Human resource management system, human resource management method and program

Abstract

A human resources management system, a human resources management method, and a program are provided that can facilitate the management of output information using a machine learning model. [Solution] In a generating step (S12-S13), the control unit generates first output information from one or more pieces of material information using a trained model. The trained model is a model that has been trained to receive input information including at least a portion of the material information and output first output information including at least a portion of the material information. In a determining step (S14), the control unit determines attributes of the generated first output information based on attributes of the one or more pieces of material information.

Description

[Technical Field]

[0001] The present disclosure relates to a human resources management system, a human resources management method, and a program. [Background technology]

[0002] In recent years, technology for generating new information by utilizing machine learning models such as generative AI (Artificial Intelligence) has been attracting attention. For example, Patent Document 1 discloses a writing support system that uses artificial intelligence to create approach sentences from source sentences that include profile information related to either an organization or an individual. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 7403023 Summary of the Invention [Problem to be solved by the invention]

[0004] However, while the spread of generative AI has made it easier to generate new information, it has also made it more difficult to properly manage this information. In particular, in human resource management systems that manage information about human resources, the raw information often contains personal and private information, so careful handling of the output of machine learning models is required. Therefore, there is a need to make it easier to manage the information output by machine learning models.

[0005] In view of the above-mentioned problems, an object of the present disclosure is to provide a human resources management system, a human resources management method, and a program that can facilitate the management of output information using machine learning models. [Means for solving the problem]

[0006] In a human resources management system according to one aspect of the present disclosure, the one or more processors are capable of executing a generating step and a determining step. In the generating step, the one or more processors generate first output information from one or more pieces of material information using a trained model. The trained model receives input information including at least a portion of the material information and outputs first output information including at least a portion of the material information. In the determining step, the one or more processors determine attributes of the generated first output information based on attributes of the one or more pieces of material information.

[0007] A human resources management method according to one aspect of the present disclosure is implemented by one or more processors. The human resources management method includes a generating step and a determining step. The generating step is a step of generating first output information from one or more pieces of material information using a trained model. The trained model receives input information including at least a portion of the material information, and outputs first output information including at least a portion of the material information. The determining step is a step of determining attributes of the generated first output information based on attributes of the one or more pieces of material information.

[0008] A program according to one aspect of the present disclosure operates a computer including one or more processors and a memory. The program causes the one or more processors to execute a generating step and a determining step. In the generating step, the one or more processors generate first output information from one or more pieces of material information using a trained model. The trained model receives input information including at least a portion of the material information and outputs first output information including at least a portion of the material information. In the determining step, the one or more processors determine attributes of the generated first output information based on attributes of the one or more pieces of material information. [Effects of the Invention]

[0009] The present disclosure provides a human resources management system, a human resources management method, and a program that can facilitate the management of output information using machine learning models. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a block diagram showing the overall configuration of a human resources management system according to a first embodiment. [Figure 2] FIG. 2 is a block diagram showing a hardware configuration of a registrant terminal according to the first embodiment. [Figure 3] 1 is a block diagram showing a hardware configuration of a human resources management device according to a first embodiment. [Figure 4] 1 is a block diagram showing a functional configuration of a human resources management device according to a first embodiment. [Figure 5] 1 is a flowchart showing the flow of a human resources management method according to a first embodiment. [Figure 6] FIG. 10 is a diagram illustrating an example of a data structure of a material attribute table. [Figure 7] FIG. 10 is a diagram illustrating an example of a data structure of a material output table. [Figure 8] FIG. 10 is a diagram illustrating an example of a data structure of an output attribute table. [Figure 9] FIG. 10 is a block diagram showing the overall configuration of a human resources management system according to a second embodiment. [Figure 10] FIG. 10 is a block diagram showing the functional configuration of a human resources management device according to a second embodiment. [Figure 11] 10 is a flowchart showing the flow of an attribute determination method according to the second embodiment. [Figure 12] FIG. 10 is a diagram illustrating an example of a screen. [Figure 13] FIG. 10 is a diagram illustrating an example of a screen. [Figure 14] FIG. 10 is a diagram illustrating an example of a screen. [Figure 15] 10 is a flowchart showing the flow of a search processing method according to a second embodiment. [Figure 16] FIG. 10 is a diagram illustrating an example of a screen. [Figure 17] FIG. 10 is a diagram illustrating an example of a screen. [Figure 18] FIG. 10 is a diagram illustrating an example of a screen. [Figure 19]FIG. 10 is a diagram illustrating an example of a screen. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In each drawing, the same or corresponding elements are designated by the same reference numerals, and for clarity of explanation, duplicate explanations will be omitted as necessary.

[0012] <Embodiment 1> First, a first embodiment of the present disclosure will be described. 1. Hardware Configuration This section describes the hardware configuration of the human resource management system 1 according to the present embodiment. Fig. 1 is a block diagram showing the overall configuration of the human resource management system 1 according to the first embodiment.

[0013] 1.1 Human Resource Management System1 The human resources management system 1 is a computer system that manages information related to human resources. In particular, the human resources management system 1 manages information related to members of an organization. The organization may be a company, government agency, voluntary association, or other organization. The members may be employees, executives, or other members of the organization. The human resources management system 1 comprises a human resources management device 2 and a registrant terminal 3. The human resources management device 2 and the registrant terminal 3 are connected to each other so that they can communicate with each other via a network NW such as the Internet.

[0014] 1.2 Registered User Terminal 3 The registrant terminal 3 is a computer terminal operated by a registrant who registers material information about members. For example, the registrant may be the member himself / herself, a human resources or general affairs officer of an organization, or a person in a position to manage members. For example, the registrant terminal 3 may be a smartphone, a tablet terminal, a personal computer, or other communication terminal capable of input and output. The registrant terminal 3 transmits material information about members received through operation by the registrant to the human resources management device 2 via the network NW. The material information about members is the personnel information of the members. Material information will be described in detail later.

