Business support device, business support method, and business support program

Abstract

This enables efficient work processes to be carried out in accordance with established procedures. [Solution] The business support device 100 receives prompts, including commands, from a large-scale language model trained using information about the organization and information about the behavior of users who are members of the organization, to generate information about suggestions for business procedures within the organization that correspond to the user's behavior, thereby generating information about suggestions for business procedures within the organization. The business support device 100 outputs the generated information about suggestions for business procedures within the organization to the user in a predetermined format.

Description

Technical Field

[0001] The present invention relates to a business support device, a business support method, and a business support program.

Background Art

[0002] In order to efficiently promote business activities in an organization, business procedures for each organization may be defined. In that case, employees belonging to the organization perform daily work according to the predetermined business procedures. Therefore, in the organization, in order to make employees and the like comply with the above business procedures, a "workflow system" or the like that defines the order of performing work based on a predetermined system may be used.

[0003] The above workflow system is a system in which an employee inputs information into the system according to a predetermined procedure, and the next step is presented based on the input information. For example, as a conventional technique for complying with the above business procedures, in a workflow system, when an input such as a document is made, a review / approval request notice or the like is sent to the reviewer and approver of the document, and a review / approval request notice is sent not for each document but in a batch to one reviewer or approver (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, conventional technologies can make it difficult to efficiently carry out tasks according to established procedures. For example, conventional technologies support employees in adhering to workflows by notifying reviewers and approvers, etc., according to a predetermined approval route for approval of entered documents. However, conventional technologies only manage workflows according to the approval route, and there are challenges in enabling employees to efficiently carry out tasks using the workflow management system. [Means for solving the problem]

[0006] Therefore, in order to solve the above-mentioned problems and achieve the objective, the business support device of the present invention is characterized by having a generation unit that generates information on suggestions for business procedures in an organization by receiving a prompt that includes a command to generate information on suggestions for business procedures in an organization in accordance with the user's behavior to a large-scale language model that has been learned using information on an organization and information on the behavior of users who are members of the organization, and output unit that outputs the generated information on suggestions for business procedures in an organization to the user in a predetermined format. [Effects of the Invention]

[0007] The present invention has the effect of enabling efficient work to be carried out in accordance with work procedures. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a diagram illustrating the overall process of generating behavioral suggestion information according to this embodiment. [Figure 2] Figure 2 shows the configuration of a business support device according to an embodiment. [Figure 3] Figure 3 is a table diagram showing an example of organizational information according to the embodiment. [Figure 4] Figure 4 is a table diagram showing an example of user behavior information according to the embodiment. [Figure 5]Figure 5 is a table diagram showing an example of training data according to the embodiment. [Figure 6] Figure 6 is a table diagram showing an example of checklist information according to the embodiment. [Figure 7] Figure 7 shows an example of self-learning according to the embodiment. [Figure 8] Figure 8 shows an example of system operation support according to the embodiment. [Figure 9] Figure 9 shows an example of user support based on the options according to the embodiment. [Figure 10] Figure 10 shows an example of checklist information generation according to the embodiment. [Figure 11] Figure 11 is a flowchart showing the processing performed by the business support device according to the embodiment. [Figure 12] Figure 12 shows an example of a computer that implements a business support device according to the embodiment. [Modes for carrying out the invention]

[0009] Hereinafter, embodiments for carrying out the present invention (hereinafter referred to as "embodiments") will be described with reference to the drawings. However, each embodiment is not limited to those described below.

[0010] <Overview> (background) In organizations, operational procedures are sometimes established for each department in order to promote efficient work and standardize operations. For example, for high-value purchases or business activities requiring management decisions, it is necessary to submit a proposal and obtain approval in advance.

[0011] When submitting a proposal for approval, a "workflow system" may be used to circulate the proposal to approvers sequentially and obtain approval from each approver. In such a workflow system, employees submitting the proposal enter information into the system according to predetermined procedures, and the next steps are presented based on the entered information.

[0012] As an example of a workflow system, there is known a reference technique that, when input such as a document is made, sends a review / approval request notice or the like to the reviewer and approver of the document, and sends the review / approval request notice in a batch rather than for each document to one reviewer or approver.

[0013] However, since the conventional reference technique is a technique that supports employees to proceed with their work while following the workflow along a predetermined approval route, there are problems in enabling employees to efficiently proceed with their work using the workflow management system. For example, in the reference technique, even when an employee does not know what information should be input to which item, it is difficult to provide appropriate support for the employee's troubles.

[0014] (Processing by Business Support Device 100) Therefore, the business support device 100 according to the present embodiment outputs to the user information suggesting the next action of the user, which is generated using information related to the organization including regulations, business procedures, past approval information, etc. in the organization and information related to the actions of users who are members of the organization.

[0015] Note that the above "information related to the organization" is information including information related to the regulations of the organization, information related to users belonging to the organization, the conversation logs of the users, and approval information related to the organization, and may hereinafter be referred to as "organization information". Also, the above "information related to the actions of the user" is information related to the history of the physical actions of the user including the logs of the systems operated by the user, the movement of the user's line of sight, the fluctuations in body temperature / sweating amount, the movement of the cursor operated by the user, etc., and may hereinafter be referred to as "user action information". Further, the above "information suggesting the next action of the user" is information related to suggestions of business procedures in the organization according to the actions of the user, and may hereinafter be referred to as "action suggestion information".

