Information processing method, information processing device, and computer program

Abstract

Provided are an information processing method, an information processing device, and a computer program for outputting a response corresponding to a request of a user with respect to an analysis device.　In the present invention, a query pertaining to an analysis device is acquired, a language generation model is used to acquire a response pertaining to control of the analysis device corresponding to the query, and the response is output.

Description

Information processing method, information processing device, and computer program

[0001] The present invention relates to an information processing method, an information processing device, and a computer program for sample analysis.

[0002] There are various analytical devices for analyzing samples. For example, analytical devices include an X-ray analytical device that analyzes X-rays emitted from a sample to identify components contained in the sample, or a Raman analytical device that analyzes Raman light emitted from a sample. An example of an X-ray analytical device is disclosed in Patent Document 1. When an analytical device is made to perform some kind of analysis, the analytical device needs to be appropriately controlled.

[0003] JP 2023-74559 A

[0004] In some cases, such as when a user wants to perform an analysis but does not know how to operate the analyzer, it can be difficult to control the analyzer in accordance with the user's requests. Therefore, a technology is desired that allows the user to obtain information regarding the control of the analyzer in accordance with the user's requests.

[0005] An object of the present invention is to provide an information processing method, an information processing device, and a computer program for obtaining information relating to the control of an analytical device in response to a user's request.

[0006] An information processing method according to one embodiment of the present invention is characterized in that it acquires a query related to an analysis device, uses a language generation model to acquire a response related to control of the analysis device in response to the query, and outputs the response.

[0007] In one aspect of the present invention, a query including a user's request regarding an analysis device is acquired, and a response regarding control of the analysis device in response to the query is acquired using a language generation model, and the response is output. This allows a response in response to the user's request to be output. For example, in response to acquiring a query including the content of a process that the user wants the analysis device to execute, a response for causing the analysis device to execute the process that the user wants is output.

[0008] An information processing method according to one embodiment of the present invention is characterized in that it stores command data recording the specifications of commands for causing the analysis device to execute processing, and uses the language generation model to obtain a response containing a command for causing the analysis device to execute a specific process in accordance with the query and the command data, and inputs the command contained in the response to the analysis device.

[0009] In one aspect of the present invention, a response including a command is obtained in response to a query and command data recording specifications of a command for causing an analytical device to execute a process, and the command is input to the analytical device. The analytical device executes the process in response to the command, thereby performing the process desired by the user.

[0010] An information processing method according to one embodiment of the present invention is characterized in that it determines whether a command included in the response is a command that can be executed by the analysis device, and if the command included in the response is a command that cannot be executed by the analysis device, it uses the language generation model to obtain a new response that includes a new command in accordance with the query, the command data, and the command that cannot be executed by the analysis device.

[0011] In one aspect of the present invention, it is determined whether a command included in the response is executable by the analysis device. If the command is executable by the analysis device, processing according to the command is executed by the analysis device. If the command is not executable by the analysis device, a new response including a new command is output using a language generation model. This allows the analysis device to reliably execute processing according to the command.

[0012] An information processing method according to an aspect of the present invention is characterized in that the command data is modified depending on whether a command included in the response is a command that can be executed by the analytical device.

[0013] In one aspect of the present invention, command data is modified depending on whether a command included in a response is executable by an analytical device. The command data reflects whether the command is executable by the analytical device. By using the modified command data, a command for controlling the analytical device is efficiently obtained.

[0014] An information processing method according to one aspect of the present invention includes obtaining a processing result of the analysis device in accordance with a command input to the analysis device, and outputting the processing result.

[0015] In one aspect of the present invention, the results of processing performed by the analyzer in accordance with the command are output, allowing the user to confirm whether the desired processing has been performed.

[0016] An information processing method according to one embodiment of the present invention is characterized in that it acquires, as the query, a query including a request for creation of a report on analysis results by the analysis device, acquires the analysis results by the analysis device and a format of the report, uses the language generation model to acquire a report on the analysis results created in accordance with the format in response to the query, the analysis results, and the format, and outputs the report.

[0017] In one embodiment of the present invention, a language generation model is used to obtain a report in response to a query including a request for the creation of a report on the analysis results by the analysis device and a report format. The analysis results are output in the form of a report whose contents are organized according to the format, allowing a user to confirm the analysis results by the analysis device.

[0018] An information processing method according to one embodiment of the present invention includes acquiring a plurality of analysis results from a plurality of analysis devices, acquiring as the query a query including a request for creating a report that integrates the plurality of analysis results, and using the language generation model to acquire a report that integrates the plurality of analysis results created in accordance with the format, depending on the query, the acquired plurality of analysis results, and the format.

[0019] In one aspect of the present invention, multiple analysis results are obtained by multiple analysis devices, and a language generation model is used to obtain a report that integrates the multiple analysis results in accordance with a query including a request for report creation and a report format. The contents of the multiple analysis results are output in the form of a report organized according to the format. A user can confirm the multiple analysis results obtained by the multiple analysis devices in an easy-to-understand format.

[0020] An information processing method according to one aspect of the present invention is characterized in that it acquires the type of the analysis device, and uses the language generation model to acquire the response according to the query and the type of the analysis device.

[0021] In one aspect of the present invention, a language generation model is used to obtain a response according to the type of analysis device, so that even if the type of analysis device is different, the analysis device can execute the processing desired by the user.

[0022] An information processing method according to one embodiment of the present invention is characterized in that it uses the language generation model to obtain the response including a notification to a user regarding control of the analysis device in response to the query, and outputs the notification.

[0023] In one aspect of the present invention, a language generation model is used to obtain a response including a notification to a user regarding control of an analytical device, and the notification is output. The user can use the notification to cause the analytical device to execute a process desired by the user, such as operating the analytical device in accordance with the output notification.

[0024] An information processing method according to one embodiment of the present invention is characterized in that it acquires measurement data consisting of measurements taken by a sensor provided in the analysis device, and uses the language generation model to acquire the response in accordance with the query and the measurement data.

[0025] In one aspect of the present invention, a language generation model is used to obtain a response related to control of the analysis device in accordance with measurement data from a sensor provided in the analysis device, thereby enabling the analysis device to execute processing in accordance with the measurement results of the sensor.

[0026] An information processing method according to one aspect of the present invention includes acquiring an image, and using the language generation model to acquire the response corresponding to the query and the image.

[0027] In one embodiment of the present invention, a language generation model is used to obtain a response related to control of an analytical device in response to an image, thereby enabling the analytical device to be controlled in response to a user's request expressed using not only natural language but also images.

[0028] An information processing method according to one embodiment of the present invention is characterized in that, as the image, a photographed image of a sample to be analyzed by the analytical device is obtained, and the response including information for causing the analytical device to analyze the sample is obtained using the language generation model in accordance with the query and the photographed image.

[0029] In one aspect of the present invention, a language generation model is used to obtain a response related to control of an analytical device in accordance with a photographed image of a sample. The analytical device can be controlled in accordance with a request from a user using the photographed image, such as specifying a specific point on the sample as a measurement point.

[0030] An information processing method according to one aspect of the present invention is characterized in that it acquires a processing result from the analysis device, and uses the language generation model to acquire the response in accordance with the query and the processing result.

[0031] In one aspect of the present invention, a processing result of an analysis device is acquired, and a response related to control of the analysis device is acquired according to the processing result using a language generation model, thereby allowing the analysis device to execute an appropriate next process according to the processing result of the analysis device.

[0032] An information processing method according to one embodiment of the present invention is characterized in that it obtains an evaluation from a user of the output response, and trains the language generation model based on training data including the query, the response, and the evaluation.

[0033] In one aspect of the present invention, user evaluations of responses obtained using a language generation model are obtained, and the evaluations are used to train the language generation model. By using the trained language generation model, more appropriate responses can be obtained.

[0034] An information processing method according to one embodiment of the present invention is characterized in that it stores text data related to the analysis device, obtains related information related to the query from the text data, and uses the language generation model to obtain the response in accordance with the query and the extracted related information.

