Information processing device, information processing method, and program

Abstract

To provide technology that can generate explanations for data. [Solution] An information processing device according to one aspect of the present disclosure includes: a selection unit that selects one or more types as the type of explanation to be generated for data relating to equipment; a data acquisition unit that acquires the data according to the selected one or more types; an explanation generation unit that generates one or more explanations corresponding to each of the one or more types based on the acquired data and a generation AI realized by a machine learning model including a large-scale language model; and an integration unit that creates a first integrated explanation representing an explanation that integrates the one or more explanations corresponding to each of the one or more types.

Description

[Technical Field]

[0001] This disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] Methods are known for determining the optimal operating plan for various types of plants (e.g., power plants, steel plants, chemical plants, oil plants, energy plants, etc.) by solving optimization problems. For example, Patent Document 1 discloses a method for calculating the optimal operating plan for a heat source system. Also, for example, Patent Document 2 discloses a method for creating the optimal production plan for production equipment. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Patent No. 6810597 [Patent Document 2] Patent No. 7349608 [Overview of the project] [Problems that the invention aims to solve]

[0004] However, verifying the correctness of driving plans obtained by solving optimization problems is generally difficult, so there is a need for technology to generate explanations for the data representing those driving plans. This applies not only to data representing driving plans obtained by solving optimization problems, but also to data representing actual driving performance, for example.

[0005] This disclosure is made in view of the above points and aims to provide a technology that can generate explanations for data. [Means for solving the problem]

[0006] An information processing device according to one aspect of the present disclosure includes: a selection unit that selects one or more types as the type of description to be generated for data relating to equipment; a data acquisition unit that acquires the data according to the selected one or more types; an explanation generation unit that generates one or more descriptions corresponding to each of the one or more types based on the acquired data and a generation AI realized by a machine learning model including a large-scale language model; and an integration unit that creates a first integrated explanation representing an explanation that integrates the one or more descriptions corresponding to each of the one or more types. [Effects of the Invention]

[0007] It is possible to generate explanations for the data. [Brief explanation of the drawing]

[0008] [Figure 1] This diagram shows an example of the configuration of the target equipment. [Figure 2] This figure shows an example of the hardware configuration of the data explanation device according to this embodiment. [Figure 3] This figure shows an example of the functional configuration of the data explanation device according to this embodiment. [Figure 4] This flowchart shows an example of the data explanation process according to this embodiment. [Figure 5] This figure shows an example of operation history data in array format. [Figure 6] This figure shows an example of optimal driving plan data in array format. [Figure 7] This flowchart shows an example of the explanation generation process according to this embodiment. [Figure 8] This figure shows an example of a prompt (part 1). [Figure 9] This figure shows an example of a prompt (part 2). [Figure 10] This figure shows an example of a prompt (part 3). [Figure 11] This figure shows an example of data explanation information (part 1). [Figure 12]This is a diagram showing an example of data description information (Part 2). [Figure 13] This is a flowchart showing a modification example (Part 1) of the explanation generation process according to this embodiment. [Figure 14] This is a flowchart showing a modification example (Part 2) of the explanation generation process according to this embodiment.

Modes for Carrying Out the Invention

[0009] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the following embodiments, a data description device 10 capable of generating an explanation for data related to a certain predetermined facility (hereinafter referred to as "target facility") will be described. Examples of the target facility include various plants such as a power generation plant, a steel plant, a chemical plant, an oil plant, an energy plant, and the like. Further, examples of the data related to the target facility include data representing the operation results of the target facility (hereinafter referred to as "operation result data"), data representing the optimal operation plan of the target facility (hereinafter referred to as "optimal operation plan data"), and the like. The optimal operation plan data is, for example, data obtained by solving an optimization problem for obtaining the optimal operation plan of the target facility.

[0010] Note that the form of the optimization problem for obtaining the optimal operation plan data is not limited to a specific form, and it is possible to use the optimal operation plan data obtained by solving an optimization problem of any form. That is, the optimal operation plan data may be, for example, data obtained by solving a continuous optimization problem or data obtained by solving a discrete optimization problem (which is also called "combinatorial optimization problem", etc.). Also, various methods (e.g., metaheuristic, reinforcement learning, other existing methods, etc.) can be used for solving the optimization problem according to the problem. Hereinafter, as an example, it is assumed that the optimal operation plan data is data obtained by solving a continuous optimization problem.

[0011] Furthermore, targeting data related to equipment is just one example; it is also possible to target data related to machinery, system data, etc.