[0016] Here, the overall picture of the processing by the business support device 100 will be described. FIG. 1 is a diagram for explaining the overall picture of the generation processing of action suggestion information according to the embodiment. The business support device 100 shown in FIG. 1 is an example of a computer that provides a technology for realizing the information processing described below.

[0017] First, the business support device 100 uses, as prior knowledge, organization information including regulations, business procedures, past approval information, etc. about the target organization, and user behavior information ( (1-1) in FIG. 1) including system operation logs and history information of physical movements by users belonging to the target organization, and provides it to a predetermined large language model ( (1-2) in FIG. 1).

[0018] The above "provision of prior knowledge" means inputting predetermined information into the large language model in advance, and includes learning and tuning of the large language model, input of information into the large language model, insertion or substitution of information related to the prior knowledge into the prompt input into the large language model, etc.

[0019] Next, the business support device 100 inputs a generation command for action suggestion information ( (2-1) in FIG. 1) including information such as a suggestion of the next action (next action) that the user should take, into the large language model using a prompt ( (2-2) in FIG. 1) expressed in natural language text, and causes the large language model to generate action suggestion information ( (2-3) in FIG. 1).

[0020] The above "next action" includes, for example, information that requests additional input from a user who uses a system or the like, and suggestions for the user's actions that suggest what actions the user should take next.

[0021] The business support device 100 then outputs action suggestion information, such as the next action mentioned above, to the target user in a predetermined format (Figure 1 (3-1)). For example, if the user is stuck on a predetermined input screen because they don't know which item to enter or what information to enter, the business support device 100 displays information on the screen such as "Next, please enter XX in this item" (Figure 1 (3-2)).

[0022] In this way, the business support device 100 according to this embodiment can output action suggestion information generated based on organizational information and user behavior information. As a result, even if a user's work stops because they don't know what to do next, the business support device 100 can present the user with information suggesting the next action, thereby enabling them to proceed with their work efficiently in accordance with the work procedure.

[0023] <Description of Business Support Device 100> The configuration of the business support device 100 according to this embodiment will now be described. Figure 2 is a diagram showing the configuration of the business support device 100 according to this embodiment. As shown in Figure 2, the business support device 100 has a communication unit 110, a storage unit 120, and a control unit 130.

[0024] Although not shown in Figure 2, the business support device 100 may be equipped with an input unit such as a keyboard or mouse to receive input such as operations from administrators. Furthermore, the business support device 100 may be equipped with a display or the like to show acquired organizational information, user behavior information, generated behavioral suggestion information, etc., to administrators.

[0025] (Communications Department 110) The communication unit 110 performs data communication related to the input of organizational information, user behavior information, and commands for generating behavioral suggestion information, which are input by the person in charge or acquired from an external information processing device. The communication unit 110 also performs data communication related to the output of the generated behavioral suggestion information, etc.

[0026] The communication unit 110 is implemented using a NIC (Network Interface Card) or the like, and controls communication via telecommunication lines such as a LAN (Local Area Network) or the Internet. The communication unit 110 can be connected to the network via wired or wireless connection as needed, and can send and receive information bidirectionally with terminal devices, etc.

[0027] (Storage unit 120) The memory unit 120 stores data and programs used for various processes by the control unit 130, as well as various data acquired through the operation of the control unit 130. The memory unit 120 is implemented using semiconductor memory elements such as RAM (Random Access Memory) and flash memory, or storage devices such as hard disks and optical discs. As shown in Figure 2, the memory unit 120 includes an organization information DB 121, a user behavior information DB 122, a learning data DB 123, a checklist information DB 124, and a generation model DB 125.

[0028] (Organization information DB121) Organizational Information DB121 is a database that stores information about an organization (organizational information), including information about the organization's regulations, information about users belonging to the organization, user conversation logs, and approval information related to the organization. Specifically, Organizational Information DB121 stores information such as regulations concerning the organization (organizational regulations), information defining the procedures for operations within the organization (operational procedures), and log information of conversations between users belonging to the organization (conversation logs).

[0029] Here, an example of organizational information stored in the organizational information DB121 will be explained using Figure 3. Figure 3 is a table diagram showing an example of organizational information according to the embodiment. The organizational information DB121 associates information related to each item, such as "No," which is information that identifies individual data included in the organizational information, with "organizational identification information," "organizational regulations," "work procedures," and "conversation logs," and stores it in a table format, for example, as shown in Figure 3. The letters "A to D" written in each item of the table diagram shown in Figure 3 are legends for the information included in each item.

[0030] The "organizational identification information" mentioned above refers to information used to identify the target organization, and includes, for example, the organization name, organizational number, and other information combining text, numbers, symbols, etc., to identify the organization. "Organizational regulations" refers to regulations established within the organization, and includes, for example, information on rules that must be followed within the organization. "Work procedures" refers to information that defines the procedures for carrying out work within the organization. "Conversation logs" refer to conversations between users belonging to the organization, and include, for example, log information from phone calls, emails, chats, etc.

[0031] (User behavior information DB122) The User Behavior Information DB122 is a database that stores information about user behavior (user behavior information), including logs of systems operated by the user and information about the history of the user's physical movements. Specifically, the User Behavior Information DB122 stores information such as operation logs of systems operated by the user, and a history of changes in the user's physical state and physical movements (physical movement history).