[0035] In one aspect of the present invention, related information related to a query is obtained from text data related to an analysis device, such as a manual, and a response corresponding to the query and the related information is obtained using a language generation model. The response obtained using the language generation model is a response generated based on the related information, and therefore is a response related to the analysis device. For example, a notification to the user included in the response is a notification related to the analysis device, such as a method of operating the analysis device.

[0036] An information processing device according to one embodiment of the present invention includes a calculation unit that acquires a query related to an analysis device, uses a language generation model to acquire a response related to control of the analysis device in response to the query, and outputs the response.

[0037] A computer program according to one embodiment of the present invention is characterized in that it causes a computer to execute a process of acquiring a query related to an analysis device, using a language generation model to acquire a response related to control of the analysis device in response to the query, and outputting the response.

[0038] In one aspect of the present invention, an information processing device acquires a query including a user's request regarding an analysis device, and uses a language generation model to acquire a response regarding control of the analysis device in response to the query, and outputs the response. The information processing device executes processing according to a computer program. In response to the query including the content of the processing that the user wants the analysis device to perform, a response is output for causing the analysis device to perform the processing that the user wants. The output response can be used to cause the analysis device to perform the processing that the user wants.

[0039] The present invention has excellent effects such as making it possible to obtain information relating to the control of an analytical device in accordance with the user's requests.

[0040] 1 is a schematic diagram showing an example of the configuration of an information processing system according to a first embodiment for controlling an analytical device. FIG. 2 is a block diagram showing an example of the functional configuration of an analytical device. FIG. 3 is a block diagram showing an example of the internal configuration of a terminal device. FIG. 4 is a block diagram showing an example of the internal configuration of an information processing device according to the first embodiment. FIG. 5 is a diagram showing example content of command data. FIG. 6 is a flowchart showing a first example of a processing procedure executed by the information processing device according to the first embodiment to control an analytical device. FIG. 7 is a flowchart showing a second example of a processing procedure executed by the information processing device to control an analytical device. FIG. 8 is a flowchart showing a third example of a processing procedure executed by the information processing device to control an analytical device. FIG. 9 is a flowchart showing a fourth example of a processing procedure executed by the information processing device to control an analytical device. FIG. 10 is a conceptual diagram showing an example of an analysis result. FIG. 11 is a conceptual diagram showing example content of a report format. FIG. 12 is a flowchart showing an example of a processing procedure executed by the information processing system 100 to create a report on the analysis results in the analytical device. FIG. 13 is a schematic diagram showing an example of a report on the analysis results. FIG. 14 is a flowchart showing an example of a learning processing procedure executed by the information processing device. FIG. 15 is a schematic diagram showing an example of the configuration of an information processing system according to a second embodiment. FIG. 16 is a block diagram showing an example of the internal configuration of an information processing device according to the second embodiment.

[0041] The present invention will be described in detail below with reference to the drawings illustrating embodiments. <Embodiment 1> In this embodiment, an analytical device is controlled in accordance with a user's request using a language generation model. FIG. 1 is a schematic diagram showing an example configuration of an information processing system 100 according to embodiment 1 for controlling an analytical device. The information processing system 100 includes a terminal device 2 used by a user 5, an information processing device 1, and an analytical device 3. The information processing device 1 executes an information processing method. The terminal device 2 communicates with the information processing device 1 via a first communication network 41, such as a local area network (LAN) or the Internet. The terminal device 2 also communicates with the analytical device 3 via a second communication network 42. The analytical device 3 is an instrument that measures phenomena occurring in a sample and analyzes the measurement results. The analytical device 3 is, for example, an X-ray analytical device.

[0042] FIG. 2 is a block diagram showing an example of the functional configuration of the analysis apparatus 3. FIG. 1 shows an example in which the analysis apparatus 3 is an X-ray analysis apparatus. The analysis apparatus 3 includes a sample stage 36 on which the sample 6 is placed, an irradiation unit 35 that irradiates the sample 6 with X-rays, an imaging unit 37 that images the sample 6, and a radiation detector 34. The irradiation unit 35 includes an X-ray tube that generates X-rays. The imaging unit 37 has an optical system and an image sensor. Radiation is irradiated from the irradiation unit 35 to the sample 6, and fluorescent X-rays are generated in the sample 6. The fluorescent X-rays are incident on the radiation detector 34, which detects the incident fluorescent X-rays. In the figure, arrows indicate the X-rays irradiating the sample 6 and the fluorescent X-rays incident on the radiation detector 34.

[0043] The radiation detector 34 includes a radiation detection element 341 and a temperature sensor 342. The radiation detection element is an element made of semiconductor. When fluorescent X-rays are incident on the radiation detection element 341, the radiation detection element 341 outputs a signal corresponding to the energy of the incident fluorescent X-rays. When fluorescent X-rays from the sample 6 are incident on the radiation detector 34, the radiation detector 34 detects the fluorescent X-rays by outputting a signal from the radiation detection element 341 corresponding to the incident fluorescent X-rays. The temperature sensor 342 measures the temperature inside the radiation detector 34. The temperature sensor 342 is made of, for example, a thermistor or a thermocouple.

[0044] The analysis device 3 includes a sample chamber 38. The sample chamber 38 is box-shaped and has an openable and closable lid 381. The sample chamber 38 is sealed when the lid 381 is closed. The sample stage 36, the irradiation unit 35, the imaging unit 37, and the radiation detector 34 are disposed inside the sample chamber 38. The sample 6 is placed inside the sample chamber 38 by being placed on the sample stage 36. With the lid 381 open, the user 5 places the sample 6 on the sample stage 36 or removes the sample 6. With the lid 381 closed, the inside of the sample chamber 38 is depressurized, and the sample 6 is irradiated with X-rays and the fluorescent X-rays are detected. The irradiation unit 35 may be configured to irradiate the sample 6 placed inside the sample chamber 38 with radiation from outside the sample chamber 38. The imaging unit 37 may be configured to image the sample 6 placed inside the sample chamber 38 from outside the sample chamber 38.

[0045] A signal processing unit 33 is connected to the radiation detector 34. The signal processing unit 33 receives the signal output by the radiation detector 34 and detects the intensity of the signal, thereby detecting a signal value corresponding to the energy of the radiation detected by the radiation detector 34. The signal processing unit 33 counts the signals for each signal value and outputs data indicating the relationship between the signal value and the count number.

[0046] The signal processing unit 33 is connected to the analysis unit 32. The analysis unit 32 is configured using a computer. The analysis unit 32 receives data indicating the relationship between the signal value output by the signal processing unit 33 and the count number. The analysis unit 32 generates a spectrum of the fluorescent X-rays incident on the radiation detector 34 based on the data from the signal processing unit 33. The spectrum indicates the relationship between the energy and intensity of the fluorescent X-rays. The process of counting the signals output by the radiation detector 34 by signal value may be performed by the analysis unit 32 instead of the signal processing unit 33. The generation of the fluorescent X-ray spectrum may be performed by the signal processing unit 33. The analysis unit 32 stores spectral data representing the fluorescent X-ray spectrum. In this way, the analysis device 3 measures the fluorescent X-rays. The analysis unit 32 may further perform information processing based on the radiation spectrum. For example, the analysis unit 32 performs qualitative or quantitative analysis of elements contained in the sample 6. The analysis unit 32 stores the analysis results.

[0047] A drive unit 361 that drives the sample stage 36 is connected to the sample stage 36. The drive unit 361 is configured using, for example, a stepping motor. The sample stage 36 moves when the drive unit 361 drives the sample stage 36. For example, the sample stage 36 moves in the horizontal direction. When the sample stage 36 moves, the sample 6 placed on the sample stage 36 moves.

[0048] The analysis device 3 includes a control unit 31. The control unit 31 is configured using a computer. The control unit 31 is connected to the irradiation unit 35, the imaging unit 37, the radiation detector 34, the signal processing unit 33, the analysis unit 32, and the drive unit 361. The control unit 31 controls the operations of the irradiation unit 35, the imaging unit 37, the radiation detector 34, the signal processing unit 33, the analysis unit 32, and the drive unit 361. The control unit 31 also communicates with the terminal device 2 via a second communication network 42. The control unit 31 and the analysis unit 32 may be integrated. The configuration of the analysis device 3 shown in FIG. 2 is an example, and the analysis device 3 may have other configurations.