[0015] FIG. 2 is a block diagram showing the hardware configuration of the registrant terminal 3 according to the first embodiment. The registrant terminal 3 comprises a control unit 31, a communication unit 35, a memory unit 36, an input unit 37, and a display unit 38, and these components are electrically connected via a communication bus inside the registrant terminal 3. The control unit 31 comprises a processor 32, a ROM 33, and a RAM 34. Note that the hardware configurations of the control unit 31, communication unit 35, and memory unit 36 ​​are substantially the same as the control unit 21, communication unit 25, and memory unit 26 in the human resources management device 2 described below, and therefore will not be described here.

[0016] The input unit 37 may be included in the housing of the registrant terminal 3, or may be externally attached. For example, the input unit 37 may be integrated with the display unit 38 and implemented as a touch panel. Of course, switch buttons, a mouse, a QWERTY keyboard, or the like may be used instead of a touch panel. That is, the input unit 37 accepts operational inputs made by the registrant. The inputs are transferred as command signals to the control unit 31 via a communication bus, and the control unit 31 can execute predetermined control or calculations as necessary.

[0017] Display unit 38 may be, for example, included in the housing of registrant terminal 3, or may be externally attached. Display unit 38 displays a graphical user interface (GUI) screen that can be operated by the registrant. This is preferably implemented by using display devices such as a CRT display, liquid crystal display, organic EL display, and plasma display, each selected depending on the type of registrant terminal 3.

[0018] 1.3 Personnel Management Device 2 The human resource management device 2 is a computer such as a server. For example, the human resource management device 2 is managed by an administrator who is in a position to manage members of an organization. The human resource management device 2 stores the material information acquired from the registrant terminal 3 in a member database (not shown) of the storage unit 26, which will be described later, or in an external storage area accessible by the human resource management device 2. The human resource management device 2 then acquires first output information related to members of the organization from the material information using a trained model, and manages the first output information. The trained model will be described in detail later.

[0019] 3 is a block diagram showing the hardware configuration of the human resource management device 2 according to embodiment 1. The human resource management device 2 includes a control unit 21, a communication unit 25, a storage unit 26, and an input unit 27, and these components are electrically connected within the human resource management device 2 via a communication bus.

[0020] The control unit 21 includes a processor 22, a ROM 23, and a RAM 24. The processor 22 executes programs stored in the ROM 23 or the storage unit 26 to realize various functions. Each function will be described in detail in the next section. The ROM 23 is a non-transitory storage medium such as a read-only memory (ROM), and pre-stores various programs or data for the processor 22 to perform various controls or calculations. The RAM 24 is a volatile storage medium such as a random access memory (RAM), and is a storage area for temporarily holding information when the processor 22 is operating. Areas for performing various processes according to this embodiment can be secured in the RAM 24.

[0021] The communication unit 25 is a communication interface for connecting to the network NW. The memory unit 26 is a storage device including a non-transitory computer-readable medium or a tangible storage medium. The hardware configuration of the input unit 27 is substantially the same as that of the above-mentioned input unit 37. Note that the input unit 27 is not essential for the human resource management device 2.

[0022] 2. Functional configuration This section describes the functional configuration of the first embodiment, in particular the functional configuration of the human resource management device 2. The human resource management method described below is executed by the human resource management device 2.

[0023] 4 is a block diagram showing the functional configuration of the human resource management device 2 according to embodiment 1. The control unit 21 realizes the functions of an acquisition unit 211, a generation unit 212, and a determination unit 213 by executing a program.

[0024] In the acquisition step, the acquisition unit 211 acquires one or more pieces of material information related to a target member (hereinafter referred to as a target member). For example, the acquisition unit 211 reads one or more pieces of material information related to the target member from the member database of the storage unit 26 or an external storage area.

[0025] In the generation step, the generation unit 212 generates first output information from one or more pieces of material information using a trained model. The trained model is a model that receives input information including at least a portion of the material information as input and outputs first output information. The first output information is different from the input information but is configured to include at least a portion of the material information. In other words, the first output information is newly constructed information that includes at least information about the target member. The input information and the first output information will be described in detail later.

[0026] In the determination step, the determination unit 213 determines attributes of the generated first output information based on attributes of one or more pieces of material information. Attributes include the range of users who can use the information, the time when the information was created, the expiration date of the information, or categories that classify the characteristics of the information from a predetermined perspective. Here, use of information refers to at least one of viewing, editing, publishing, sharing, and processing of the information.

[0027] 3. Operation of the personnel management device 2 In Section 3, the flow of the human resource management method executed by the human resource management device 2 will be described with reference to the drawings.

[0028] 3.1 Overview of Human Resource Management Methods 5 is a flowchart showing the flow of the human resource management method according to the first embodiment. The order of the processes can be changed as appropriate, and multiple processes may be executed simultaneously. Also, some processes may be omitted. The same applies to the subsequent flowcharts.

[0029] First, the acquisition unit 211 acquires one or more pieces of material information related to the target member (S11: acquisition step). For example, the acquisition unit 211 reads out one or more pieces of material information from the member database of the storage unit 26 or an external storage area. Next, the generation unit 212 inputs input information including the acquired one or more pieces of material information to the trained model (S12). Next, the generation unit 212 acquires first output information from the trained model (S13). Next, the determination unit 213 determines attributes of the first output information based on the attributes of the one or more pieces of material information (S14). Next, the determination unit 213 stores information in the storage unit 26 that associates the first output information acquired in S13 with the attributes determined in S14 (S15).

[0030] [effect] Until now, it has been necessary to properly manage trained models, particularly output information output from generation AI, by appropriately setting the range of available users and expiration date, or by appropriately assigning categories that represent the characteristics of the information. However, manually performing these tasks each time a trained model is generated is time-consuming. However, according to embodiment 1, the human resources management device 2 automatically assigns attributes to the first output information output from the trained model based on the attributes of the material information. This facilitates management of trained models, particularly output information output from generation AI.