[0032] Here, an example of user behavior information stored in the user behavior information DB122 will be explained using Figure 4. Figure 4 is a table diagram showing an example of user behavior information according to the embodiment. The user behavior information DB122 stores information related to each item, such as "No," which is information that identifies individual data included in the user behavior information, "user identification information," "operation log," and "physical movement history," in a table format as shown in Figure 4. The letters "E to G" written in each item of the table diagram shown in Figure 4 are legends for the information included in each item.

[0033] The "user identification information" mentioned above is information used to identify users belonging to the target organization, and includes, for example, usernames, user numbers, and other information combining text, numbers, symbols, etc., to identify the user, but with information that could identify the individual user excluded. The "operation log" is log information of systems operated by the user, and includes, for example, information such as the date of operation and the content of the operation. The "physical movement history" includes information about the history of the user's physical movements, including the user's eye movements, changes in body temperature / sweating, and the movement of the cursor operated by the user.

[0034] (Training data DB123) The learning data DB123 is a database that stores learning data, such as updated information on business procedures, which the learning unit 132 (described later) uses to pre-train a large-scale language model. Specifically, the learning data DB123 stores information such as updated information on business procedures (learning business procedures) associated with each organization.

[0035] Here, an example of training data stored in the training data DB123 will be explained using Figure 5. Figure 5 is a table diagram showing an example of training data according to the embodiment. The training data DB123 associates information related to each item, such as "No," which is information that identifies individual data included in the training data, with "organization identification information" and "training business procedure," and stores this information in a table format, for example, as shown in Figure 5. The letters "A" and "H" written in each item of the table diagram shown in Figure 5 are legends for the information included in each item.

[0036] The "organization identification information" mentioned above is the same information as the "organization identification information" included in the organization information shown in Figure 3. Furthermore, the "learning work procedures" are updated work procedures based on the difference between predetermined work procedures and the procedures performed by users who actually carry out the work, and are used for the self-learning of the large-scale language model by the learning unit 132 described later.

[0037] (Checklist Information DB124) The checklist information DB124 is a database that stores checklist information generated by the generation unit 135, which will be described later. Specifically, the checklist information DB124 stores information such as a first checklist in which predetermined confirmation items have been updated to be confirmed, and a second checklist in which the predetermined confirmation items included in the first checklist that were not yet confirmed have been updated to be confirmed.

[0038] Here, an example of checklist information stored in the checklist information DB124 will be explained using Figure 6. Figure 6 is a table diagram showing an example of checklist information according to the embodiment. The checklist information DB124 stores the information related to each item of the "first checklist" and the "second checklist" by associating it with "No," which is information that identifies individual data included in the checklist information, in a table format such as that shown in Figure 6. The letters "I and J" written in each item of the table diagram shown in Figure 6 are legends for the information included in each item.

[0039] The "First Checklist" described above is a checklist that includes items checked by the generation unit 135 for parts of the predetermined business-related checklist where it is possible to check for problems based on log information of the organization's systems, etc. The "Second Checklist" is a checklist that includes items checked by the generation unit 135 for checklist items other than those covered in the "First Checklist," based on the results of user interviews, etc.

[0040] (Generative model DB125) The generative model DB125 is a database that stores predetermined models used by the generation unit 135 (described later) to generate behavioral suggestion information. For example, the generative model DB125 can store large-scale language models as generative models.

[0041] Specifically, the business support device 100 according to this embodiment can use at least one of the following as a large-scale language model: "ChatGPT®", a large-scale language model possessing general-purpose knowledge, and "tsuzumi®", a predetermined large-scale language model on which adapter tuning is performed (see, for example, References 1 and 2).

[0042] (Reference 1):ChatGPT(OpenAI),<URL:https: / / openai.com / chatgpt> ,<Searched on October 7, 2021> (Reference 2): NTT version of large-scale language model "tsuzumi",<URL:https: / / www.rd.ntt / research / LLM_tsuzumi.html> ,<Searched on October 7, 2021>

[0043] Furthermore, the business support device 100 according to this embodiment can perform processing using a predetermined learning model that falls within the category of machine learning models, in addition to the aforementioned "ChatGPT®" and "tsuzumi®".

[0044] (Control unit 130) Now, let's return to Figure 2 and continue the explanation. The control unit 130 has an internal memory for temporarily storing programs and processing data that define various processing procedures of the business support device 100, and is realized by electronic circuits such as a CPU (Central Processing Unit) and an MPU (Micro Processing Unit), and integrated circuits such as an ASIC (Application Specific Integrated Circuit) and an FPGA (Field Programmable Gate Array). As shown in Figure 2, the control unit 130 has an acquisition unit 131, a learning unit 132, a detection unit 133, a reception unit 134, a generation unit 135, and an output unit 136.

[0045] (Acquisition part 131) The acquisition unit 131 acquires information used by the generation unit 135 (described later) to generate behavioral suggestion information from an external information processing device, etc., via the communication unit 110, etc.

[0046] For example, the acquisition unit 131 acquires organizational information from an information processing device that manages information related to the organization, including information such as the organization's regulations ("Organizational Regulations" in Figure 3), the procedures for operations within the organization ("Operational Procedures" in Figure 3), and the conversation logs of users belonging to the organization ("Conversation Logs" in Figure 3), and stores it as organizational information in the organizational information DB 121.

[0047] Furthermore, the acquisition unit 131 acquires the operation log of the system (the "operation log" in Figure 4) from the information processing device, etc., on which the system operated by the user is running, and stores it in the user behavior information DB 122 as user behavior information. In addition, the acquisition unit 131 acquires the user's physical movement history (the "physical movement history" in Figure 4) from the information processing device, etc., such as a camera that detects the user's body movements, a human presence sensor, a body temperature sensor, a sweat sensor, and an eye-tracking device, and stores it in the user behavior information DB 122 as user behavior information.

