seeker, and the multiple categories indicating the desires of the respondent are desire categories indicating what the job seeker desires for the job.

9. The processing apparatus according to claim 1, wherein the recommendation information is generated based on a second indicator indicating the degree of compatibility between the first indicator and the subject information, which is calculated based on the first indicator and the subject information.

10. The aforementioned request information includes multiple types of request information with different content, The first indicator is generated based on different types of request information in each of the multiple categories representing the characteristics of the respondent or the multiple categories representing the desires of the respondent. The apparatus according to claim 1.

11. The requested information includes information about location, The aforementioned subject information includes location information entered by the subject, The processing apparatus according to claim 1, wherein the recommendation information is generated based on a first indicator of a request information selected from a plurality of request information based on the location information and the location information of the subject.

12. The processing apparatus according to claim 9, wherein the second indicator is generated based on the first indicator and target information, and the weighting of each of a plurality of categories indicating the characteristics of the respondent or a plurality of categories indicating the desired content of the respondent.

13. The processing apparatus according to claim 12, wherein the weighting is modified based on feedback from the subject.

14. The processing apparatus according to claim 1, wherein the weighting changes based on the subject's feedback are changed based on reaction information indicating the subject's reaction to the recommendation information.

15. The processing apparatus according to claim 1, which changes the content of the recommendation information displayed on the target user terminal device based on the selection information.

16. By being executed by at least one processor provided in the computer, The system receives subject information from the subject terminal device, which includes at least selection information indicating the respondent's choices for each of the multiple categories indicating the respondent's characteristics or the multiple categories indicating the respondent's preferences. Multiple request information items, each specifying the content requested by one or more nominees, are received and stored from each nominee's terminal device. For each of the aforementioned multiple request information, a first index is generated based on each request information, indicating the degree of recommendation in each of the multiple categories representing the characteristics of the respondent or the multiple categories representing the desired content of the respondent. Based on the first indicator and the target information, recommendation information is generated that indicates the recommended request information from among the multiple request information. A processing program that causes the aforementioned at least one processor to function in this manner.

17. A processing method that is performed by at least one processor provided in a computer, The steps include receiving subject information from a subject terminal device, which includes at least selection information indicating the respondent's choices for each of the multiple categories indicating the respondent's characteristics or the multiple categories indicating the respondent's preferences, The steps include: storing multiple request information, each specifying the content requested by one or more nominees, by receiving it from each nominee's terminal device; For each of the aforementioned multiple pieces of request information, a first index is generated based on each piece of request information, indicating the degree of recommendation in each of the multiple categories representing the characteristics of the respondent or the multiple categories representing the desired content of the respondent. A step of generating recommendation information that indicates the recommended request information from among the plurality of request information based on the first indicator and the target person information, A processing method that includes this.

18. The processing apparatus according to any one of claims 1 to 15, A target user terminal device configured for use by the target user and connected to the processing device in a communicative manner, A terminal device for recommended persons, configured to be usable by the recommended person and connected to the processing device in a communicative manner, A processing system that includes this.

Images