[0031] 3.2 Details of human resource management methods Next, the above-mentioned personnel management method will be described in detail.

[0032] [Material information] Material information about members is personnel information about the members. The material information may include at least one of career information indicating the career history of the members of the organization, skill information indicating the skills of the members, and goal information indicating the goals of the members. For example, the material information may include the member's basic personal information, work history, transfer history, skills, goals set by the members within the organization, information about the members before they joined the organization (information at the time of hiring, such as pre-employment test results), evaluations of the members by superiors, engagement, aptitude test results, information about inner characteristics, attendance information, or interview history. The basic personal information may include the member's name, gender, age, employee code, contact information, affiliation, employment status, work location, or current job.

[0033] [Trained model] A trained model is a model that takes input information as input and outputs first output information. For example, a trained model is a large language model (LLM), which is an example of generative AI (artificial intelligence). The trained model is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ) or a dedicated model tuned for the human resource management system 1. In the latter case, the trained model may be trained to take input information as input and output first output information using a training dataset including training prompts containing learning material information and first output information that is the correct answer. Training may be performed in the control unit 21 of the human resource management device 2 or by an external device (not shown).

[0034] [Input information] The input information input to the trained model is configured to include at least a portion of the material information. For example, the input information is a prompt including at least a portion of the material information. The prompt may include not only the material information but also other information. Examples of other information include a condition on the data size (e.g., number of characters) of the first output information, a condition on the absolute or relative importance of each type of material information (e.g., prioritizing work history over skills), a condition on characteristics desired for the first output information (e.g., conciseness or understandability), or a combination thereof. By specifying a condition on a particularly small number of characters or conciseness, the generation unit 212 can generate, as the first output information in S13, a summary text summarized by combining one piece of material information or multiple pieces of material information, or a text newly constructed from these. When the amount of data in the material information is large or when the material information used for the input information is distributed across multiple tables, integrating the material information into a single concise piece of information makes it easier to grasp and manage the characteristics of the members. Note that instead of including the other information in the prompt, the parameters of the trained model may be adjusted through training to satisfy the above conditions.

[0035] [First output information] The first output information of the trained model is information about the characteristics of the members. In particular, the first output information may be information about the members' careers, skills, or inner characteristics. The first output information is configured to include at least a portion of the material information, but is information different from the input information. For example, the first output information may be a summary text of the material information, or a text newly constructed by combining information (keywords, phrases, etc.) contained in the material information. The first output information may be a summary text of one piece of material information, or a text newly constructed by combining information contained in one piece of material information. However, the first output information may instead be a text summarized by concatenating multiple pieces of material information, or a text newly constructed by combining information contained in each piece of material information. Furthermore, the first output information is not limited to text, and may be in other data formats such as numbers, vectors, or matrices.

[0036] [Attribute determination process: S14] In S14, the determination unit 213 determines the attribute of the first output information based on the attributes of one or more pieces of material information. At this time, the determination unit 213 may generate an output attribute table T3 using the material attribute table T1 and the material output table T2 stored in the storage unit 26.

[0037] (Material attribute table T1) FIG. 6 is a diagram showing an example of the data structure of the material attribute table T1. The material attribute table T1 is a table showing the correspondence between material information and attributes. For example, the material attribute table T1 associates material information T11 with attributes T12. More specifically, as shown in FIG. 6, the material attribute table T1 shows which attribute out of multiple types of attributes T12 (groups a to c) each of multiple types of material information T11 (materials A to F) has. Note that one type of material information T11 may have only one type of attribute T12, or may have multiple types of attributes T12.

[0038] In the example of FIG. 6, materials A to C have the attribute of group a, material D has the attributes of groups a and b, material E has the attributes of groups a, b and c, and material F has the attribute of group c.

[0039] (Material output table T2) 7 is a diagram showing an example of the data structure of the material output table T2. The material output table T2 is a table showing the correspondence between material information and first output information. For example, the material output table T2 associates material information T21 with first output information T22. More specifically, as shown in FIG. 7, the material output table T2 shows combinations of material information T11 used as input information when each piece of first output information T22 (output A, B, or C) is generated.

[0040] In the example of Fig. 7, output A is generated based on materials A to E, output B is generated based on materials D and E, output C is generated based on materials E and F, and output D is generated based on materials A and F. Note that Fig. 7 shows an example in which the first output information T22 is generated from input information using multiple types of material information T21, but it may also be generated from input information using only one type of material information T21.

[0041] (Output attribute table T3) 8 is a diagram showing an example of the data structure of the output attribute table T3. The output attribute table T3 is a table showing the correspondence between first output information and attributes. For example, the output attribute table T3 associates first output information T31 with attributes T32. More specifically, as shown in FIG. 8, the output attribute table T3 shows which attribute among multiple types of attributes T32 (groups a to d) each of multiple types of first output information T31 (output A to output D) has. Note that one type of first output information T31 may have only one type of attribute T32, or may have multiple types of attributes T32.

[0042] Here, for each piece of first output information, the generation unit 212 identifies the underlying material information from the material output table T2, and identifies the attributes of the identified material information from the material attribute table T1. Then, the generation unit 212 determines the attributes of the first output information based on the attributes of the underlying material information.

[0043] For example, when the first output information is based on a single piece of material information or on multiple pieces of material information with the same attribute, the generation unit 212 may determine the same attribute as the attribute of the material information as the attribute of the first output information. As an example, since output A is generated based on materials A to E, the attribute of output A may be group a, which is the attribute of materials A to E.

[0044] Furthermore, for example, when the first output information is based on a plurality of pieces of material information with different attributes, the generation unit 212 may determine the range (AND) in which the attributes of the plurality of pieces of material information overlap as the attribute of the first output information. As an example, output C is generated based on materials E and F, and material E has attributes of groups a to c, and material F has attributes of group c. In this case, the attribute of output C may be group c, which is an overlapping attribute.