[0048] (Learning Section 132) The learning unit 132 trains a large-scale language model using updated organizational procedures based on the difference between the user's work procedures and the organizational procedures calculated using organizational information.

[0049] Specifically, the learning unit 132 updates the current work procedure based on the difference between the predetermined work procedure and the work procedure actually performed by the user, thereby generating a "new work procedure." Then, the learning unit 132 learns a large-scale language model using the updated new work procedure, thereby performing self-learning to always generate action suggestion information using the latest work procedure.

[0050] The "user's work procedures" described above refer to the procedures performed by the user, identified using information about the user's actions, including at least one of the following: logs of systems operated by the user, and / or a history of the user's physical movements. Furthermore, the "updated organizational procedures" described above refer to the organizational procedures that have been updated based on the difference between the organizational procedures and the organizational procedures calculated using organizational information, which includes at least one of the following: information about organizational regulations, information about users belonging to the organization, user conversation logs, and / or organizational approval information.

[0051] (Detection unit 133) The detection unit 133 detects a predetermined action by the user based on the operation log of the system operated by the user. Then, triggered by the detection of the predetermined action by the user, the detection unit 133 transmits a command to the generation unit 135 to generate action suggestion information.

[0052] For example, if the detection unit 133 detects an error in input items or a halt in input work when a user belonging to the organization operates a system related to the organization via a terminal device, it sends a command to generate a next action to the generation unit 135.

[0053] (Reception desk 134) The reception unit 134 accepts answers to questions entered by users regarding points they are unsure of, as well as answers based on the provided options.

[0054] (Generation unit 135) The generation unit 135 generates behavioral suggestion information by inputting prompts that include commands to a large-scale language model, which has been trained using organizational information and user behavior information, to generate behavioral suggestion information corresponding to the user's actions. For example, the generation unit 135 generates next action and checklist information as behavioral suggestion information. A specific example of the behavioral suggestion information generation process by the generation unit 135 will be explained in the following sections.

[0055] (Output section 136) The output unit 136 outputs the generated action suggestion information to the user in a predetermined format. For example, the output unit 136 displays the next action generated by the generation unit 135 superimposed on the screen of the system used by the user. The output unit 136 also outputs checklist information, including the first checklist and the second checklist generated by the generation unit 135, to the target user. A specific example of the output processing of action guidance information by the output unit 136 in a predetermined format will be explained in the following sections.

[0056] (An example of processing) From here, using Figures 7 to 10, we will explain an example of the generation and output processing of behavioral suggestion information by the business support device 100. The first example shown below is an example of the business support device 100 self-learning the learning model used to generate behavioral suggestion information. The second example is an example of the business support device 100 outputting behavioral suggestion information to a user operating a predetermined system. The third example is an example of the business support device 100 automatically generating a checklist related to business procedures in a predetermined organization.

[0057] (Example 1) First, as a first example, we will explain with reference to Figure 7 an example in which the business support device 100 updates the organization's business procedures using organizational information and user behavior information. Figure 7 is a diagram showing an example of self-learning according to the embodiment. In this embodiment, "self-learning" means that the business support device 100 automatically learns a large-scale language model using information on business procedures that has been updated based on organizational information and user behavior information acquired by the business support device 100.

[0058] In the first example, the business support device 100 (learning unit) inputs commands to update business procedures using organizational information and user behavior information, and commands to train the large-scale language model using the updated business procedures, to the large-scale language model using prompts (Figure 7(1)), thereby performing self-learning of the large-scale language model.

[0059] For example, the business support device 100 (learning unit) learns a large-scale language model using prompts that include "<role>", "<constraints>", "<commands>", etc., as shown in (1) of Figure 7.

[0060] The "<Role>" mentioned above is a text that defines the role of a person who embodies a specific role in the large-scale language model, thereby enabling the model to perform a predetermined learning process. As described above, the business support device 100 (learning unit) can use prompts that include "role definitions" to enable the large-scale language model to perform the desired learning process with high accuracy.

[0061] For example, the business support device 100 (learning unit) can assign the role of "monitoring the behavior of users belonging to an organization and creating optimal business procedures according to the users' behavior" to the large-scale language model by using prompts that have the "<role>" indicated as shown in (1-1) of Figure 7. As a result, the business support device 100 (learning unit) can generate more accurate business procedures by having the large-scale language model execute the process as an expert in creating or updating business procedures.

[0062] Furthermore, the "<Constraints>" mentioned above is text that describes predetermined constraints on the execution of processing when the large-scale language model generates information. The business support device 100 (learning unit) learns the large-scale language model using prompts that include "processing instructions based on processing constraints" as described above, thereby enabling the large-scale language model to accurately generate desired behavioral suggestion information.

[0063] For example, the business support device 100 (learning unit) can use prompts containing the "<constraints>" shown in (1-2) of Figure 7 to impose constraints on the large-scale language model, such as "use organizational information and user behavior information," "when updating business procedures, update them strictly in accordance with organizational information," and "learn the large-scale language model using the updated business procedures." As a result, the business support device 100 (learning unit) prevents the generation of behavioral suggestion information unrelated to the store when generating behavioral suggestion information in the large-scale language model, enabling the generation of more accurate behavioral suggestion information.