[0049] FIG. 3 is a block diagram showing an example of the internal configuration of the terminal device 2. The terminal device 2 is a computer used by a user 5, such as a smartphone, tablet computer, or personal computer. The terminal device 2 includes a calculation unit 21, a memory 22, a storage unit 23, a reading unit 24, an operation unit 25, a display unit 26, a first communication unit 27, and a second communication unit 28. The calculation unit 21 is a processor, and is configured using, for example, a central processing unit (CPU), a graphics processing unit (GPU), or a multi-core CPU. The calculation unit 21 may also be configured using a quantum computer. The memory 22 stores temporary data generated in connection with calculations. The memory 22 is, for example, a random access memory (RAM). The reading unit 24 reads information from a recording medium 20, such as an optical disc or a portable memory. The storage unit 23 is non-volatile, and is, for example, a hard disk or non-volatile semiconductor memory.

[0050] The operation unit 25 receives input of information such as text by receiving operations from the user 5. The operation unit 25 is, for example, a touch panel, a keyboard, or a pointing device. The display unit 26 displays images. The display unit 26 is, for example, a liquid crystal display or an EL display (Electroluminescent Display). The operation unit 25 and the display unit 26 may be integrated. The first communication unit 27 communicates with the information processing device 1 via a first communication network 41 by wired or wireless communication. The second communication unit 28 communicates with the control unit 31 of the analysis device 3 via a second communication network 42 by wired or wireless communication.

[0051] The calculation unit 21 causes the reading unit 24 to read a computer program (program product) 231 recorded on the recording medium 20, and stores the read computer program 231 in the storage unit 23. The calculation unit 21 executes processing for realizing the functions of the terminal device 2 in accordance with the computer program 231. The computer program 231 may be stored in the storage unit 23 in advance, or may be downloaded from outside the terminal device 2. In this case, the terminal device 2 does not need to be equipped with the reading unit 24.

[0052] The computer program 231 can be deployed to run on a single computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communications network, i.e., the terminal device 2 may be made up of multiple computers, and the computer program 231 may run on multiple computers connected via a communications network.

[0053] FIG. 4 is a block diagram showing an example of the internal configuration of the information processing device 1 according to the first embodiment. The information processing device 1 is configured using a computer such as a server device. The information processing device 1 includes a calculation unit 11, a memory 12 for storing temporary data generated during calculation, a storage unit 13, a reading unit 14, and a communication unit 15. The calculation unit 11 is a processor, and is configured using, for example, a CPU, a GPU, or a multi-core CPU. The calculation unit 11 may also be configured using a quantum computer. The memory 12 is, for example, a RAM. The storage unit 13 is nonvolatile, and is, for example, a hard disk or nonvolatile semiconductor memory. The reading unit 14 reads information from a recording medium 10, such as an optical disk or a portable memory. The communication unit 15 communicates with the outside of the information processing device 1 via a first communication network 41 by wired or wireless communication. Specifically, the communication unit 15 communicates with a terminal device 2.

[0054] The calculation unit 11 causes the reading unit 14 to read a computer program (program product) 131 recorded on the recording medium 10, and stores the read computer program 131 in the storage unit 13. The calculation unit 11 executes processing to realize the functions of the information processing device 1 in accordance with the computer program 131. The computer program 131 may be stored in the storage unit 13 in advance, or may be downloaded from outside the information processing device 1. In this case, the information processing device 1 does not need to include the reading unit 14.

[0055] The computer program 131 can be deployed to run on a single computer, or on multiple computers located at one site or distributed across multiple sites and interconnected by a communications network. That is, the information processing device 1 may be configured with multiple computers, and the computer program 131 may be executed on multiple computers connected via a communications network. The information processing device 1 may be configured using a cloud server.

[0056] The processing of each step described below for executing the information processing method can be executed by multiple computers. The processing of each step can also be executed by different computers. The processing of each step can also be executed using a virtual machine. The processing of each step may be executed by multiple calculation units. The processing of each step may also be executed by different calculation units.

[0057] The information processing device 1 includes a language generation model 132. The language generation model 132 is a trained model that has been trained to output a sentence corresponding to a prompt when a prompt consisting of an instruction sentence requesting some kind of answer is input. Alternatively, the language generation model 132 is a trained model that has been trained to output a sentence corresponding to the input prompt and information when a prompt and information associated with the prompt are input. For example, the language generation model 132 is a large-scale language model. More specifically, the language generation model 132 is BERT, GPT-4, Bard, LLaMA, or the like. The language generation model 132 has been trained in advance.

[0058] The language generation model 132 is realized by the calculation unit 11 executing information processing in accordance with the computer program 131. The storage unit 13 stores data for realizing the language generation model 132. The language generation model 132 may be configured with hardware. The language generation model 132 may be realized using a quantum computer. Alternatively, the language generation model 132 may be provided outside the information processing device 1, and the information processing device 1 may execute processing using the external language generation model 132. The language generation model 132 may be realized using multiple computers connected via a communication network, or may be realized using the cloud.

[0059] The memory unit 13 stores command data 133. The command data 133 records the specifications of commands for causing the analytical device 3 to execute a process. FIG. 5 is a diagram showing an example of the content of the command data 133. The command data 133 describes in text the specifications of commands for causing the analytical device 3 to execute a specific process. The command data 133 is determined in accordance with the specifications of the analytical device 3. When a command is input, the analytical device 3 executes a process according to the command. Information about multiple commands is recorded in the command data 133. FIG. 5 shows an example in which the command data 133 records a command MECR for adding or removing a measurement point to or from the analytical device 3, and a command STSR for acquiring the status of the analytical device 3. A measurement point is a point on the sample 6 that is irradiated with X-rays, and fluorescent X-rays generated from the measurement point are measured. Information about other commands is also recorded in the command data 133.

[0060] The memory unit 13 stores text data 134. The text data 134 records text related to the analysis device 3. For example, the text data 134 includes text describing how to operate the analysis device 3, an explanation of information output from the analysis device 3, how to handle samples 6 used in the analysis device 3, or how to maintain the analysis device 3. For example, the text data 134 includes a manual for the analysis device 3 or a collection of troubleshooting tips related to the analysis device 3.

[0061] The information processing system 100 controls the analysis device 3 using the language generation model 132. Fig. 6 is a flowchart showing a first example of a processing procedure executed by the information processing device 1 according to the first embodiment to control the analysis device 3. Hereinafter, steps are abbreviated as S. The calculation unit 11 executes information processing according to the computer program 131, thereby causing the information processing device 1 to execute the following processing. Furthermore, the calculation unit 21 executes information processing according to the computer program 231, thereby causing the terminal device 2 to execute processing.

[0062] The information processing device 1 acquires a query related to the analysis device 3 (S101). In S101, the user 5 operates the operation unit 25 to input the query to the terminal device 2. The calculation unit 21 causes the first communication unit 27 to transmit the input query to the information processing device 1 via the first communication network 41. The information processing device 1 acquires the query by receiving the query transmitted from the terminal device 2 via the communication unit 15.

[0063] The query contains the details of the process that the user 5 wants the analytical device 3 to execute. For example, the query is data written in natural language. An example of a query to execute a specific process in the analytical device 3 is shown below: "Generate Python code to add a measurement point. The faster the measurement, the better. Automatic current control. Leave the rest to us. Wait until the device is in an idle state. When it is in an idle state, call print("finished.") and exit." This query requests the generation of code to cause the analytical device 3 to execute the process of adding a measurement point.

[0064] The query may include content requesting notification to the user 5. For example, if the operating method for instructing the analytical device 3 to execute a desired process is unknown, a query requesting notification of the operating method is input. An example of a query requesting notification to the user 5 in natural language is shown below: "What is spot size selection? What are the advantages of each spot? Which one should I choose?" This query requests notification including a method for selecting the size of the spot for irradiating the sample 6 with X-rays. In this way, a query including the user 5's request regarding the analytical device 3 is acquired.