[0045] Furthermore, for example, when the first output information is based on a plurality of pieces of material information with different attributes, the generation unit 212 may determine a range (OR) that includes some or all of the attributes of the plurality of pieces of material information as the attribute of the first output information. As an example, output D is generated based on materials A and F, and material A has an attribute of group a, and material F has an attribute of group c. In this case, the attribute of output D may be group d that includes group a and group c.

[0046] The rules for determining the attributes of the first output information are not limited to the above-described method and may be flexibly set as long as they are based on the attributes of the underlying material information.

[0047] Furthermore, determining the attributes may mean determining the attributes regardless of the registrant's operation, or determining the attributes after proposing the attributes to the registrant and obtaining their consent. The determined attributes may be unable to be modified, or may be modifiable by the registrant.

[0048] <Embodiment 2> Next, a second embodiment of the present disclosure will be described. The second embodiment is a specific example of the first embodiment. In the second embodiment, the attribute of the material information is the range of users who can use the material information (hereinafter referred to as the usable range), particularly the range of users who can view the material information (hereinafter referred to as the viewable range). The attribute of the first output information is the usable range of the first output information, particularly the viewable range. That is, the human resources management device 2 automatically determines the usable range of the material information based on the usable range of the first output information. This prevents a user who is not originally able to use the material information from being able to use the output of the trained model. It also eliminates the need to manually set the usable range of the output of the trained model.

[0049] The human resources management system according to the second embodiment is used to manage information relating to employees of a company, which are an example of members of an organization, and the first output information is employee characteristic text that indicates text that expresses the characteristics of the employee.

[0050] 4. Overall structure 4.1 Human Resource Management System 1a 9 is a block diagram showing the overall configuration of a human resources management system 1a according to embodiment 2. The human resources management system 1a is basically similar in configuration and function to the human resources management system 1, but differs from the human resources management system 1 in that it includes a human resources management device 2a instead of the human resources management device 2, and a viewer terminal 4.

[0051] 4.2 Human resources management device 2a The human resource management device 2a is basically similar in configuration and function to the human resource management device 2, but differs from the human resource management device 2 in that it presents first output information, which is information about human resources, to the viewer terminal 4. Specifically, the human resource management device 2a presents employee characteristic text according to the viewer's viewing authority to the viewer terminal 4. For example, when the human resource management device 2a receives a search request related to employees from the viewer terminal 4, it presents employee characteristic text that matches the search conditions and according to the viewer's viewing authority to the viewer terminal 4 as a search result. Note that a viewer having viewing authority for the first output information means that the viewer belongs to the viewable range of the employee characteristic text.

[0052] 4.3 Viewer terminal 4 The viewer terminal 4 is a computer terminal operated by the viewer. For example, the viewer may be an employee, a human resources officer, or a general affairs officer belonging to a company, or a person in a position to manage employees. For example, the viewer terminal 4 may be a smartphone, a tablet terminal, a personal computer, or other communication terminal capable of input and output. The hardware configuration of the viewer terminal 4 is substantially the same as that of the registrant terminal 3, so a description thereof will be omitted. The viewer terminal 4 receives employee characteristic text according to the viewer's viewing authority from the human resources management device 2a and displays it on a display unit (not shown).

[0053] 5. Functional Configuration 10 is a block diagram showing the functional configuration of a human resources management device 2a according to embodiment 2. A control unit 21a executes a program to realize the functions of an acquisition unit 211a, a generation unit 212a, a determination unit 213a, a search processing unit 214a, and a display control unit 215a.

[0054] The acquisition unit 211a, the generation unit 212a, and the determination unit 213a have basically the same functions as the acquisition unit 211, the generation unit 212, and the determination unit 213 of embodiment 1. However, the material information is information about employees, and the first output information is employee characteristic text.

[0055] The search processing unit 214a executes the following search processing steps. First, in response to receiving a search request from the viewer terminal 4, the search processing unit 214a acquires employee characteristic text that corresponds to an employee and that the viewer has viewing authority for. For example, the search processing unit 214a acquires, for each of all employees, employee characteristic text that corresponds to that employee and that the viewer has viewing authority for. Then, the search processing unit 214a generates information to be presented to the viewer as search results based on the acquired employee characteristic text and the search conditions included in the search request. In the second embodiment, the search conditions are search conditions related to the desired personnel or job type, but are not limited to this.

[0056] The display control unit 215a controls the various processing results executed by the control unit 21 so that they can be displayed on the display unit of the operator's terminal, such as the registrant terminal 3 or the viewer terminal 4. The above control may involve the display control unit 214 transmitting images showing the various processing results to the operator's terminal, or transmitting information necessary for generating images showing the various processing results on the operator's terminal to the operator's terminal. The operator's terminal that has received the necessary information displays the images on its display unit. For example, the display control unit 215a controls the search results generated by the search processing unit 214a so that they can be displayed on the display unit of the viewer terminal 4. In other words, the display control unit 215a causes the viewer terminal 4 to display information generated based on the search conditions and employee characteristic text included in the search request.

[0057] 6. Operation of the personnel management device 2a In this section, the flow of a human resource management method executed by the human resource management device 2a will be described with reference to the drawings. In the second embodiment, the human resource management method includes an attribute determination method and a search processing method.

[0058] 6.1 Overview of Attribute Determination Methods 11 is a flowchart showing the flow of the attribute determination method according to the second embodiment. First, the acquisition unit 211a receives and sets a group indicating a viewable range (viewable group) for each piece of material information from the registrant terminal 3 (S21). The viewable group is an example of the attribute of the first embodiment. The viewable group is determined in accordance with the user attribute. Next, the acquisition unit 211a receives designation of the user attribute of the viewer from the registrant terminal 3 (S22).

[0059] Next, the acquisition unit 211a identifies and acquires one or more pieces of material information corresponding to the specified user attributes, which is material information of the target employee who is the employee for whom employee characteristic text is to be generated (S23). Next, the generation unit 212a inputs a prompt including the one or more pieces of material information into the first trained model (S24). The first trained model is an example of the trained model of embodiment 1. Next, the generation unit 212a acquires employee characteristic text representing the characteristics of the target employee from the first trained model (S25).