[0064] Furthermore, the "<command>" mentioned above is a text containing instructions for causing the large-scale language model to perform a desired process. The business support device 100 (learning unit) can cause the large-scale language model to perform a desired process by inputting a prompt containing the "<command>" that describes the desired process into the large-scale language model.

[0065] For example, the business support device 100 (learning unit) can use prompts containing "<command>" as shown in (1-3) of Figure 7 to cause the large-scale language model to perform processes such as "extracting the difference between the current business procedure and the procedure performed by the user," "generating an updated version of the current business procedure based on the difference," "executing the update of the current business procedure if the updated version meets the conditions," and "learning using the updated business procedure."

[0066] By using the prompts described above, the business support device 100 (learning unit) generates an updated business procedure in which a new step (Figure 7(2-1)) is added to the business procedure, for example, as shown in Figure 7(2). The business support device 100 (learning unit) then uses the updated business procedure to learn a large-scale language model. By repeating the above steps of "updating the business procedure" and "learning using the updated business procedure," the business support device 100 (learning unit) can self-learn a large-scale language model.

[0067] (Second example) Next, as a second example, an example in which the business support device 100 outputs behavior suggestion information generated in response to the actions of a user operating a predetermined system will be explained using Figure 8. Figure 8 is a diagram showing an example of system operation support according to the embodiment. The large-scale language model used in the second example is a trained large-scale language model that has undergone the "self-learning" described in the first example.

[0068] In the second example, the business support device 100 (generation unit) generates behavior suggestion information by inputting, using prompts, into a large-scale language model a command to generate user behavior information, which includes at least one of the acquired user operation history of the terminal device and information regarding the user's gaze, and behavior suggestion information, which includes at least one of the information to request additional input from the user and suggestions for actions to be taken by the user.

[0069] For example, as shown in (1) of Figure 8, the business support device 100 (generation unit) uses prompts including "<role>", "<constraints>", "<commands>", etc., to cause a large-scale language model to generate behavioral suggestion information.

[0070] Specifically, the business support device 100 (generation unit) assigns the role of "supporting users belonging to an organization when they are having trouble with how to proceed with their work" to the large-scale language model by using prompts that contain the "<Role>" shown in (1-1) of Figure 8.

[0071] Furthermore, the business support device 100 (generation unit) imposes constraints on the large-scale language model, such as "use organizational information and user behavior information" and "generate behavior suggestion information only if the user does not operate for a certain period of time and deviates from the business procedure," by using prompts that contain the "<Constraints>" shown in (1-2) of Figure 8.

[0072] Furthermore, the business support device 100 (generation unit) uses prompts containing "<command>" as shown in (1-3) of Figure 8 to cause the large-scale language model to perform actions such as "generating information suggesting the next business procedure" and "generating questions to confirm points of confusion for the user."

[0073] The business support device 100 (output unit) then outputs to the user the generated "information suggesting the next business procedure" and "questions to confirm any points the user is unsure of" (Figure 8 (2)).

[0074] For example, as shown in (2-1) of Figure 8, the business support device 100 (output unit) displays information on the system screen operated by the user that suggests the next business procedure, such as "Next, please enter XX in this item."

[0075] Furthermore, for example, the business support device 100 (output unit), as shown in (2-2) of Figure 8, displays questions on the system screen operated by the user to confirm any points of confusion the user may have, such as "Is there anything I can help you with?". The user then answers these questions in a chat format or similar.

[0076] Furthermore, a second example of the business support device 100 presenting predetermined options to the user, identifying the user's problem for which it is difficult to verbalize their problem, and then outputting action suggestion information to the user according to that problem will be explained using Figure 9. Figure 9 is a diagram showing an example of user support based on options according to the embodiment.

[0077] In one example shown in Figure 9, the business support device 100 (generation unit) generates options to clarify unclear points based on a predetermined framework such as 6W1H related to information transmission. Next, the business support device 100 (reception unit) receives action information that expresses the action desired by the user when the user selects an option displayed based on the predetermined framework related to information transmission.

[0078] The business support device 100 (generation unit) generates behavior suggestion information by inputting commands to a large-scale language model using prompts, which include at least one of the following: information that the user is asked to input additionally, and suggestions for actions to be taken by the user, in accordance with the behavior information.

[0079] For example, as shown in (1) of Figure 9, the business support device 100 (generation unit) uses prompts including "<role>", "<constraints>", "<commands>", etc., to cause a large-scale language model to generate behavioral suggestion information.

[0080] Specifically, the business support device 100 (generation unit) assigns the role of "supporting users belonging to an organization when they are having trouble with how to proceed with their work" to the large-scale language model by using prompts that contain the "<Role>" shown in (1-1) of Figure 9.

[0081] Furthermore, the business support device 100 (generation unit) imposes constraints on the large-scale language model, such as "use organizational information and user behavior information" and "present the user with options based on '6W1H'," by using prompts that contain the "<Constraints>" shown in (1-2) of Figure 9.

[0082] Furthermore, the business support device 100 (generation unit) uses prompts containing "<command>" as shown in (1-3) of Figure 9 to cause the large-scale language model to perform actions such as "presenting options to clarify unclear points for the next user" and "presenting the user's next action based on the selected option."

[0083] The business support device 100 (output unit) then outputs the generated "options to clarify the user's unclear points" to the user. For example, as shown in (2-1) of Figure 9, the business support device 100 (output unit) displays to the user a message that says, "Is there anything I can help you with? Please select from the following options," along with options such as "When, who, where, what..."