[0012] In this embodiment, the data explanation device 10 generates explanations for data relating to the target equipment using generation AI (Artificial Intelligence). At this time, the data explanation device 10 generates multiple explanations for the data and then creates an integrated explanation. As a result, by referring to this explanation, the user can, for example, understand current operational problems from operational performance or verify the validity of the optimal operational plan. Furthermore, it becomes possible to confirm improvement measures for current operational problems and compare the current operation with the optimal operational plan. Therefore, by using the data explanation device 10 according to this embodiment, it is possible to support the user's decision-making when making any kind of decision regarding the target equipment (e.g., deciding whether or not to actually apply the optimal operational plan).

[0013] The data explanation device 10 according to this embodiment is implemented, for example, by an information processing device (computer) or an information processing system (computer system) consisting of one or more information processing devices. Specific examples of information processing devices include personal computers, general-purpose servers, smartphones, tablet terminals, and the like.

[0014] Generative AI refers to technologies or systems that generate content such as natural language sentences or images that satisfy given instructions or questions (these instructions or questions are called "prompts," etc.). Generative AI is generally implemented using machine learning models, such as large language models (LLMs).

[0015] <Example of equipment configuration> The following example of target equipment is the energy plant shown in Figure 1. The energy plant shown in Figure 1 consists of a gas turbine (+ exhaust gas boiler), a once-through boiler, N turbo chillers, and N steam absorption chillers. The gas turbine (+ exhaust gas boiler) consists of a gas turbine that generates electricity from city gas and an exhaust gas boiler that outputs steam from the gas discharged from the gas turbine. The once-through boiler outputs steam from city gas. The turbo chillers output cooling using electricity generated by the gas turbine and electricity purchased as needed. The steam absorption chillers output cooling using steam output from the exhaust gas boiler and steam output from the once-through boiler. The electricity, cooling, and steam output from the energy plant are used as power load, heat load, and steam load, respectively.

[0016] Note that the configuration of the energy plant shown in Figure 1 is just one example and is not limited to this configuration.

[0017] <Example hardware configuration of data explanation device 10> Figure 2 shows an example of the hardware configuration of the data explanation device 10 according to this embodiment. As shown in Figure 2, the data explanation device 10 according to this embodiment includes an input device 101, a display device 102, an external I / F 103, a communication I / F 104, a RAM (Random Access Memory) 105, a ROM (Read Only Memory) 106, an auxiliary storage device 107, and a processor 108. Each of these hardware components is connected to each other via a bus 109 so as to be able to communicate.

[0018] The input device 101 is, for example, a keyboard, mouse, touch panel, or physical button. The display device 102 is, for example, a display or display panel. The data explanation device 10 does not necessarily have to have at least one of the input device 101 and the display device 102.

[0019] External I / F 103 is an interface with external devices such as recording media 103a. Examples of recording media 103a include CD (Compact Disc), DVD (Digital Versatile Disk), SD memory card (Secure Digital memory card), and USB (Universal Serial Bus) memory card.

[0020] The communication interface 104 is an interface for connecting to a communication network. The RAM 105 is a volatile semiconductor memory (storage device) that temporarily holds programs and data. The ROM 106 is a non-volatile semiconductor memory (storage device) that can retain programs and data even when the power is turned off. The auxiliary storage device 107 is a non-volatile storage device such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory. The processor 108 is a processing unit such as a CPU (Central Processing Unit) or GPU (Graphic Processing Unit).

[0021] Note that the hardware configuration shown in Figure 2 is just one example, and the hardware configuration of the data explanation device 10 is not limited to this. For example, the data explanation device 10 may have multiple auxiliary storage devices 107 or multiple processors 108, it may not have some of the hardware shown, or it may have various other hardware components besides the hardware shown.

[0022] <Example of the functional configuration of the data explanation device 10> Figure 3 shows an example of the functional configuration of the data explanation device 10 according to this embodiment. As shown in Figure 3, the data explanation device 10 according to this embodiment includes an explanation type selection unit 201, a data acquisition unit 202, an information organization unit 203, an explanation generation processing unit 204, an explanation output unit 205, and a user input acquisition unit 206. The data explanation device 10 according to this embodiment also includes a data storage unit 301.

[0023] The explanation type selection unit 201, data acquisition unit 202, information organization unit 203, explanation generation processing unit 204, explanation output unit 205, and user input acquisition unit 206 are implemented, for example, by a process executed by a processor 108 or the like, which is run by one or more programs installed on the data explanation device 10. At least one of these units may be implemented by a process executed by, for example, an external cloud server or the like.