[0065] The information processing device 1 acquires related information related to the acquired query from the text data 134 (S102). In S102, the calculation unit 11 searches the text data 134 for related information related to the acquired query and acquires the searched related information. For example, the text data 134 is divided into multiple parts, such as by page, paragraph, or item, and the calculation unit 11 extracts keywords included in the query, searches the text data 134 for parts including the keywords, and acquires the parts including the keywords as related information. For example, the text data 134 is vectorized, and the calculation unit 11 calculates similarities, such as cosine similarities, between the vectorized keywords and each part of the vectorized text data 134 and acquires the part with the greatest similarity as related information.

[0066] The information processing device 1 then creates a prompt including the acquired query, the acquired related information, and command data 133 (S103). In S103, the calculation unit 11 reads the command data 133 from the storage unit 13 and creates a prompt including the query, the related information, and the command data 133. The calculation unit 11 creates a prompt requesting that a response including a command for causing the analysis device 3 to execute processing according to the query and a notification to the user 5 be created by referring to the related information and the command data 133. For example, the storage unit 13 stores a fixed phrase included in the prompt, and the calculation unit 11 creates a prompt including the fixed phrase, the query, the related information, and the command data 133. For example, the fixed phrase includes information specifying the role of the language generation model 132 and the output format of the command.

[0067] An example of a boilerplate prompt is shown below: "You are a generator that generates Python scripts based on what the user says in natural language. The script generation should be as simple as possible. It is not a Python method, but rather a code-based executable. There is no need to explain the generated code; just provide the code. There is no need to explain arguments or variables; just provide the code. Commands must start with the command name in ASCII format, followed by the entire parameter list, with words separated by commas. For example, the complete command for the "ABC" command would be "ABC,1,2,3,4". The response command (if defined) would be "ABC,2,3,4,5,...". Because this is just a string, it must be split to obtain each parameter. The command type is always placed at the beginning. Other parameters follow the command type. In Python, your response must be in the command sending style. The method "SendCommand(string_)" is the command used to send the generated command. This method should be used only once per complete command text. The method "SendAndGetCommand(string _)" is a command that sends the generated command and retrieves the response value in string format if it exists." This template instructs the analyzer 3 to generate code that will execute some kind of process, and specifies the format of the generated code.

[0068] The information processing device 1 acquires a response regarding control of the analysis device 3 in accordance with the query, command data 133, and related information from the language generation model 132 (S104). In S104, the calculation unit 11 inputs the created prompt to the language generation model 132. The language generation model 132 performs calculations in response to the input prompt and outputs a response. Since the prompt includes a query regarding control of the analysis device 3, the language generation model 132 outputs a response regarding control of the analysis device 3. The calculation unit 11 acquires the response output by the language generation model 132. The calculation unit 11 stores the acquired response in the memory 12 or the storage unit 13.

[0069] Note that the information processing device 1 may create a prompt that does not include the command data 133 and the related information in S103, and then, in S104, perform processing to input the prompt with the command data 133 and the related information attached to it into the language generation model 132. In S103, the information processing device 1 may create a prompt that does not include the command data 133 and the related information, but that includes information for referencing the command data 133 and the related information, such as an address where the command data 133 and the related information are stored. Alternatively, in S104, the information processing device 1 may train the language generation model 132 using the command data 133 and the related information, and, after the training, input the prompt that does not include the command data 133 and the related information into the language generation model 132 to obtain a response.

[0070] The information processing device 1 outputs a notification to the user 5 included in the acquired response (S105). In S105, the calculation unit 11 extracts a notification to the user 5 from the acquired response and causes the communication unit 15 to transmit the extracted notification to the terminal device 2 via the first communication network 41. The terminal device 2 receives the notification to the user 5 via the first communication unit 27. The calculation unit 21 outputs the notification to the user 5 by displaying the received notification on the display unit 26. At this time, the calculation unit 21 generates an image including the notification to the user 5 and displays the generated image on the display unit 26. The user 5 checks the output notification.

[0071] The notification to the user 5 is included in the response generated in response to the related information, and is therefore a notification related to the analytical device 3. The notification to the user 5 is a notification instructing the user 5 to perform an operation on the analytical device 3 in order to control the analytical device 3, such as "Please turn on the analytical device" or "Please open the lid, place the sample on the sample stage, and close the lid." The notification to the user 5 is, for example, a notification requesting the input of further information necessary to control the analytical device 3.

[0072] The notification to the user 5 is a notification in response to the query. If the query includes content requesting a notification to the user 5, the requested notification is included in the response. For example, in response to a query requesting a notification explaining how to operate the analysis device 3, a notification explaining how to operate the analysis device 3 is output. An example of a notification in response to a query requesting a notification including how to select the spot size is shown below. "There are two options for selecting the spot size: 1 is 100 μm, and 2 is 7 mm. The advantages of each spot size are as follows: 1. A spot size of 100 μm improves spatial resolution and improves the detection limit for small particles and features on the sample surface. However, the measurement time may be longer and the X-ray absorption effect may be higher. 2. A spot size of 7 mm increases X-ray intensity and reduces measurement time. This is suitable for large samples or samples with high X-ray absorption. However, it sacrifices spatial resolution and detection limit. The spot size should be selected based on the sample size, sample features, and the trade-off between spatial resolution and measurement speed." In this example, a notification is obtained that explains in natural language how to select the spot size.

[0073] The user 5, who has confirmed the output notification, operates the analysis device 3 in accordance with the notification. By outputting a notification to the user 5, it is possible to control the analysis device 3 in a way that cannot be achieved by simply inputting a command to the analysis device 3. If the response does not include a notification to the user 5, the information processing device 1 omits the processing of S105.

[0074] The information processing device 1 determines whether the acquired response includes a command (S106). In S106, the calculation unit 11 determines whether the response includes a command based on the command data 133. The command here means an instruction to cause the analysis device 3 to execute some kind of processing. The command may be a single instruction or a statement including multiple instructions. The command may be a program or a script. The command may include binary code. If the response does not include a command (S106: NO), the information processing device 1 proceeds to S114, which will be described later.

[0075] If the response includes a command (S106: YES), the information processing device 1 determines whether the command included in the response is executable by the analysis device 3 (S107). In S107, the calculation unit 11 extracts the command included in the response and causes the communication unit 15 to transmit the command to the terminal device 2. The terminal device 2 receives the command via the first communication unit 27. The calculation unit 21 causes the second communication unit 28 to transmit the received command to the analysis device 3 via the second communication network 42. In the analysis device 3, the control unit 31 receives the command transmitted from the terminal device 2 and determines whether it is possible to execute processing in accordance with the received command. For example, the control unit 31 verifies whether the command is executable by trying to execute the command. The control unit 31 transmits the determination result of whether the command is executable to the terminal device 2, and the terminal device 2 transmits the determination result to the information processing device 1. The information processing device 1 receives the determination result via the communication unit 15. The information processing device 1 may perform processing to determine by itself whether or not a command can be executed by the analysis device 3 .

[0076] If the command included in the response is not executable by the analysis device 3 (S107: NO), the information processing device 1 creates a prompt corresponding to the command that is not executable by the analysis device 3 (S108). In S108, the calculation unit 11 creates a prompt including the query, related information, command data 133, and the command that is not executable. The calculation unit 11 creates a prompt requesting the analysis device 3 to create a response including a command different from the command that is not executable, in order to have the analysis device 3 execute processing corresponding to the query. For example, the calculation unit 11 specifies the command that is not executable and creates a prompt including a request to create a new command that does not include the specified command. Alternatively, the calculation unit 11 obtains an error message that occurs when the control unit 31 attempts to execute the command that is not executable, and creates a prompt including a request to create a new command that does not generate the obtained error message.

[0077] The information processing device 1 then returns the process to S104. The calculation unit 11 inputs the prompt created in S108 to the language generation model 132, the language generation model 132 outputs a response corresponding to the prompt, and the calculation unit 11 acquires the response. The command included in the response is a command different from the command that cannot be executed by the analysis device 3. In this way, the information processing device 1 acquires a new response including a new command in accordance with the query, related information, command data 133, and the command that cannot be executed by the analysis device 3.