[0084] Then, as shown in (2-2) of Figure 9, the business support device 100 (output unit) displays action suggestion information generated based on the options selected by the user, such as "You're having trouble with XX, I understand. I'll explain the next steps you (the user) should take regarding XX."

[0085] (Third example) Next, as a third example, we will explain, using Figure 10, an example in which the business support device 100 uses organizational information and user behavior information to automatically check the checkable parts of a checklist related to business procedures, and for parts that cannot be checked, it conducts interviews with the user to perform the checks. Figure 10 is a diagram showing an example of checklist information generation according to the embodiment. The large-scale language model used in the third example is a trained large-scale language model that has undergone the "self-learning" described in the first example.

[0086] In the third example, the business support device 100 (generation unit) uses organizational information, including logs related to a predetermined system, to generate a first checklist updated with predetermined confirmation items marked as confirmed (Figure 10 (1-1)).

[0087] Here, the business support device 100 (generation unit) inputs a command to the large-scale language model using prompts to generate predetermined questions for the user in order to update any unconfirmed items in the checklist related to the business to confirmed items, thereby generating predetermined questions. For example, the business support device 100 (generation unit) generates predetermined questions for the user as shown in (1-2) of Figure 10.

[0088] Next, the business support device 100 (generation unit) uses the results of predetermined questions to generate a second checklist in which the predetermined confirmation items included in the first checklist that were not confirmed have been updated to confirmed (not shown in Figure 10).

[0089] Then, the business support device 100 (generation unit) generates a confirmed checklist in which all predetermined confirmation items are updated to "confirmed" based on the first and second checklists that have been generated (Figure 10 (1-3)).

[0090] For example, as shown in (2) of Figure 10, the business support device 100 (generation unit) uses prompts including "<role>", "<constraints>", "<commands>", etc. to cause the large-scale language model to generate a first checklist and a second checklist.

[0091] Specifically, the business support device 100 (generation unit) assigns the role of "checking whether the organization's operations are being carried out without problems" to the large-scale language model by using prompts that contain the "<Role>" shown in (2-1) of Figure 10.

[0092] Furthermore, the business support device 100 (generation unit) uses prompts containing the "<Constraints>" shown in (2-2) of Figure 10 to impose constraints on the large-scale language model, such as "use organizational information and user behavior information," "record a flag of "OK" for items on the checklist that meet the conditions and "NG" for items that do not," and "ask the user if it is unclear whether an item on the checklist meets the conditions."

[0093] Furthermore, the business support device 100 (generation unit) uses prompts containing "<command>" as shown in (2-3) of Figure 10 to cause the large-scale language model to perform actions such as "generate a first checklist using organizational information and user behavior information," "ask the user questions about unchecked items in the first checklist," "generate a second checklist based on the user's answers to the questions," and "generate checklist information using the first and second checklists."

[0094] (Procedure for processing by the business support device 100) Next, the processing procedure implemented by the business support device 100 according to this embodiment will be explained using Figure 11. Figure 11 is a flowchart showing the processing performed by the business support device 100 according to this embodiment.

[0095] The learning unit 132 learns a generative model based on user behavior information and organizational information (S101). If no behavioral suggestion information is generated (No. in S102), the business support device 100 waits for processing.

[0096] On the other hand, if behavioral suggestion information is to be generated (Yes in S102), the business support device 100 performs the following processing. For example, the business support device 100 may start the process of generating behavioral suggestion information when the detection unit 133 detects a predetermined action of the user.

[0097] The generation unit 135 inputs a command to generate behavioral suggestion information to the large-scale language model using prompts, and generates the behavioral suggestion information (S103). Next, the output unit 136 outputs the generated behavioral suggestion information in a predetermined format (S104). Then, the business support device 100 terminates processing.

[0098] (effect) Next, we will explain the effects of the business support device 100 according to this embodiment. In conventional workflow systems, even when a user does not know which items to input and what kind of information to enter, it was sometimes difficult to provide appropriate support to the user's problem.

[0099] Therefore, the generation unit 135 of the business support device 100 according to this embodiment receives prompts, including commands to generate behavioral suggestion information corresponding to user behavior, from a large-scale language model learned using organizational information and user behavior information, and generates behavioral suggestion information. The output unit 136 of the business support device 100 outputs the generated behavioral suggestion information to the user in a predetermined format.

[0100] Therefore, the business support device 100 according to this embodiment has the effect of enabling efficient business operations in accordance with business procedures. Furthermore, the business support device 100 according to this embodiment achieves predetermined effects by executing the processes described below.

[0101] The learning unit 132 learns a large-scale language model using updated organizational business procedures based on the difference between user work procedures identified using user behavior information, which includes at least one of the logs of systems operated by the user and the history of the user's physical movements, and organizational business procedures calculated using organizational information, which includes at least one of the following: information on organizational regulations, information on users belonging to the organization, user conversation logs, and organizational approval information.

[0102] Through the processes described above, the business support device 100 enables the large-scale language model to self-learn by repeatedly executing processes to update predetermined business procedures and to learn the large-scale language model using the updated business procedures. As a result, the business support device 100 generates action suggestion information based on the latest business procedures, thereby enabling the user to efficiently carry out their work in accordance with the latest business procedures.