[0024] Furthermore, the data storage unit 301 may be implemented by a storage area such as an auxiliary storage device 107. Alternatively, the data storage unit 301 may be implemented by a storage area of ​​a storage device (e.g., a storage device provided by a database server) that is communicatively connected to the data explanation device 10.

[0025] The explanation type selection unit 201 allows the user to select one or more explanation types from a predetermined set of one or more explanation types for which they wish to receive an explanation. An explanation type is a type of explanation for data related to the target equipment. As an example, four explanation types are assumed below: "Problems with current operation," "Measures to improve the problems," "Comparison and consideration of current operation and optimization results," and "Reasons for the optimization results." Here, "Problems with current operation" is the explanation type for when the user wants to receive an explanation about problems with the current operation of the target equipment. "Measures to improve the problems" is the explanation type for when the user wants to receive an explanation about measures to improve the problems with the current operation of the target equipment. "Comparison and consideration of current operation and optimization results" is the explanation type for when the user wants to receive an explanation about a comparison and consideration of the current operation of the target equipment and the optimal operation plan. "Reasons for the optimization results" is the explanation type for when the user wants to receive an explanation about the reasons for the optimal operation plan of the target equipment. However, these four explanation types are all examples, and the explanation types are not limited to these four.

[0026] The data acquisition unit 202 acquires at least one of the driving performance data and the optimal driving plan data from the data storage unit 301 according to the explanation type selected by the explanation type selection unit 201. Here, for example, if the explanation type selected by the explanation type selection unit 201 includes "problems with current operation" or "measures to improve the problems," the data acquisition unit 202 acquires the driving performance data from the data storage unit 301; if it includes "reasons for the optimization result," it acquires the optimal driving plan data from the data storage unit 301; and if it includes "comparison and consideration of current operation and optimization result," it acquires the driving performance data and the optimal driving plan data from the data storage unit 301.

[0027] The information organization unit 203 organizes or converts the data acquired by the data acquisition unit 202 (operation performance data or optimal operation plan data or both) into a predetermined format (e.g., array format) that can be input into the generating AI.

[0028] The explanation generation processing unit 204 generates data explanation information that represents an explanation for the data acquired by the data acquisition unit 202. Here, the explanation generation processing unit 204 includes a prompt generation unit 211, an explanation generation unit 212, and an integration unit 213. The prompt generation unit 211 creates a prompt corresponding to each explanation type selected by the explanation type selection unit 201. The explanation generation unit 212 generates data explanation information that represents an explanation for the data acquired by the data acquisition unit 202, using the prompt generated by the prompt generation unit 211 and the generation AI for each explanation type selected by the explanation type selection unit 201. That is, the explanation generation unit 212 generates data explanation information using the generation AI by inputting the prompt generated by the prompt generation unit 211 into the generation AI. The integration unit 213 creates data explanation information (hereinafter also referred to as "integrated data explanation information") by integrating one or more data explanation information generated by the explanation generation unit 212.

[0029] Furthermore, large-scale language models that realize generative AI acquire various domain knowledge during training. Therefore, when data descriptive information is generated by generative AI, it can be expected that this domain knowledge will be utilized in the generation of data descriptive information.

[0030] Furthermore, the explanation generation unit 212 inputs user input information, described later, as a prompt to the generation AI, which then generates new data explanation information. This new data explanation information may include, for example, information representing additional explanations for the previously generated data explanation information by the generation AI, information representing a summary of the previously generated data explanation information by the generation AI, or information representing answers to any questions or instructions regarding the previously generated data explanation information by the generation AI.

[0031] The explanation output unit 205 outputs the data explanation information or integrated data explanation information generated by the explanation generation unit 212 to a predetermined output destination. Examples of predetermined output destinations include a display device 102 such as a display, a storage area such as an auxiliary storage device 107, and other devices or equipment connected via communication.

[0032] The user input acquisition unit 206 inputs information such as natural language sentences entered by the user (hereinafter also referred to as "user input information"). Examples of user input information include information representing natural language sentences requesting additional explanations for data description information, information representing natural language sentences requesting summaries of data description information, information representing natural language sentences such as standardized questions about data description information, and information representing natural language sentences that mean some kind of question or instruction about other data description information.

[0033] The data storage unit 301 stores data related to the target equipment (such as operational performance data and optimal operation plan data). Operational performance data is collected from the target equipment (more precisely, collected from sensors installed on the target equipment or its constituent devices or equipment). On the other hand, optimal operation plan data is obtained by solving an optimization problem to determine the optimal operation plan for the target equipment. The entity that solves the optimization problem may be the data explanation device 10, or it may be a server (e.g., a cloud server) that is communicatively connected to the data explanation device 10. When the data explanation device 10 solves the optimization problem, the data explanation device 10 may have a functional unit for solving the optimization problem, for example, called an "optimization calculation unit".