[0078] If the command included in the response is executable by the analysis device 3 (S107: YES), the information processing device 1 modifies the command data 133 according to the determination result of whether the command is executable by the analysis device 3 (S109). In S109, for example, the calculation unit 11 adds information indicating that the command is executable to a command determined to be executable by the analysis device 3 among the commands recorded in the command data 133. For example, the calculation unit 11 adds information indicating that the command is prohibited from use to a command determined to be unexecutable by the analysis device 3 among the commands recorded in the command data 133. For example, the calculation unit 11 deletes from the command data 133 the command determined to be unexecutable by the analysis device 3.

[0079] After the command data 133 is corrected, the information processing device 1 uses the language generation model 132 to acquire a response including a command corresponding to the corrected command data 133. The corrected command data 133 reflects whether the command is executable by the analysis device 3, so that the response is more likely to include a command that is executable by the analysis device 3. This makes it possible to efficiently acquire commands for controlling the analysis device 3. Note that the processing of S109 may be executed when it is determined that the command included in the response is not executable by the analysis device 3. The processing of S109 may be omitted.

[0080] The information processing device 1 then inputs the command included in the response to the analysis device 3 (S110). In S110, the calculation unit 11 causes the communication unit 15 to transmit the command included in the response to the terminal device 2. The terminal device 2 receives the command and transmits it to the analysis device 3. In the analysis device 3, the control unit 31 receives the command transmitted from the terminal device 2. In this way, the information processing device 1 inputs the command included in the response to the analysis device 3, thereby outputting the response acquired from the language generation model 132. Note that in S110, the analysis device 3 may execute processing in accordance with the command received in S107 without inputting the command again to the analysis device 3.

[0081] The control unit 31 controls each part of the analysis device 3 in accordance with the commands. The analysis device 3 executes processing in accordance with the commands by operating each part of the analysis device 3 under the control of the control unit 31. For example, the analysis device 3 moves the sample 6, irradiates the sample 6 with X-rays, detects the generated fluorescent X-rays, and generates a fluorescent X-ray spectrum. The analysis unit 32 stores the processing results. For example, the analysis unit 32 stores a fluorescent X-ray spectrum as the processing result. For example, the analysis device 3 continuously or intermittently measures a physical quantity such as the intensity of fluorescent X-rays, and the processing result is time-series data representing the result of continuously or intermittently measuring the physical quantity.

[0082] The command included in the response output by the language generation model 132 is a command corresponding to the query. The command is a command for executing the process that the query has requested the analysis device 3 to execute. An example of a command corresponding to a query requesting the analysis device 3 to execute the process of adding a measurement point is shown below: "# MECR,1,2,30,1 SendCommand("MECR,1,2,30,1") # wait until machine status is idle status = "" while status != "0": reply = SendAndGetCommand("STSR") params = reply.split(",") status = params[1] # print finished if machine is idle if status == "0": print("finished.")" In this example, a command is obtained that causes the analysis device 3 to execute the process of adding a measurement point.

[0083] The information processing device 1 acquires the processing results of the analysis device 3 (S111). The control unit 31 transmits the processing results of the analysis device 3 to the terminal device 2. The terminal device 2 receives the processing results transmitted from the analysis device 3 and transmits them to the information processing device 1. The information processing device 1 receives the processing results via the communication unit 15. The calculation unit 11 stores the received processing results in the memory unit 13. In this way, the information processing device 1 acquires the processing results of the analysis device 3.

[0084] The information processing device 1 outputs the processing results of the analysis device 3 (S112). In S112, the calculation unit 11 causes the communication unit 15 to transmit the processing results to the terminal device 2. The terminal device 2 receives the processing results, and the calculation unit 21 outputs the processing results by displaying the received processing results on the display unit 26. At this time, the calculation unit 21 generates an image including the processing results and displays the generated image on the display unit 26. For example, a fluorescent X-ray spectrum is displayed on the display unit 26 as the processing result. For example, a graph representing time-series data is displayed as the processing result. The user 5 checks the output processing results of the analysis device 3. Having checked the processing results, the user 5 can know whether the processing that the user 5 wanted the analysis device 3 to perform was actually performed.

[0085] The information processing device 1 acquires an evaluation from the user 5 (S113). In S113, the user 5 operates the operation unit 25, and the terminal device 2 acquires an evaluation on the result of the information processing device 1 performing a process of outputting a response from the language generation model 132. For example, the calculation unit 11 causes the communication unit 15 to transmit data of an input image for inputting the evaluation to the terminal device 2, the terminal device 2 receives the input image data, and the calculation unit 21 displays an input image based on the received data on the display unit 26. The user 5 inputs the evaluation to the terminal device 2 using the input image, and the terminal device 2 acquires the evaluation. The terminal device 2 transmits the acquired evaluation to the information processing device 1, and the information processing device 1 receives the evaluation. The calculation unit 11 stores the received evaluation from the user 5 in the storage unit 13.

[0086] The evaluation from the user 5 is, for example, information specifying whether the processing of the information processing device 1 is good or bad by selecting between two options. The evaluation from the user 5 is, for example, information specifying whether the processing is good or bad by more than two levels. The processing of S113 may be omitted.

[0087] The information processing device 1 determines whether to terminate the process for controlling the analysis device 3 (S114). In S114, for example, the user 5 operates the operation unit 25 to input an instruction to terminate to the terminal device 2, the terminal device 2 transmits the instruction to terminate to the information processing device 1, and if the information processing device 1 receives the instruction to terminate, the calculation unit 11 determines to terminate the process. If the instruction to terminate is not received, the calculation unit 11 determines not to terminate the process.

[0088] If it is determined that the process for controlling the analysis device 3 should not be terminated (S114: NO), the information processing device 1 returns the process to S101. In S101, the user 5 inputs a query based on the processing result. The information processing device 1 acquires the query and repeats the processes from S102 onwards. If it is determined that the process for controlling the analysis device 3 should be terminated (S114: YES), the information processing device 1 terminates the process for controlling the analysis device 3.

[0089] Although S101 to S114 include processing that utilizes related information acquired from the text data 134, the information processing device 1 may perform processing that does not utilize the related information. That is, the processing of S102 may be omitted. Although S101 to S114 include processing that acquires a response that includes a notification to the user 5, the information processing device 1 may perform processing that acquires a response that does not include a notification to the user 5. That is, the processing of S105 may be omitted.

[0090] In S101 to S114, the information processing device 1 acquires a response corresponding to the query, command data 133, and related information from the language generation model 132. The information processing device 1 may perform processing to acquire a response corresponding to the processing result in the analysis device 3. FIG. 7 is a flowchart showing a second example of the processing procedure executed by the information processing device 1 to control the analysis device 3.

[0091] The information processing device 1 acquires the processing results of the analysis device 3 (S21). In S21, the calculation unit 11 acquires the processing results by reading out the processing results of the analysis device 3 stored in the memory unit 13. At this time, the information processing device 1 may cause the terminal device 2 to output the processing results. The user 5 may operate the operation unit 25 to input an instruction to the terminal device 2 to acquire the processing results of the analysis device 3, and the information processing device 1 may execute the processing of S21 in response to receiving the instruction from the terminal device 2. For example, the processing results of the analysis device 3 are a spectrum of fluorescent X-rays. For example, the processing results of the analysis device 3 are time-series data.

[0092] The information processing device 1 then executes the process of S101. In S101, the information processing device 1 acquires a query corresponding to the processing result of the analysis device 3. For example, the user 5 inputs a query to the terminal device 2 to cause the analysis device 3 to execute processing under conditions changed from the obtained processing result. For example, the processing result is a spectrum of fluorescent X-rays emitted from a specific point on the sample 6, and a query is acquired instructing to acquire a spectrum of fluorescent X-rays emitted from a point near the specific point. For example, the processing result is time-series data, and a query is acquired instructing to perform analysis based on the time-series data.