[0103] The reception unit 134 receives behavioral information in which the user selects an option displayed based on a predetermined framework for information transmission, thereby expressing the action desired by the user. The generation unit 135 generates behavioral suggestion information by prompting a large-scale language model to input commands to generate behavioral suggestion information that includes at least one of the following: information to request additional input from the user, and suggestions for actions to be taken by the user, in accordance with the behavioral information.

[0104] Through the process described above, the business support device 100 can generate appropriate action-oriented information by clarifying the user's points of confusion, even if the user is not clear about what they don't understand. As a result, even if the user does not understand their own points of confusion, the business support device 100 can present appropriate action-oriented information, thereby enabling the user to efficiently proceed with their work according to the work procedures.

[0105] The generation unit 135 generates behavior suggestion information by inputting a command to a large-scale language model using prompts, which includes user behavior information that includes at least one of the acquired user operation history of the terminal device and information regarding the user's gaze, as well as information that the user is asked to input additionally and a suggestion for action to the user.

[0106] Through the process described above, the business support device 100 can suggest the next action based on the user's own system operation if the user has difficulty with an operation or if the work stops. Therefore, the business support device 100 enables the user to understand the next task without having to look up unclear points each time. As a result, the business support device 100 has the effect of enabling the user to proceed with work efficiently in accordance with the work procedure.

[0107] The generation unit 135 uses organizational information including logs related to a predetermined system to generate a confirmed checklist in which all predetermined confirmation items have been updated to be confirmed, based on a first checklist in which predetermined confirmation items have been updated to be confirmed, and a second checklist in which the parts of the first checklist that were not confirmed have been updated to be confirmed, using the results of predetermined questions.

[0108] Specifically, the generation unit 135 inputs a command to the large-scale language model using prompts to generate a predetermined question for the user in order to update any unconfirmed items in the checklist related to the work to confirmed items, thereby generating the predetermined question.

[0109] Through the process described above, the business support device 100 automatically checks areas that do not require human verification based on accumulated information such as system logs within the organization. Furthermore, for areas that cannot be automatically checked, the business support device 100 conducts interviews with the user to perform the check, thereby eliminating the need for user verification work. As a result, the business support device 100 enables users to perform their tasks efficiently.

[0110] <Variation> The following describes modifications implemented by the business support device 100 according to this embodiment.

[0111] (Data, etc.) The action suggestion information, next action, checklist checks, names of the functional parts of the business support device 100, steps, processes, names of steps or processes, etc., used in the description of the above embodiment are merely examples and can be changed at will.

[0112] For example, while it was explained that Organization Information DB121 stores information related to each of the following items in a table format, such as "No," which is information that identifies individual data included in Organization Information, "Organization Identification Information," "Organization Regulations," "Work Procedures," and "Conversation Logs," the items and information within each item that are stored are not limited to those shown in Figure 3. Similarly, while it was explained that User Behavior Information DB122 stores information related to each of the following items in a table format, such as "No," which is information that identifies individual data included in User Behavior Information, "User Identification Information," "Operation Logs," and "Physical Movement History," the items and information within each item that are stored are not limited to those shown in Figure 4. Furthermore, while it was explained that Learning Data DB123 stores information related to each of the following items in a table format, such as "No," which is information that identifies individual data included in Learning Data, "Organization Identification Information," and "Learning Work Procedures," the items and information within each item that are stored are not limited to those shown in Figure 5. Furthermore, while it was explained that the checklist information DB124 stores the information in a table format, etc., by associating "No," which is information that identifies individual data included in the checklist information, with the information related to each item in "First Checklist" and "Second Checklist," the items stored and the information within each item are not limited to those shown in Figure 6.

[0113] (Regarding the use of generative models) In this embodiment, the model (large-scale language model) used by the business support device 100 is described as being stored in the generated model DB 125 of the storage unit 120, but this is not limited to this. For example, the business support device 100 can access an external information processing device (server, etc.) and use a predetermined model.

[0114] (Flowcharts, etc.) In flowcharts, each step may be rearranged as long as it does not create inconsistencies, and some steps may be omitted. Furthermore, conjunctions such as "next," "continue," "in addition," "at this time," and "on this occasion" in flowchart descriptions do not limit the order or timing of the processes in the flowchart.

[0115] <Hardware Configuration> Each component of the illustrated device is a functional concept and does not necessarily have to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions. Furthermore, each processing function performed by each device can be implemented, all or any part of it, by a CPU and the program that is analyzed and executed by that CPU, or by hardware using wired logic.

[0116] Furthermore, among the processes described in this embodiment, all or part of those described as being performed automatically can be performed manually using known methods. In addition, the processing procedures, control procedures, specific names, and information including various data and parameters shown in the drawings can be arbitrarily changed unless otherwise specified.

[0117] <Program> In one embodiment, the various devices constituting the business support device 100 can be implemented by installing a business support program as packaged software or online software on a desired computer. For example, by having the above-mentioned business support program run on an information processing device, it can function as various devices constituting the business support device 100. The information processing device referred to here includes desktop or notebook personal computers. In addition, the information processing device also includes mobile communication terminals such as smartphones and mobile phones, and slate terminals such as PDAs (Personal Digital Assistants).

[0118] Figure 12 shows an example of a computer that implements the business support device 100 according to the embodiment. The computer 1000 has, for example, memory 1010 and CPU 1020. The computer 1000 also has a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0119] Memory 1010 includes ROM (Read Only Memory) 1011 and RAM 1012. ROM 1011 stores, for example, a boot program such as BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1090. The disk drive interface 1040 is connected to the disk drive 1100. For example, a removable storage medium such as a magnetic disk or optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to, for example, a mouse 1110 and a keyboard 1120. The video adapter 1060 is connected to, for example, a display 1130.