[0034] <Example of data explanation processing> An example of the data explanation processing according to this embodiment will be described with reference to Figure 4. For example, steps S101 to S105 in Figure 4 are executed whenever data explanation information is needed, and steps S106 to S108 are executed one or more times as needed after the data explanation information has been obtained.

[0035] The explanation type selection unit 201 selects one or more explanation types from one or more predetermined explanation types for which the user wishes to receive an explanation (step S101). For example, the explanation type selection unit 201 may select the explanation types specified by the user from one or more predetermined explanation types. In the following explanation, we will assume that four explanation types have been selected as an example: "Problems with current operation," "Measures to improve the problems," "Comparison and consideration of current operation and optimization results," and "Reasons for the optimization results."

[0036] The data acquisition unit 202 acquires at least one of the driving performance data and the optimal driving plan data from the data storage unit 301 according to the explanation type selected in step S101 (step S102). For example, if the explanation type selected in step S101 includes "problems with current operation" or "measures to improve the problems," the data acquisition unit 202 acquires the driving performance data from the data storage unit 301. Also, for example, if the explanation type selected in step S101 includes "reasons for the optimization result," the data acquisition unit 202 acquires the optimal driving plan data from the data storage unit 301. Furthermore, for example, if the explanation type selected in step S101 includes "comparison and consideration of current operation and optimization result," the data acquisition unit 202 acquires the driving performance data and the optimal driving plan data from the data storage unit 301.

[0037] Here, the operational performance data and the optimal operation plan data are represented as time-series data of data (hereinafter referred to as "state data") composed of one or more variables (when the target equipment is a plant, these variables are also called "state variables," etc.). When the target equipment is the energy plant shown in Figure 1, examples of state variables include the amount of electricity purchased, the amount of city gas purchased, the amount of city gas used by the gas turbine, the power output of the gas turbine, the steam output of the exhaust gas boiler, the amount of electricity used by the turbo chiller, the heat output of the turbo chiller, the amount of steam used by the steam absorption chiller, the heat output of the steam absorption chiller, the amount of city gas used by the once-through boiler, the steam output of the once-through boiler, the surplus steam of the once-through boiler, the power load, the heat load, the steam load, etc.

[0038] Furthermore, if operational performance data is acquired in step S102 above, the data acquisition unit 202 may, for example, acquire status data constituting the operational performance data sequentially in real time from the target equipment.

[0039] The information organization unit 203 organizes or converts the data acquired in step S102 (operational performance data or optimal operation plan data or both) into a predetermined format (e.g., array format) that can be input to the generating AI (step S103). As an example, Figure 5 shows the operational performance data organized or converted into array format for the energy plant shown in Figure 1 as the target facility. Similarly, Figure 6 shows the optimal operation plan data organized or converted into array format for the energy plant shown in Figure 1 as the target facility. Figures 5 and 6 show examples of how the time-series data of each state variable is represented as an array. In Figures 5 and 6, each row represents one array, and each array contains the name of the state variable corresponding to that array and the unit of the value set for that state variable. In addition, each array stores the time-series data of the state variable from time s=1 to time s=t.

[0040] If the data acquired in step S102 above (driving performance data or optimal driving plan data or both) is in a format that can be input into the generating AI, then step S103 above does not need to be performed. Hereafter, the data after being organized or converted in step S103 above (driving performance data or optimal driving plan data or both) will also be simply referred to as "driving performance data" or "optimal driving plan data".

[0041] The explanation generation processing unit 204 executes an explanation generation process (step S104) to generate data explanation information that represents an explanation for the data organized or transformed in step S103. This results in integrated data explanation information. Details of the explanation generation process will be described later.

[0042] The explanation output unit 205 outputs the integrated data explanation information obtained in step S104 to a predetermined output destination (step S105).

[0043] The user input acquisition unit 208 acquires user input information entered by the user (step S106).

[0044] The explanation generation unit 212 of the explanation generation processing unit 204 generates new data explanation information using the user input information obtained in step S106 and the generation AI (step S107). That is, the explanation generation unit 212 inputs the user input information obtained in step S106 as a prompt to the generation AI, and the generation AI generates new data explanation information.

[0045] The explanation output unit 205 outputs the data explanation information generated in step S107 to a predetermined output destination (step S108).