[0093] The information processing device 1 then executes the processes of S102 to S114. In S103, the information processing device 1 creates a prompt that includes the query, related information, command data 133, and the acquired processing result. In the process of S104, the information processing device 1 acquires a response according to the query and the processing result in the analysis device 3 from the language generation model 132. In this way, the information processing device 1 can cause the analysis device 3 to execute an appropriate next process according to the processing result. In S103, the information processing device 1 may create a prompt that does not include the processing result, and in S104, perform a process of inputting the prompt with the processing result attached to it into the language generation model 132.

[0094] The information processing device 1 may perform processing to acquire a response corresponding to the image. FIG. 8 is a flowchart showing a third example of processing procedures executed by the information processing device 1 to control the analysis device 3. The information processing device 1 acquires an image (S22). In S22, the calculation unit 11 acquires the image by reading out an image stored in the memory unit 13. At this time, the information processing device 1 may cause the terminal device 2 to display the image. The user 5 may operate the operation unit 25 to input an instruction to acquire an image to the terminal device 2, and the information processing device 1 may execute the processing of S22 in response to receiving the instruction from the terminal device 2. For example, the image is an image representing a fluorescent X-ray spectrum or a mapping image showing the distribution of elements in the sample 6 obtained by fluorescent X-ray analysis.

[0095] The information processing device 1 then executes the process of S101. In S101, the information processing device 1 acquires a query corresponding to the image. For example, the user 5 inputs a query referring to the image to the terminal device 2. For example, a query inquiring about the cause of a specific peak included in the spectrum of fluorescent X-rays is acquired. For example, a query requesting to obtain the position within the sample 6 of a substance containing a specific element included in the mapping image is acquired.

[0096] The information processing device 1 then executes the processes of S102 to S114. In S103, the information processing device 1 creates a prompt including the query, related information, command data 133, and the acquired image. In the process of S104, the information processing device 1 acquires a response corresponding to the query and the image from the language generation model 132. In this process, the information processing device 1 uses a multimodal model such as GPT-4, which can use images, as the language generation model 132. In this way, the information processing device 1 can cause the analysis device 3 to execute processing corresponding to the image. In S103, the information processing device 1 may create a prompt that does not include an image, and in S104, perform a process of inputting the prompt with the attached image to the language generation model 132.

[0097] The image acquired by the information processing device 1 in S22 may be a photographed image of the sample 6. The user 5 operates the operation unit 25 to input an instruction to acquire the photographed image to the terminal device 2, and the terminal device 2 transmits the instruction to photograph to the analysis device 3. Under the control of the control unit 31 in accordance with the instruction to photograph, the photographing unit 37 photographs the sample 6 and creates a photographed image. The control unit 31 transmits the photographed image to the terminal device 2, and the terminal device 2 transmits the photographed image to the information processing device 1. The information processing device 1 receives the photographed image and thereby acquires the photographed image.

[0098] In S101, the information processing device 1 acquires a query corresponding to the captured image. For example, the user 5 inputs a query referring to the captured image to the terminal device 2. For example, a query requesting elemental analysis by specifying an arbitrary point on the sample 6 shown in the captured image and irradiating the specified point with X-rays is acquired. For example, a query requesting elemental analysis by specifying an arbitrary portion of the captured image and enlarging the specified portion is acquired. For example, a query requesting the position of an arbitrary portion included in the captured image within the sample 6 is acquired. By the processing of S104, the information processing device 1 acquires from the language generation model 132 a response including information for causing the analysis device 3 to execute a process for analyzing the sample 6 according to the query and the captured image. For example, a response including a command for causing the analysis device 3 to execute a process for elemental analysis by irradiating a specified point on the sample 6 with X-rays, or a notification guiding the analysis device 3 to execute the process, is acquired. In this way, the information processing device 1 can cause the analysis device 3 to execute a process according to the captured image of the sample 6.

[0099] As described above, the information processing device 1 acquires a query including a request from the user 5 regarding the analysis device 3, acquires a response regarding control of the analysis device 3 in response to the query from the language generation model 132, and outputs the acquired response. This outputs information regarding control of the analysis device 3 in response to the user's request. For example, in response to acquiring a query including the content of a process that the user 5 wants the analysis device 3 to execute, a response is output for causing the analysis device 3 to execute the process requested by the user 5. If the response includes a command for causing the analysis device 3 to execute the process, the analysis device 3 executes the process requested by the user 5 by executing the process in response to the command. If the response includes a notification to the user 5, the user 5 can take action, such as operating the analysis device 3, in accordance with the output notification. Even if the method of operating the analysis device 3 is unknown, a response regarding control of the analysis device 3 in response to the user 5's request is output. A command for causing the analysis device 3 to execute the process requested by the user 5, or a notification for causing the analysis device 3 to execute the process requested by the user 5, is output, causing the analysis device 3 to execute the process requested by the user 5.

[0100] When an analysis needs to be performed using an unusual method, such as when analyzing a special sample, a sample being analyzed for the first time, or an unknown sample using the analysis device 3, the method for operating the analysis device 3 may be unclear. Alternatively, different types of analysis devices 3 may have different operation methods. For example, even if the same X-ray analysis device is used, the operation method may differ depending on the manufacturer or manufacturing date. Even if the user 5 is familiar with operating one type of analysis device 3, the user 5 may not know how to operate another type of analysis device 3. Even in such cases, by inputting a query containing the content of the processing the user 5 wants the analysis device 3 to perform, a response for causing the analysis device 3 to perform the processing requested by the user 5 can be obtained. In this way, the analysis device 3 can perform the processing requested by the user 5.

[0101] Even if the method for operating or controlling the analysis device 3 is clear, a response related to the control of the analysis device 3 may be required in response to a request from the user 5. For example, there is a desire to automate an analysis procedure consisting of multiple steps, such as inputting commands to the analysis device 3, outputting processing results or errors from the analysis device 3, and re-inputting commands in response to the output. Creating a program for automation by combining multiple commands is usually difficult. In this embodiment, the user 5 inputs a query including the analysis procedure and a request to automate the analysis procedure, and the information processing device 1 can output a response including a combination of commands to automate the analysis procedure in response to the query. The analysis procedure can be automated by having the analysis device 3 execute processing in accordance with the output response.

[0102] The information processing device 1 may perform processing to acquire a response corresponding to the measurement data from the sensor. FIG. 9 is a flowchart showing a fourth example of processing procedures executed by the information processing device 1 to control the analysis device 3. The information processing device 1 acquires measurement data created using the temperature sensor 342 (S23). The user 5 operates the operation unit 25 to input an instruction to acquire the measurement data to the terminal device 2, and the terminal device 2 transmits the instruction to acquire the measurement data to the analysis device 3. The control unit 31 creates measurement data consisting of measurement values ​​obtained by measuring the temperature of the radiation detector 34 by the temperature sensor 342 in accordance with the instruction. The control unit 31 may store the measurement data created in advance. The control unit 31 transmits the measurement data to the terminal device 2, and the terminal device 2 transmits the measurement data to the information processing device 1. The information processing device 1 receives the measurement data to acquire the measurement data.

[0103] The information processing device 1 then executes the process of S101. In S101, the information processing device 1 acquires a query corresponding to the measurement data. For example, the user 5 inputs to the terminal device 2 a query requesting control of the temperature of the radiation detector 34 or a query requesting notification advising on an operation corresponding to the temperature.

[0104] The information processing device 1 then executes the processes of S102 to S114. In S103, the information processing device 1 creates a prompt including a query, related information, command data 133, and measurement data. In the process of S104, the information processing device 1 acquires a response corresponding to the query and measurement data from the language generation model 132. In this way, the information processing device 1 can cause the analysis device 3 to execute processing corresponding to the measured temperature of the radiation detector 34. In S103, the information processing device 1 may create a prompt that does not include measurement data, and in S104, perform processing to input the prompt accompanied by the measurement data to the language generation model 132.

[0105] The information processing system 100 can use the language generation model 132 to create a report on the analysis results of the analysis device 3. The analysis device 3 stores analysis results that record the results of analyses that have been performed so far in the control unit 31 or the analysis unit 32. The analysis results may be the results of a single process, or may be the results of analysis performed by the analysis unit 32 based on the results of multiple processes.