[0120] The hard disk drive 1090 stores, for example, an OS (Operating System) 1091, an application program 1092, a program module 1093, and program data 1094. That is, the programs that define the various processes of the various devices constituting the business support system 100 are implemented as program modules 1093 in which computer-executable code is written. The program modules 1093 are stored, for example, in the hard disk drive 1090. For example, a program module 1093 for performing processes similar to the functional configurations of the various devices constituting the business support system 100 is stored in the hard disk drive 1090. Note that the hard disk drive 1090 may be replaced by an SSD (Solid State Drive).

[0121] Furthermore, the configuration data used in the processing of the embodiment described above is stored as program data 1094 in, for example, memory 1010 or hard disk drive 1090. The CPU 1020 then reads the program module 1093 and program data 1094 stored in memory 1010 or hard disk drive 1090 into RAM 1012 as needed and executes the processing of the embodiment described above.

[0122] Furthermore, the program module 1093 and program data 1094 are not limited to being stored in the hard disk drive 1090; for example, they may be stored in a removable storage medium and read by the CPU 1020 via a disk drive 1100 or the like. Alternatively, the program module 1093 and program data 1094 may be stored in another computer connected via a network (LAN, WAN (Wide Area Network), etc.). The program module 1093 and program data 1094 may then be read from the other computer by the CPU 1020 via a network interface 1070.

[0123] <Other> Although this embodiment has been described above, this embodiment is not limited by the description and drawings that constitute part of the disclosure. That is, all other embodiments, examples, and operational techniques made by those skilled in the art based on this embodiment are included in the scope of this embodiment. [Explanation of Symbols]

[0124] 100 Business support equipment 110 Communications Department 120 Storage section 121 Organizational information DB 122 User Behavior Information Database 123 Training Data Database 124 Checklist Information Database 125 Generative Model DB 130 Control Unit 131 Acquisition Department 132 Learning Department 133 Detection unit 134 Reception Department 135 Generation part 136 Output section

Claims

1. A generation unit that generates information on suggestions for business procedures in an organization by receiving prompts that include commands to generate information on suggestions for business procedures in an organization in response to user behavior, to a large-scale language model that has been trained using information on an organization and information on the behavior of users who are members of the organization. An output unit that outputs information regarding suggestions of business procedures within the organization that has been generated to the user in a predetermined format, A business support device characterized by having the following features.

2. The procedures performed by the user, identified using information about the user's actions, which includes at least one of the user's system logs and the user's physical movement history, The system further includes a learning unit that learns the large-scale language model using the updated business procedures related to the organization, based on the difference between the business procedures related to the organization calculated using information about the organization, which includes at least one of the following: information about the organization's regulations, information about users belonging to the organization, the conversation logs of the users, and approval information related to the organization. The business support device according to feature 1.

3. The system further includes a receiving unit that receives behavioral information representing the action desired by the user when the user selects an option displayed based on a predetermined framework for information transmission. The generating unit is Based on the aforementioned behavioral information, a command is input to the large-scale language model using the prompt expressed in natural language text to generate information regarding suggestions for business procedures within the organization, which includes at least one of the following: information to request additional input from the user, and suggestions for user actions. The business support device according to claim 1 or 2.

4. The generating unit is Information regarding the user's behavior, including at least one of the user's terminal device operation history and information regarding the user's gaze, A command to generate information regarding suggestions for business procedures within the organization, which includes at least one of the following: information to request additional input from the user, and suggestions for actions to be taken by the user, is input to the large-scale language model using the prompt expressed in natural language text. To generate information that suggests business procedures within the aforementioned organization, The business support device according to claim 1 or 2.

5. The generating unit is Using information about the organization, including logs related to a specified system, a first checklist updated to confirm the specified items, Using the results of the prescribed questions, the second checklist is created by updating the parts of the first checklist that are not yet confirmed to confirmed, Based on this, a confirmed checklist is generated in which all of the aforementioned specified confirmation items are updated to "confirmed". The business support device according to claim 1 or 2.

6. The generating unit is The system generates the predetermined questions for the user to update any unconfirmed items in the checklist related to the work to confirmed items by inputting a command to the large-scale language model using a prompt expressed in natural language text, thereby generating the predetermined questions. The business support device according to feature 5.

7. The generating unit is As the aforementioned large-scale language model, at least one of the following is used: a large-scale language model possessing general knowledge, and a predetermined large-scale language model on which adapter tuning is performed. The business support device according to claim 1 or 2.

8. A business support method to be executed by a business support device, A generation step of generating information that suggests business procedures in an organization by inputting prompts that include commands to a large-scale language model, which has been trained using information about the organization and information about the behavior of users who are members of the organization, to generate information that suggests business procedures in the organization in response to user behavior, An output step which outputs information regarding suggestions of business procedures within the organization that has been generated to the user in a predetermined format, A business support method characterized by including the following.

9. A generation step of generating information about suggestions for business procedures in an organization by inputting prompts that include commands to a large-scale language model trained using information about an organization and information about the behavior of users who are members of the organization, to generate information about suggestions for business procedures in the organization in response to user behavior, An output step which outputs information regarding suggestions of business procedures in the organization that has been generated to the user in a predetermined format, A business support program that allows a computer to execute tasks.

Images