[0046] ≪Example of explanation generation process≫ An example of the explanation generation process according to this embodiment will be explained with reference to Figure 7. Hereinafter, the explanation type selected in step S101 of Figure 4 will be referred to as the "explanation target type". The number of explanation target types will be referred to as the "number of explanation target types" and will be represented by I. Furthermore, the i-th explanation target type will be referred to as "explanation target type i". Hereinafter, as an example, the first explanation target type will be "Problems with current operation", the second explanation target type will be "Improvement measures for the problems", the third explanation target type will be "Comparison and consideration of current operation and optimization results", and the fourth explanation target type will be "Reasons for the optimization results".

[0047] The explanation generation processing unit 204 initializes the variable i, which represents the number of the type of item to be explained, to 1 (step S201).

[0048] The prompt generation unit 211 of the explanation generation processing unit 204 generates a prompt corresponding to the explanation target type i (step S202). The prompt generation unit 211 may generate a prompt corresponding to the explanation target type i by, for example, following steps 1 to 3 below.

[0049] Step 1: The prompt generation unit 211 obtains a prompt template, which is a template for a prompt corresponding to the type of explanation target i. A prompt template is template information that includes a natural language sentence representing instructions for generating an explanation about the type of explanation target i, and zero or more variable parts. A variable part is the part in which data corresponding to the type of explanation target i is set (i.e., the variable part of the prompt template). Prompt templates are created in advance by, for example, a user, and then stored in a predetermined memory area.

[0050] Step 2: If the prompt template obtained in Step 1 above contains a variable portion, the prompt generation unit 211 sets the data corresponding to the explanation target type i (driving performance data or optimal driving plan data or both) for that variable portion. Hereinafter, the data set for the variable portion included in the prompt template will be referred to as "setting information".

[0051] Step 3: If the prompt template obtained in Step 1 above does not contain a variable portion, the prompt generation unit 211 sets the prompt template as the prompt corresponding to the type i to be explained.

[0052] The prompt corresponding to the type i of the subject to be explained is generated by step 2 or 3 above. Examples of prompt templates that include a variable part include the prompt template corresponding to "Problems with the current operation", the prompt template corresponding to "Comparison and discussion of the current operation and the optimization results", and the prompt template corresponding to "Reasons for the optimization results". On the other hand, an example of a prompt template that does not include a variable part is the prompt template corresponding to "Measures to improve the problems".

[0053] As an example, Figure 8 shows a prompt corresponding to the explanation target type "Problems with current operation." Prompt 1100 shown in Figure 8 is a prompt for generating data explanation information based on the results of identifying problems with the current operation of the target equipment (i.e., the operation performance data shown in Figure 5) from the perspective of energy cost efficiency. Note that in prompt 1100 shown in Figure 8, the setting information is shown in bold and underlined.

[0054] As another example, Figure 9 shows a prompt corresponding to the explanation target type "Reason for optimization result". Prompt 1200 shown in Figure 9 is a prompt for generating data explanation information that explains the reason for the optimal solution obtained by mathematical optimization technique (i.e., the optimal operation plan data shown in Figure 6). Note that in prompt 1200 shown in Figure 9, the setting information is shown in bold and underlined.

[0055] As another example, Figure 10 shows a prompt corresponding to the explanation target type "Improvement measures for the problem". The prompt 1300 shown in Figure 10 is a prompt for generating improvement measures for the problem obtained as an explanation of the explanation target type "Problems with current operation" as data explanation information. Note that there is no setting information for the prompt 1300 shown in Figure 10.

[0056] The explanation generation unit 212 of the explanation generation processing unit 204 generates data explanation information corresponding to the explanation target type i using the prompt generated in step S202 (step S203). That is, the explanation generation unit 212 inputs the prompt generated in step S202 to the generation AI, and the generation AI generates data explanation information corresponding to the explanation target type i. Hereinafter, the data explanation information corresponding to the explanation target type i will also be referred to as "data explanation information i".

[0057] Furthermore, in the generation AI, when generating data description information i (where i≧2), data description information i is generated based on data description information 1 to data description information i-1. Therefore, when generating data description information i (where i≧2), it is possible to generate data description information i that also takes into account the contents of data description information 1 to data description information i-1.

[0058] As an example, Figure 11 shows data description information corresponding to the explanation target type "Problems with current operation". In the data description information 2100 shown in Figure 11, the problems with the current operation of the target equipment are explained in natural language from the perspective of energy cost efficiency. Note that the data description information 2100 shown in Figure 11 was generated by inputting the prompt 1100 shown in Figure 8 into the generating AI.