[0060] Next, the determination unit 213a determines the viewable group of the employee characteristic text based on the viewable group of the material information used in S24 (S26). Next, the determination unit 213a stores information associating the employee characteristic text with the viewable group determined in S26 in the storage unit 26 (S27). Note that S23 to S27 in Fig. 11 correspond to S11 to S15 in Fig. 5, respectively.

[0061] In this way, the second embodiment can also achieve the same effects as the first embodiment.

[0062] 6.2 Details of attribute determination method Next, the above-mentioned attribute determination method will be described in detail.

[0063] [Setting the group that can view material information: S21] In S21, the acquisition unit 211a sets a group that can view the material information. At this time, the display control unit 215a may cause the display unit of the registrant terminal 3 to display a screen 6 for setting a group that can view the material information.

[0064] 12 is a diagram showing an example of screen 6. Screen 6 includes a material attribute table 60 for setting a viewable group for each piece of material information. The material attribute table 60 includes material information 61, a corresponding viewable group 62, and an input area 63 for inputting the correspondence between them.

[0065] The material information 61 shown in Fig. 12 indicates the identification information of the material information, specifically, the name. In Fig. 12, the material information includes "basic personal information," "transfer history," "career history," "career design sheet," "skills," "pre-employment test results," and "evaluation results." Each piece of material information may be stored in a different storage area (e.g., a table).

[0066] The viewable group 62 indicates the identification information of the viewable group, specifically the name. Here, the viewable group of material information is determined in correspondence with the user attributes of the viewer, particularly the viewer's job position. In Figure 12, three types of viewable groups are shown: "Superior," "Regular employee," and "Self." "Superior" indicates a group of users who hold the position of superior to the target employee. "Self" indicates a user who is the target employee himself. "Regular employee" indicates a group of users who are neither "Superior" nor "Self," for example, a group of colleagues. By simplifying the types of viewable range in this way, it is possible to make it easier to manage the viewable range.

[0067] The input area 63 is an area for inputting information to associate each piece of material information 61 with each viewable group 62. In the example shown in Fig. 12, "Basic personal information" can be viewed by users belonging to "Superior," "Regular employee," and "Self." Furthermore, "Evaluation results" can be viewed by users belonging to "Superior" and "Self," but cannot be viewed by users belonging to "Regular employee."

[0068] The screen 6 may have an operation area 64 for registering the material attribute table 60 reflecting the input. When the registrant clicks the operation area 64, the acquisition unit 211a registers the material attribute table 60 as the material attribute table T1. Then, the display control unit 215a transitions the display screen from the screen 6 to the screen 7.

[0069] 13 is a diagram showing an example of screen 7. Screen 7 is a screen for setting raw material information used as input information when generating employee characteristic text. Screen 7 includes a raw material output table 70. The raw material output table 70 includes raw material information 61, employee characteristic text 72, and the correspondence between them.

[0070] The employee characteristic text 72 indicates the identification information of the employee characteristic text, specifically the name. In FIG. 13, the same number of types of employee characteristic text are prepared as the types of viewable groups of material information. For example, the employee characteristic text 72 shows three types of employee characteristic text: "for supervisor," "for general employee," and "for the employee." "For supervisor" indicates employee characteristic text intended for a user who holds the position of supervisor of the target employee. "For the employee" indicates employee characteristic text intended for the target employee himself. "For general employee" indicates employee characteristic text intended for a user who is neither a "supervisor" nor "the employee," for example, a colleague.

[0071] The display control unit 215a may display, for the employee characteristic text "for supervisors," the material information associated with the viewable group "supervisor" registered on the screen 6 as corresponding material information. Note that the screen 7 may include an operation area 74 for correcting the correspondence. When the registrant clicks on the operation area 74, the display control unit 215a may cause the registrant terminal 3 to display a screen that accepts corrections to the correspondence between the material information and the viewable group or the correspondence between the material information and the employee attribute text.

[0072] The screen 7 may have an operation area 75 for registering the material output table 70. When the registrant clicks on the operation area 75, the acquisition unit 211a registers the material output table 70 as the material output table T2.

[0073] [Determining the viewable group of employee characteristic text: S26] In S26, the determination unit 213a may determine the same range as the viewable groups of the material information as the viewable groups of the employee characteristic text. When the display control unit 215a has determined the viewable groups of the employee characteristic text, the display control unit 215a may display a screen 8 indicating the viewable groups on the display unit of the registrant terminal 3.

[0074] FIG. 14 is a diagram showing an example of screen 8. Screen 8 includes an output attribute table 80. The output attribute table 80 includes employee characteristic text 72, viewable groups 82, and the correspondence between them. In the example shown in FIG. 14, the viewable groups of the employee characteristic text are determined in correspondence with the user attributes of the viewer, particularly the viewer's job position, in the same way as the viewable groups of the material information. FIG. 14 shows three viewable groups: "Superior," "Regular Employee," and "Self." The employee characteristic text "for Supervisor" can be viewed by users belonging to the "Superior" category. The employee characteristic text "for Self" can be viewed by users belonging to the "Self" category. The employee characteristic text "for Regular Employee" can be viewed by users who are neither "Superior" nor "Self," such as coworkers.

[0075] The screen 8 may include an operation area 84 for correcting the correspondence. When the registrant clicks on the operation area 84, the display control unit 215a may cause the registrant terminal 3 to display a screen that accepts corrections to the correspondence between the employee characteristic text and the viewable group. The control unit 21a can ensure flexibility in determining attributes by automatically determining the viewable group for the employee characteristic text while leaving room for later correction.

[0076] The screen 8 may have an operation area 85 for registering the output attribute table 80. When the registrant clicks on the operation area 85, the determination unit 213a registers the output attribute table 80 as the output attribute table T3.