[0106] Fig. 10 is a conceptual diagram showing an example of an analysis result. In the example shown in Fig. 10, the analysis result includes a title, fluorescent X-ray spectrum, quantitative results, comments, the date of the experiment, the experimenter, and measurement conditions. The fluorescent X-ray spectrum is recorded in the form of image data, and in the example shown in Fig. 10, the file name of the image data is recorded.

[0107] The storage unit 13 stores a report format 135. The report format 135 is information that defines the format of a report regarding analysis results. FIG. 11 is a conceptual diagram showing an example of the contents of the report format 135. For example, the report format 135 specifies the data format of the report or the format of characters in the report. For example, the report format 135 specifies each item included in the report, such as the title, date, experimenter, or experimental conditions. For example, the report format 135 specifies the layout of each item included in the report. For example, the report format 135 specifies images included in a report such as a spectrum, or charts included in a report such as a table of quantitative results. For example, the report format 135 specifies the display format or layout of images or charts included in the report.

[0108] 12 is a flowchart showing an example of the processing procedure executed by the information processing system 100 to create a report related to the analysis results of the analysis device 3. The information processing device 1 acquires a query including a request for creating a report (S31). In S31, the user 5 operates the operation unit 25 to input the query to the terminal device 2, which then transmits the query to the information processing device 1. The information processing device 1 then receives the query transmitted from the terminal device 2, thereby acquiring the query. The query includes a request for creating a report related to the analysis results, such as "Please create a report applying the analysis results ***.** to a report format." The "***.**" included in the example query indicates the file name of the analysis results.

[0109] The information processing device 1 acquires the analysis results from the analysis device 3 (S32). In S32, the calculation unit 11 causes the communication unit 15 to send a request for the analysis results to the terminal device 2. At this time, the calculation unit 11 may send a request for a specific analysis result specified in the query. The terminal device 2 receives the request for the analysis results and sends it to the analysis device 3. In response to the request for the analysis results, the control unit 31 reads out the analysis results and sends them to the terminal device 2. The terminal device 2 receives the analysis results and sends them to the information processing device 1. The information processing device 1 acquires the analysis results by receiving them via the communication unit 15. The calculation unit 11 stores the acquired analysis results in the memory unit 13. Note that the information processing device 1 may store the analysis results in the memory unit 13 in advance and acquire the analysis results by reading them from the memory unit 13.

[0110] The information processing device 1 acquires the report format 135 (S33). In S33, the calculation unit 11 acquires the report format 135 by reading it from the storage unit 13. Note that multiple report formats 135 may be stored in the storage unit 13, and information specifying one of the report formats 135 may be included in the query, and the information processing device 1 may acquire the specific report format 135 specified in the query. Alternatively, the information processing device 1 may acquire the report format by acquiring a query including the report format in S31. In this case, the information processing device 1 does not need to perform the process of S33 separately from S31.

[0111] The information processing device 1 then creates a prompt including the acquired query, the acquired analysis results, and the report format 135 (S34). In S34, the calculation unit 11 creates a prompt requesting the creation of a report on the analysis results in accordance with the report format 135 in response to the query.

[0112] The information processing device 1 obtains a report from the language generation model 132 in response to the prompt (S35). In S35, the calculation unit 11 inputs the created prompt to the language generation model 132. The language generation model 132 performs calculations in response to the input prompt and outputs a response. Since the prompt requests the creation of a report on the analysis results, the language generation model 132 outputs a report on the analysis results as a response. The calculation unit 11 obtains the report output by the language generation model 132. The calculation unit 11 stores the obtained report in the memory unit 13.

[0113] Note that the information processing device 1 may create a prompt that does not include the analysis result and report format 135 in S34, and may perform processing to input the prompt accompanied by the analysis result and report format 135 into the language generation model 132 in S35. Alternatively, the information processing device 1 may create a prompt that does not include the analysis result and report format 135 but includes information for referencing the analysis result and report format 135, such as an address where the analysis result and report format 135 are stored, in S34.

[0114] The information processing device 1 outputs a report on the analysis results (S36). In S36, the communication unit 15 transmits the acquired report to the terminal device 2 via the first communication network 41. The terminal device 2 receives the report, and the calculation unit 21 outputs the report by displaying the received report on the display unit 26. At this time, the calculation unit 21 generates an image including the report and displays the generated image on the display unit 26.

[0115] 13 is a schematic diagram showing an example of a report on the analysis results. The report is configured, for example, in HTML (Hyper Text Markup Language) format, PDF (Portable Document Format), or as an image. The contents of the analysis results are arranged according to the report format 135, and are displayed in an easily viewable and organized manner. The user 5 checks the output report. Displaying the analysis results in the form of an appropriately organized report makes it easier for the user 5 to understand the analysis results obtained by the analysis device 3.

[0116] The information processing device 1 performs a process of training the language generation model 132 using the evaluation acquired from the user 5. The information processing device 1 stores training data 136 including the evaluation acquired in S113 in the storage unit 13. The training data 136 stores a query, a response output from the language generation model 132 in response to a prompt including the query, and an evaluation for the response, in association with each other. Each time an evaluation is acquired in S113, the calculation unit 11 generates the training data 136 by storing the query, the response, and the evaluation in association with each other in the storage unit 13. A combination of a query, a response, and an evaluation is treated as one dataset, and multiple datasets are recorded in the training data 136.

[0117] The information processing device 1 uses training data to train the language generation model 132. Fig. 14 is a flowchart showing an example of the procedure of the learning process performed by the information processing device 1. The calculation unit 11 reads out the training data 136 stored in the storage unit 13, and the information processing device 1 acquires the training data 136 (S41). In S41, the information processing device 1 may acquire the training data 136 from an external source.

[0118] The information processing device 1 then trains the language generation model 132 using the training data 136 (S42). In S42, the calculation unit 11 inputs a prompt including a query included in the training data 136 to the language generation model 132, and calculates the difference between the response output by the language generation model 132 and the response associated with the query in the training data 136. The calculation unit 11 adjusts the parameters of the language generation model 132 so that the difference is smaller when the rating associated with the query in the training data 136 is high and larger when the rating is low. The calculation unit 11 trains the language generation model 132 by repeating processing using multiple data sets recorded in the training data 136 and adjusting the parameters for the calculation of the language generation model 132.

[0119] The calculation unit 11 stores the final parameters adjusted in S42 in the storage unit 13. After S42 is completed, the information processing device 1 ends the learning process. The processes of S41 to S42 may be executed by a device other than the information processing device 1. After the language generation model 132 has been learned, the learned language generation model 132 is used to execute processes such as S101 to S114. By using the language generation model 132 that has been learned using evaluations from the user 5, it becomes possible to obtain a more appropriate response.

[0120] In this embodiment, the information processing device 1 and the terminal device 2 are different devices, but the information processing device 1 and the terminal device 2 may be integrated. In this embodiment, the terminal device 2 is separate from the analysis device 3, but the terminal device 2 may be part of the analysis device 3. For example, the control unit 31 or the analysis unit 32 may also have the functions of the terminal device 2.

[0121] In this embodiment, the analysis device 3 is provided with the temperature sensor 342, but the analysis device 3 may be provided with a sensor other than the temperature sensor 342. For example, the analysis device 3 may be provided with a sensor that measures the pressure or degree of vacuum in the sample chamber 38. Even if the analysis device 3 is provided with a sensor other than the temperature sensor 342, the information processing device 1 can execute the processes of S23 and S101 to S114 using the measurement data from the sensor.