[0059] As another example, Figure 12 shows data explanation information corresponding to the explanation target type "Reason for optimization result". In the data explanation information 2200 shown in Figure 12, the reason why the optimal solution obtained by mathematical optimization technique is optimal is explained in natural language. Note that the data explanation information 2200 shown in Figure 12 is data explanation information generated by inputting the prompt 1200 shown in Figure 9 into the generating AI.

[0060] The explanation generation processing unit 204 adds 1 to the variable i, which represents the number of the type of item to be explained (step S204).

[0061] The explanation generation processing unit 204 determines whether i is less than or equal to the number of types to be explained (step S205). That is, the explanation generation processing unit 204 determines whether i ≤ I.

[0062] If it is determined in step S205 that i is less than or equal to the number of types to be explained, the explanation generation processing unit 204 returns to step S202. As a result, steps S202 to S204 are repeatedly executed until i > I.

[0063] On the other hand, if it is not determined in step S205 that i is less than or equal to the number of types to be explained, the integration unit 213 of the explanation generation processing unit 204 creates integrated data explanation information by integrating the data explanation information i (1 ≤ i ≤ I) (step S206). The method of integrating the data explanation information i (1 ≤ i ≤ I) is not limited to a specific method, but as an example, for i = 1, ..., I-1, data explanation information i+1 can be concatenated after data explanation information i. As another example, integrated data explanation information can be created by summarizing the data explanation information i (1 ≤ i ≤ I) using generation AI or existing document summarization technology.

[0064] Through the above explanation generation process, for example, integrated data explanation information is obtained that combines data explanation information corresponding to the explanation target type "Problems with current operation," "Measures to improve the problems," "Comparison and consideration of current operation and optimization results," and "Reasons for optimization results." Therefore, users can consistently understand, for example, the problems with current operation, measures to improve them, a comparison and consideration of current operation and optimization results, and the reasons for optimization results.

[0065] <Variation> • Variation 1 As a variation of the explanation generation process shown in Figure 7, we will explain the explanation generation process in which multiple data explanations are generated using the same prompt and the same generation AI, and then these data explanations are integrated, with reference to Figure 13.

[0066] The explanation generation processing unit 204 initializes the variable i, which represents the number of the type to be explained, to 1, similar to step S201 in Figure 7 (step S301).

[0067] The prompt generation unit 211 of the explanation generation processing unit 204 generates a prompt corresponding to the explanation target type i, similar to step S202 in Figure 7 (step S302).

[0068] The explanation generation processing unit 204 initializes the variable j, which represents the number of times data explanation information is generated by prompts corresponding to the explanation target type i, to 1 (step S303).

[0069] The explanation generation unit 212 of the explanation generation processing unit 204 generates the j-th data explanation information corresponding to the explanation target type i using the prompt generated in step S302 above, similar to step S203 in Figure 7 (step S304). Hereinafter, the j-th data explanation information corresponding to the explanation target type i will be referred to as "data explanation information i". j We will also express it as "."

[0070] The explanation generation processing unit 204 adds 1 to the variable j, which represents the number of times data explanation information has been generated by prompts corresponding to the explanation target type i (step S305).

[0071] The explanation generation processing unit 204 determines whether j is less than or equal to a predetermined specified number of times (step S306). That is, for example, if J(i) is the specified number of times corresponding to the type of explanation target i, the explanation generation processing unit 204 determines whether j ≤ J(i).

[0072] If it is determined in step S306 that j is less than or equal to a predetermined number of times, the explanation generation processing unit 204 returns to step S304. As a result, steps S304 to S305 are repeatedly executed until j > J(i).

[0073] On the other hand, if it is not determined in step S306 above that j is less than or equal to a predetermined specified number of times, the integration unit 213 of the explanation generation processing unit 204 generates data explanation information i j Create a first integrated data description information by integrating (1≦j≦J(i)) (Step S307). Data description information i j The method for integrating (1≦j≦J(i)) is not limited to a specific method, but as an example, for j=1,···,J(i)-1, the data explanatory information i j Data description information i j+1It is conceivable to link them together. Another example is to use generative AI or existing document summarization technologies to extract data explanatory information i j It is conceivable to create the first integrated data description information by summarizing (1 ≤ j ≤ J(i)). Below is the data description information i j The first integrated data explanation information obtained by integrating (1≦j≦J(i)) will also be referred to as "first integrated data explanation information i".

[0074] The explanation generation processing unit 204 adds 1 to the variable i, which represents the number of the type to be explained, similar to step S204 in Figure 7 (step S308).

[0075] The explanation generation processing unit 204 determines whether i is less than or equal to the number of types to be explained, similar to step S205 in Figure 7 (step S309). That is, the explanation generation processing unit 204 determines whether i ≤ I.