[0077] In the attribute determination method of the second embodiment, the display control of the screens 7 and 8 may be omitted.

[0078] 6.3 Overview of search processing methods FIG. 15 is a flowchart showing the flow of the search processing method according to the second embodiment. First, when the control unit 21a receives a search request related to employees from the viewer terminal 4 (S31), the search processing unit 214a generates search text from the search criteria included in the search request (S32). Next, the search processing unit 214a identifies the user attributes of the viewer and identifies a viewable group based on the user attributes (S33). Next, the search processing unit 214a uses the output attribute table 80 to acquire, for each employee, employee characteristic text included in the viewable group corresponding to the user attributes of the viewer, i.e., employee characteristic text viewable by the viewer (S34). Next, the search processing unit 214a matches the employee characteristic text with the search text for each employee and calculates the degree of match (S35). Next, the search processing unit 214a identifies employee candidates to be presented to the viewer based on the degree of match (S36). Next, the search processing unit 214a generates information related to the employee candidates to be presented to the viewer (S37). Then, the display control unit 215a displays the generated information about the employee candidates on the display unit of the viewer terminal 4 (S38).

[0079] [effect] In this way, the control unit 21 uses the newly constructed employee characteristic text by integrating or summarizing the material information for employee search, which makes the search process easier even when the amount of material information data is large or when the material information is distributed across multiple tables.

[0080] 6.4 Details of search processing method Next, the above-mentioned search processing method will be described in detail.

[0081] [Search request received: S31] In S31, the control unit 21a may display an input screen for searching on the display unit of the viewer terminal 4, and may receive a search request including search criteria via the input screen. There may be two ways for the viewer to input search criteria. One is to input search criteria in free text format, and the other is to input search criteria in a selection format.

[0082] Fig. 16 is a diagram showing an example of screen 9, which is one of the input screens for searching. Screen 9 includes an operation area 91 for inputting in free text format and an operation area 92 for inputting in a selection format. In other words, the viewer can select the search method between free text format and selection format. When the viewer clicks on operation area 91 in Fig. 16, control unit 21a transitions the display screen from screen 9 to screen 10.

[0083] FIG. 17 is a diagram showing an example of screen 10. Screen 10 is a screen including an input area 101 for inputting characteristics of employees to be searched for in free text format as search conditions. In the example shown in this figure, the text "has leadership experience and is overseas-oriented" has been input into input area 101. In response to this input, control unit 21a receives a search request.

[0084] In FIG. 16, when the viewer clicks on the operation area 92, the control unit 21a changes the display screen from the screen 9 to the screen 11.

[0085] FIG. 18 is a diagram showing an example of screen 11. Screen 11 is a screen including an operation area 111 for inputting, in a selection format, characteristics of the employee to be searched for as search conditions. In the example shown in this figure, the items "Leadership" and "Overseas experience" are selected in operation area 111. In response to this selection, control unit 21a receives a search request.

[0086] [Generating search text: S32] In S32, the search processing unit 214a generates search text based on the search criteria included in the search request. The search text may be generated using a method substantially similar to the method for generating employee characteristic text. That is, the search processing unit 214a may generate search text from the input search criteria using a second trained model. The second trained model is a model trained to output search text using input information including at least a portion of the search criteria as input. The second trained model is not limited to, but may be an LLM. Note that if the search criteria are input in free text format, the search text generation step may be omitted. That is, the search text may be the input text.

[0087] [Identifying the viewable group of the viewer and obtaining the employee characteristic text: S33~S34] In S33, the search processing unit 214a identifies the user attributes of the viewer and identifies a viewable group based on the user attributes. At this time, the search processing unit 214a may identify the viewable group by identifying the job position relationship between each employee and the viewer for each employee. Then, in S34, the search processing unit 214a uses the output attribute table 80 to acquire, for each employee, employee characteristic text that is viewable by the viewable group to which the viewer belongs. At this time, if employee characteristic text for each group is prepared in advance, the search processing unit 214a may acquire the employee characteristic text by selecting the employee characteristic text corresponding to the relationship (i.e., group) from among multiple employee characteristic texts corresponding to the employee. Note that the employee characteristic text may be generated in S34. That is, in S34, the attribute determination process shown in FIG. 11 may be performed.

[0088] [Matching process: S35~S36] In S35, the search processing unit 214a matches the employee characteristic text with the search text for each employee and calculates the degree of match. The degree of match corresponds to the degree of match between the employee and the search criteria corresponding to the search text. Here, "match" indicates that there is a correlation between the two, and the degree of match indicates the degree of correlation. For example, "match" not only includes a state in which the meaning of the entire text or words or phrases contained therein is similar between the texts, but also a state in which the meanings of the texts are not similar but are recognized to be correlated based on past experience.

[0089] The calculation of the match degree can use a known method for analyzing correlations. Examples of analysis methods include co-occurrence analysis using a corpus or dataset, extraction of similar concepts through clustering, and regression analysis using a correlation matrix. The analysis may use words, phrases, or the entire text contained in the text, or features extracted from these. The search processing unit 214a may then calculate the match degree based on the correlation. The search processing unit 214a may also calculate the match degree using a trained model for calculating the match degree. For example, the trained model may be trained to output the match degree using features extracted from the training employee characteristic text and the training search text as inputs.

[0090] Next, in S36, the search processing unit 214a identifies employees to be presented to the viewer as candidate employees based on the degree of match. For example, the search processing unit 214a may identify the top N employees (N is a natural number) in descending order of the degree of match as candidate employees. Alternatively, the search processing unit 214a may identify employees whose degree of match is equal to or greater than a predetermined threshold as candidate employees.

[0091] [Generation and display control of information about employee candidates: S37 to S38] In S37, the search processing unit 214a generates information on candidate employees to be presented to the viewer as a search result. Then, in S38, the display control unit 215a displays the generated information on candidate employees on the display unit of the viewer terminal 4.

[0092] 19 is a diagram showing an example of the screen 12. The screen 12 includes candidate information 121 and 122 as search results. Each of the candidate information 121 and 122 is information relating to an employee candidate. The candidate information 121, 122 includes basic personal information RG1, RG2 of the identified employee candidate, such as name, gender, age, affiliation, or current job. The candidate information 121, 122 also includes information DC1 indicating the reason for the match. The information DC1 may include at least a portion of the employee characteristic text used in the matching process, raw information used as the basis for generating the employee characteristic text, or a combination thereof. The information DC1 may also be text newly generated from the employee characteristic text used in the matching process. For example, the information DC1 may be text further summarized from the employee characteristic text used in the matching process using a first trained model, or may be text newly generated based on the employee characteristic text used in the matching process and the search text. For example, the search processing unit 214a may identify a portion of the employee characteristic text and the search text that has a particularly high degree of match and generate the information DC1 based on that portion. As an example, the search processing unit 214a may identify combinations between words or phrases contained in the employee characteristic text and words or phrases contained in the search text whose degree of match is above a predetermined threshold, and generate information DC1 based on those combinations.

[0093] The candidate information 121, 122 may further include information indicating the degree of match calculated in S35, for example, an image IM indicating the degree of match. The image IM is placed at a position corresponding to the personal basic information RG1, RG2 of the employee candidate. The length of the bar represented by the image IM indicates the degree of match. This allows the viewer to quantitatively and immediately grasp how well the employee candidate matches the search criteria. The image IM may also indicate the magnitude of the difference between the degree of match and the specified maximum score (100%).

[0094] In this figure, two pieces of candidate information 121, 122 are presented in descending order of match. The display order of the information about the employee candidates is determined by the display control unit 215a based on the match degree. However, this is not limited to this, and the order may be determined based on other information in addition to or instead of the match degree. This allows for different priorities to be placed between identifying employee candidates and determining their display order. For example, the display control unit 215a may prioritize suitability for the search criteria for employee candidates, while prioritizing the display order based on the likelihood of attracting the interest of viewers or employee candidates or the wishes of the employee candidates. In this case, the display control unit 215a may determine the display order based on other material information not used in the matching process. For example, if skill information and career information are used in the matching process, the display control unit 215a may determine the display order of the employee candidates based on the match degree and goal information. In this way, the control unit 21a selects employee candidates based on the current or past characteristics of the employee candidates, but presents the employee candidates to the viewer in a manner that is consistent with the employee candidates' future aspirations. This ensures suitability for the search conditions, while making it possible to present information that is likely to attract the viewer's interest or that meets the wishes of potential employees.

[0095] Furthermore, the display control unit 215a may determine the display order based on the degree of match as well as on the information of relatively high importance among the material information used in the matching process. For example, assume that skill information and career information are used in the matching process. In this case, if the career information is more important than the skill information, the display control unit 215a may determine the display order of the employee candidates based on the degree of match and the career information. This makes it possible to adjust the weighting between the selection of employee candidates and the determination of the display order.

[0096] The candidate information 121, 122 may include information DC2 newly generated from other material information not used in the matching process. The material information used in the matching process may be excluded from the material information used in the information DC2. For example, if skill information and career information are used in the matching process, the information DC2 may be text generated based on the goal information of the employee candidate.

[0097] 7. Variations Although the first and second embodiments have been described above, the first and second embodiments can be modified as follows.

[0098] For example, in the first and second embodiments, a series of processes are executed in the human resource management device 2, 2a, but some of the processes may be executed in an external device. For example, in the step of generating first output information (employee characteristic text in the second embodiment), the generation unit 212 of the human resource management device 2, 2a may generate a prompt including at least a part of the first input information and transmit the prompt to the external device via the network NW. In this case, the external device is a computer that stores a trained model such as an LLM (the first trained model in the second embodiment). The trained model is the same as in the first and second embodiments, and therefore its description is omitted. The generation unit 212, 212a may then receive, via the network NW, the first output information generated by the external device using the trained model. By executing part of the processes through an external service in this way, it is possible to obtain the same effects as in the first and second embodiments while reducing development resources.

[0099] In addition, for example, in the second embodiment, the raw information and the first output information are information about members of an organization. However, the raw information and the first output information may be information about a person who does not belong to the organization or a person who currently belongs to another organization, such as a job seeker.

[0100] In the second embodiment, the search processing unit 214a of the human resources management device 2a identifies the job position relationship between each of all employees in the organization and the viewer, acquires the employee characteristic text for each employee, and matches the search text with the search text in the search processing step. However, the range of employees to be searched (search range) is not limited to all employees, but may be a portion of the employees. For example, the search processing unit 214a may divide the search range in advance, sequentially identify the job position relationship between each employee included in each search range and the viewer, acquire the employee characteristic text for each employee, and perform matching. The search processing unit 214a may then terminate the search processing step when the number of matched employees is M or more (M is a natural number). Note that the search processing unit 214a acquires employee characteristic text for all employees, but may also perform matching for a portion of the employees. The search processing unit 214a may also set a search range in advance depending on the search conditions. For example, suppose a viewer inputs the text "personnel with leadership experience and an overseas orientation" along with the search condition "currently working in the human resources department." In this case, the search processing unit 214a may identify employees who are tagged as currently working in the human resources department from an employee database or the like, and set the employees as the search range.

[0101] In the first and second embodiments, the human resource management devices 2 and 2a are described as single devices. However, some or all of the components of the human resource management devices 2 and 2a may be implemented using multiple information processing devices, circuits, etc. In this case, the multiple information processing devices, circuits, etc. may be centrally or decentralized. For example, the information processing devices, circuits, etc. may be implemented as a client-server system, a cloud computing system, or the like, in a form in which each device is connected via a communication network. Furthermore, the functions of the human resource management devices 2 and 2a may be provided in the form of SaaS (Software as a Service).

[0102] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the invention.