[0076] If it is determined in step S309 that i is less than or equal to the number of types to be explained, the explanation generation processing unit 204 returns to step S302. As a result, steps S302 to S308 are repeatedly executed until i > I.

[0077] On the other hand, if it is not determined in step S309 that i is less than or equal to the number of types to be explained, the integration unit 213 of the explanation generation processing unit 204 creates a second integrated data explanation information by integrating the first integrated data explanation information i (1 ≤ i ≤ I) (step S310). The method of integrating the first integrated data explanation information i (1 ≤ i ≤ I) is not limited to a specific method, but as an example, for i = 1, ..., I-1, the first integrated data explanation information i+1 can be concatenated after the first integrated data explanation information i. As another example, the second integrated data explanation information can be created by summarizing the first integrated data explanation information i (1 ≤ i ≤ I) using a generation AI or existing document summarization technology. In step S105 of Figure 4, the second integrated data explanation information is output to a predetermined output destination.

[0078] In the explanation generation process shown in FIG. 13, the first integrated data explanation information i that integrates the data explanation information i j (1≦j≦J(i)) is created, so that the highly robust first integrated data explanation information i can be obtained as an explanation corresponding to the explanation target type i. Therefore, even when the generation AI uses a probabilistic method to generate an explanation, the generation of highly reliable first integrated data explanation information i can be expected.

[0079] · Variant 2 As a variant of the explanation generation process shown in FIG. 7, the explanation generation process in the case where a plurality of data explanation information is generated by a plurality of generation AIs from the same prompt and then these data explanation information are integrated will be described while referring to FIG. 14.

[0080] Similar to step S201 in FIG. 7, the explanation generation processing unit 204 initializes the variable i representing the number of the explanation target type to 1 (step S401).

[0081] Similar to step S202 in FIG. 7, the prompt generation unit 211 of the explanation generation processing unit 204 generates a prompt corresponding to the explanation target type i (step S402).

[0082] Using the prompt generated in step S402 above, the explanation generation unit 212 of the explanation generation processing unit 204 generates a plurality of data explanation information by a plurality of generation AIs (step S403). That is, the explanation generation unit 212 inputs the prompt generated in step S402 above to each of the plurality of generation AIs, and a plurality of data explanation information is generated by these plurality of generation AIs. Hereinafter, the number of generation AIs used when generating data explanation information from the prompt corresponding to the explanation target type i is denoted as K(i), and the data explanation information generated by the k-th generation AI is also referred to as "data explanation information i k ". Note that all of the K(i) generation AIs may be different generation AIs, or some of the generation AIs may be the same generation AIs.

[0083] The integration unit 213 of the explanation generation processing unit 204 generates data explanation information i k Create a first integrated data description information by integrating (1≦k≦K(i)) (Step S404). Data description information i k The method for integrating (1≦k≦K(i)) is not limited to a specific method, but as an example, for k=1,···,K(i)-1, the data explanatory information i k Data description information i k+1 It is conceivable to link them together. Another example is to use generative AI or existing document summarization technologies to extract data explanatory information i k It is conceivable to create the first integrated data description information by summarizing (1 ≤ k ≤ K(i)). Below, data description information i k The first integrated data explanation information obtained by integrating (1≦k≦K(i)) will also be referred to as "first integrated data explanation information i".

[0084] The explanation generation processing unit 204 adds 1 to the variable i, which represents the number of the type to be explained, in the same way as in step S204 in Figure 7 (step S405).

[0085] The explanation generation processing unit 204 determines whether i is less than or equal to the number of types to be explained, similar to step S205 in Figure 7 (step S406). That is, the explanation generation processing unit 204 determines whether i ≤ I.

[0086] If it is determined in step S406 that i is less than or equal to the number of types to be explained, the explanation generation processing unit 204 returns to step S402. As a result, steps S402 to S405 are repeatedly executed until i > I.

[0087] On the other hand, if it is not determined in step S406 above that i is less than or equal to the number of types to be explained, the integration unit 213 of the explanation generation processing unit 204 creates a second integrated data explanation information by integrating the first integrated data explanation information i (1 ≤ i ≤ I), similar to step S310 in Figure 13 (step S407). In step S105 in Figure 4, the second integrated data explanation information is output to a predetermined output destination.

[0088] In the explanation generation process shown in Figure 14, data explanation information i k Since a first integrated data explanation information i is created by integrating (1≦k≦K(i)), a highly robust first integrated data explanation information i is obtained as an explanation corresponding to the type of explanation target i. For this reason, as with the modification 1 described above, even if the generating AI uses a probabilistic method to generate the explanation, the generation of a highly reliable first integrated data explanation information i can be expected.

[0089] • Modification example 3 The prompt is not limited to natural language text; it may also include content such as images, videos, and audio. Furthermore, it may include mathematical formulas, program code, and other similar elements.

[0090] In particular, when instructing the generating AI to perform any numerical calculations or processes, the mathematical formulas or program code used in those calculations or processes may be included in the prompt. This makes it possible to reduce calculation errors when the generating AI performs numerical calculations.

[0091] <Summary> As described above, the data explanation device 10 according to this embodiment generates and outputs explanations for data relating to the target equipment using a generation AI. At this time, the data explanation device 10 according to this embodiment generates multiple explanations for the data and then creates an integrated explanation by combining these multiple explanations. As a result, the user of the data explanation device 10 according to this embodiment can make various decisions by referring to this explanation.

[0092] The present invention is not limited to the embodiments specifically disclosed above, and various modifications, changes, and combinations with known technologies are possible without departing from the scope of the claims. [Explanation of Symbols]

[0093] 10 Data Explanation Device 101 Input Device 102 Display device 103 External I / F 103a Recording medium 104 Communication I / F 105 RAM 106 ROM 107 Auxiliary storage 108 processors 109 Bus 201 Explanation Type Selection Section 202 Data Acquisition Unit 203 Information Organizing Department 204 Description Generation Processing Unit 205 Explanation Output Section 206 User Input Acquisition Unit 211 Prompt generation unit 212 Explanation Generation Unit 213 Integration Department 301 Data Storage Unit

Claims

1. A selection unit that selects one or more types as the type of description to be generated for data related to equipment, A data acquisition unit that acquires the data according to the one or more types selected, An explanation generation unit that generates one or more explanations corresponding to each of the one or more categories based on the acquired data and a generative AI realized by a machine learning model including a large-scale language model, An integration unit that creates a first integrated explanation representing an explanation that combines the one or more explanations corresponding to each of the one or more categories, An information processing device having

2. For each of the one or more types mentioned above, there is a creation unit that creates a prompt corresponding to the type using the data corresponding to the type, The explanation generation unit, The information processing apparatus according to claim 1, wherein the prompt is input to the generating AI to generate an explanation corresponding to the type.

3. The above one or more categories include a first category that describes problems with the current operation of the equipment, a second category that describes measures to improve those problems, a third category that describes a comparison and consideration between the current operation of the equipment and the optimal operation plan for the equipment, and a fourth category that describes the reasons for the optimal operation plan for the equipment. The explanation generation unit, The information processing apparatus according to claim 2, which creates prompts in order from the first type to the fourth type and generates descriptions corresponding to the types.

4. The explanation generation unit, By inputting the aforementioned prompt multiple times into the generating AI, multiple explanations corresponding to the aforementioned type are generated. The aforementioned integration unit is A second integrated explanation is created that represents an explanation that combines multiple explanations corresponding to the aforementioned categories. The information processing apparatus according to claim 2, which creates the first integrated description by integrating the second integrated descriptions corresponding to each of the one or more types mentioned above.

5. The explanation generation unit, By inputting the aforementioned prompt into multiple generating AIs, multiple explanations corresponding to the aforementioned type are generated. The aforementioned integration unit is A second integrated explanation is created that represents an explanation that combines multiple explanations corresponding to the aforementioned categories. The information processing apparatus according to claim 2, which creates the first integrated description by integrating the second integrated descriptions corresponding to each of the one or more types mentioned above.

6. The aforementioned integration unit is An information processing device according to any one of claims 1 to 5, which uses generative AI or document summarization technology to create an integrated explanation of multiple explanations.

7. A selection procedure for selecting one or more types of descriptions to be generated for data related to equipment, A data acquisition procedure for acquiring the data according to the one or more types selected, An explanation generation procedure that generates one or more explanations corresponding to each of the one or more categories based on the acquired data and a generative AI implemented with a machine learning model including a large-scale language model, A procedure for creating a first integrated description that represents an integrated description of the one or more descriptions corresponding to each of the one or more types mentioned above, A method of information processing performed by a computer.

8. A selection procedure for selecting one or more types of descriptions to be generated for data related to equipment, A data acquisition procedure for acquiring the data according to the one or more types selected, An explanation generation procedure that generates one or more explanations corresponding to each of the one or more categories based on the acquired data and a generative AI implemented with a machine learning model including a large-scale language model, A procedure for creating a first integrated description that represents an integrated description of the one or more descriptions corresponding to each of the one or more types mentioned above, A program that causes a computer to execute something.

Images