[0103] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes. (Appendix 1) A human resources management system comprising one or more processors and a memory, the one or more processors are capable of performing the generating step and the determining step; The one or more processors: In the generating step, first output information is generated from one or more pieces of material information using a trained model; The trained model receives input information including at least a portion of material information as input, and outputs first output information including at least a portion of the material information; In the determining step, the attribute of the generated first output information is determined based on the attribute of the one or more pieces of material information. Human resource management system. (Appendix 2) the one or more pieces of material information include at least one of career information indicating the careers of members of the organization, skill information indicating the skills of the members, and goal information indicating the goals of the members; The first output information is information about the characteristics of the members. The human resource management system described in Appendix 1. (Appendix 3) the attribute of the one or more pieces of material information is a range of users who can use the one or more pieces of material information; the attribute of the first output information is a range of users who can use the first output information; The use is at least one of viewing, editing, sharing, and processing of the target information. A human resources management system as set forth in Appendix 1 or 2. (Appendix 4) The attributes of the one or more pieces of material information are determined in association with the user attributes of the viewer. The human resource management system described in Appendix 3. (Appendix 5) The one or more processors are further capable of performing a first display control step; In the first display control step, a screen for accepting corrections to the attributes of the one or more pieces of material information or the attributes of the determined first output information is displayed on a registrant terminal operated by a registrant. 5. A human resources management system according to any one of appendices 1 to 4. (Appendix 6) The input information further includes a condition for the data size of the first output information, a condition for the absolute or relative importance of each type of the one or more pieces of material information, and a condition for characteristics required for the first output information. 6. A human resources management system according to any one of appendices 1 to 5. (Appendix 7) the one or more processors are further capable of executing a search processing step and a second display control step; In the search processing step, in response to receiving a search request from a viewer terminal operated by a viewer, the first output information corresponding to the member and for which the viewer has a viewing authority is acquired; In the second display control step, information generated based on the search conditions included in the search request and the first output information is displayed on the viewer terminal. The human resource management system described in Appendix 2. (Appendix 8) The one or more processors: In the search processing step, for each member, a position relationship between the member and the viewer is identified, and the first output information corresponding to the position relationship is selected from a plurality of first output information corresponding to the member; In the second display control step, information generated based on the search conditions and the selected first output information is displayed on the viewer terminal. The human resource management system described in Appendix 7. (Appendix 9) 1. A workforce management method implemented by one or more processors, comprising: The human resource management method includes a generating step and a determining step, In the generating step, first output information is generated from one or more pieces of material information using a trained model; The trained model receives input information including at least a portion of material information as input, and outputs first output information including at least a portion of the material information; In the determining step, the attribute of the generated first output information is determined based on the attribute of the one or more pieces of material information. Human resource management methods. (Appendix 10) A program for operating a computer including one or more processors and a memory, the program causing the one or more processors to execute a generating step and a determining step; In the generating step, first output information is generated from one or more pieces of material information using a trained model; The trained model receives input information including at least a portion of material information as input, and outputs first output information including at least a portion of the material information; In the determining step, the attribute of the generated first output information is determined based on the attribute of the one or more pieces of material information. program. [Explanation of symbols]

[0104] 1,1a Human Resources Management System 2,2a Human resources management device 3. Registered terminal 4. Viewer terminal 6,7,8,9,10,11,12 screens 21 Control section 22 processors 23 ROM 24 RAM 25 Communications Department 26 Memory section 27 Input section 31 Control Unit 32 processors 33 ROM 34 RAM 35 Communications Department 36 Memory section 37 Input section 38 Display section 60 Material Attribute Table 61 Material Information 62,82 Viewable groups 63,101 input area 64,74,75,84,85,91,92,111 Operation area 70 Material Output Table 72 Employee Characteristics Text 80 Output Attribute Table 121,122 Candidate Information 211,211a Acquisition Department 212,212a generation part 213,213a Decision section 214a Search processing unit 215a Display control unit T1 Material Attribute Table T11, T21 material information T12,T32 attributes T2 Material Output Table T22, T31 1st output information T3 Output Attribute Table

Claims

1. A human resources management system comprising one or more processors and a memory, the one or more processors are capable of performing the generating step, the storing step, and the retrieval processing step; The one or more processors: In the generating step, first output information is acquired for each viewable group, the first output information being output from the large-scale language model in response to inputting input information into the large-scale language model; the input information includes a plurality of pieces of material information corresponding to the viewable group; the material information is information about members of an organization; the first output information includes at least a part of the plurality of pieces of material information; the viewable group is a user group indicating a viewable range of the generated first output information, and there are a predetermined number of types of viewable groups; In the storing step, the first output information is stored in association with the viewable group for each viewable group; In the search processing step, in response to a search request from a viewer, members are searched for using first output information corresponding to a viewable group to which the viewer belongs. Human resource management system.

2. the plurality of pieces of material information include at least one of career information indicating the careers of members of the organization, skill information indicating the skills of the members, and goal information indicating the goals of the members; The first output information is information about the characteristics of the members. The human resources management system according to claim 1 .

3. The one or more processors are further capable of performing a first display control step; In the first display control step, a screen for accepting corrections to the viewable group of the first output information is displayed on a registrant terminal operated by the registrant. The human resources management system according to claim 1 .

4. The input information further includes a condition regarding the data size of the first output information, a condition regarding the absolute or relative importance of the material information, and a condition regarding characteristics required for the first output information. The human resources management system according to claim 1 .

5. The one or more processors are further capable of performing a second display control step; In the second display control step, information generated based on the search condition included in the search request and first output information corresponding to a viewable group to which the viewer belongs is displayed on a viewer terminal operated by the viewer. The human resources management system according to claim 1 .

6. The viewable group of the first output information of the member is determined in accordance with a job position relationship between the member and the viewer. The human resources management system according to claim 1 .

7. A human resources management system executes the steps set forth in claim 1. Human resource management methods.

8. A program for causing a computer having one or more processors and a memory to execute each step of claim 1.

Images