[0122] The information processing device 1 may perform processing to acquire a response corresponding to the state of the analysis device 3, such as the degree of consumption of consumables, the degree of deterioration of each part of the analysis device 3, such as the radiation detector 34 or the irradiation unit 35, or the position of the sample stage 36 moved by the drive unit 361. In this case, the information processing device 1 may perform processing to acquire the state of the analysis device 3 by a method other than a method using a sensor. For example, the control unit 31 measures the cumulative operating time of the analysis device 3, the information processing device 1 acquires the cumulative operating time from the control unit 31, and determines the degree of consumption of consumables or the degree of deterioration of each part of the analysis device 3 based on the cumulative operating time. The control unit 31 may determine the state of the analysis device 3 based on the cumulative operating time, and the information processing device 1 may acquire the determined state of the analysis device 3 from the control unit 31. The control unit 31 may determine the position of the sample stage 36 from the control history of the drive unit 361, and the information processing device 1 may acquire the position of the sample stage 36 from the control unit 31. After acquiring the state of the analysis device 3, the information processing device 1 executes processing from S101 to S114 using the acquired state of the analysis device 3. At this time, in S101, the information processing device 1 acquires a query according to the state of the analysis device 3. Through the processes of S101 to S114, the information processing device 1 acquires a response according to the state of the analysis device 3.

[0123] In this embodiment, the analysis device 3 is an X-ray analysis device, but the analysis device 3 may be an analysis device other than an X-ray analysis device. For example, the analysis device 3 may be a Raman analysis device.

[0124] 15 is a schematic diagram showing an example of the configuration of an information processing system 100 according to the second embodiment. In the second embodiment, the information processing system 100 includes a plurality of terminal devices 2 and a plurality of analysis devices 3. The plurality of terminal devices 2 communicate with the information processing device 1 via a first communication network 41. Each of the terminal devices 2 communicates with a plurality of analysis devices 3 via a second communication network 42. The plurality of analysis devices 3 includes analysis devices 3 of different types. For example, some of the analysis devices 3 are X-ray analysis devices, and other portions of the analysis devices 3 are Raman analysis devices. The configuration of the terminal device 2 is the same as that of the first embodiment.

[0125] 16 is a block diagram showing an example of the internal configuration of an information processing device 1 according to the second embodiment. The information processing device 1 is configured using a computer such as a server device. As in the first embodiment, the information processing device 1 includes a calculation unit 11, a memory 12, a storage unit 13, a reading unit 14, and a communication unit 15. As in the first embodiment, the information processing device 1 includes a language generation model 132, and the storage unit 13 stores a computer program 131 and stores a report format 135 and training data 136. The storage unit 13 also stores a command DB (database) 137 and a sentence DB 138.

[0126] The command DB 137 includes a plurality of command data such as first command data and second command data. Each piece of command data included in the command DB 137 records the specifications of a command for causing an analysis device 3 to execute a process, similar to the command data 133 in the first embodiment. The command DB 137 includes command data for each type of analysis device 3. That is, each piece of command data included in the command DB 137 records the specifications of a command for causing one of the plurality of analysis devices 3 to execute a process.

[0127] The text DB 138 includes a plurality of text data such as first text data and second text data. Each piece of text data included in the text DB 138 records text related to the analysis device 3, such as a manual, similar to the text data 134 in the first embodiment. The text DB 138 includes text data for each type of analysis device 3. That is, each piece of text data included in the text DB 138 records text related to one of the plurality of analysis devices 3.

[0128] In the second embodiment as well, the information processing system 100 controls each analysis device 3 by utilizing the language generation model 132. Fig. 17 is a flowchart showing an example of a processing procedure executed by the information processing device 1 according to the second embodiment to control the analysis device 3.

[0129] The information processing device 1 acquires the type of analysis device 3 to be controlled and a query related to the analysis device 3 (S51). In S51, the user 5 operates the operation unit 25 to input a query to the terminal device 2, and the terminal device 2 transmits the query to the information processing device 1. The terminal device 2 also transmits the type of analysis device 3 to be controlled to the information processing device 1. For example, the terminal device 2 stores information indicating the type of analysis device 3 in the storage unit 23 and transmits the stored information to the information processing device 1. For example, the terminal device 2 acquires information indicating the type of analysis device 3 from the control unit 31 of the analysis device 3 to be controlled and transmits the acquired information to the information processing device 1. The information processing device 1 receives the type of analysis device 3 and the query transmitted from the terminal device 2 via the communication unit 15, thereby acquiring the type of analysis device 3 and the query.

[0130] Next, the information processing device 1 selects command data according to the type of the analysis device 3 (S52). In S52, the calculation unit 11 selects command data corresponding to the acquired type of analysis device 3 from the plurality of command data included in the command DB 137. That is, the calculation unit 11 selects command data that records the specifications of commands for causing the analysis device 3 to be controlled to execute processing.

[0131] The information processing device 1 selects text data according to the type of analysis device 3 (S53). In S53, the calculation unit 11 selects text data corresponding to the acquired type of analysis device 3 from the multiple text data included in the text DB 138. In other words, the calculation unit 11 selects text data that records text related to the analysis device 3 to be controlled.

[0132] The information processing device 1 then executes the processes of S102 to S114 similar to those of the first embodiment. In S102, the information processing device 1 searches for related information related to the acquired query from the sentence data selected in S53, and acquires the searched related information. In S103, the information processing device 1 creates a prompt including the query, related information, and command data selected in S52. In the process of S104, the information processing device 1 acquires a response corresponding to the query, related information, and command data from the language generation model 132. In this way, the information processing device 1 controls each analysis device 3.

[0133] The information processing device 1 can individually control the multiple analysis devices 3. Even if the multiple analysis devices 3 are of different types, the information processing device 1 outputs a response corresponding to each analysis device 3 in response to the same query, and controls each analysis device 3. For example, when a query including a request to automate an analysis procedure is received, the information processing device 1 outputs a response corresponding to each analysis device 3 in response to the same query, and can automate the analysis procedure in each analysis device 3. In the second embodiment, as in the first embodiment, the information processing device 1 can also execute processing corresponding to the processing results of each analysis device 3, images such as captured images of the sample 6, or measurement data from sensors provided in each analysis device 3.

[0134] In the second embodiment, the information processing system 100 uses the language generation model 132 to perform processing for creating a report regarding multiple analysis results obtained by multiple analysis devices 3. The information processing system 100 executes the same processing of S31 to S36 as in the first embodiment. The information processing device 1 stores in the storage unit 13 a report format 135 that defines the format of a report that integrates multiple analysis results obtained by multiple analysis devices 3. In S31, the information processing device 1 acquires a query that includes a request for creating a report that integrates multiple analysis results. In S32, the information processing device 1 acquires multiple analysis results from the multiple analysis devices 3. In S33, the information processing device 1 acquires the report format 135 that defines the format of a report that integrates multiple analysis results. In S34, the information processing device 1 acquires the report that integrates multiple analysis results in accordance with the report format 135.

[0135] The processing of S31 to S36 outputs a report that integrates the multiple analysis results from the multiple analysis devices 3. The contents of the analysis results are output in a state in which the multiple analysis results are integrated according to the report format 135. For example, analysis results from multiple types of analysis devices 3, such as analysis results from an X-ray analysis device and a Raman analysis device, are output together. The user 5 checks the output report. Since the multiple analysis results are output in an appropriately integrated state, the user 5 can easily understand the analysis results from the multiple analysis devices 3. By outputting the multiple analysis results from the multiple analysis devices 3 in a combined form, the user 5 can gain insights that cannot be obtained from a single analysis result.

[0136] The present invention is not limited to the contents of the above-described embodiment, and various modifications are possible within the scope of the claims. In other words, embodiments obtained by combining technical means modified appropriately within the scope of the claims are also included in the technical scope of the present invention.

[0137] The matters described in each embodiment can be combined with each other. Furthermore, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations, regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. A multiple claim (multi-multi claim) that references at least one other multiple claim may also be used.

[0138] REFERENCE SIGNS LIST 1 Information processing device 11 Calculation unit 13 Storage unit 131 Computer program 132 Language generation model 133 Command data 134 Text data 135 Report format 136 Training data 2 Terminal device 3 Analysis device 41 First communication network 42 Second communication network 5 User

Claims

1. An information processing method comprising: acquiring a query related to an analysis device; acquiring a response related to control of the analysis device in response to the query using a language generation model; and outputting the response.

2. The information processing method according to claim 1, further comprising: storing command data recording the specifications of a command for causing the analysis device to execute a process; using the language generation model to obtain the response including a command for causing the analysis device to execute a specific process in response to the query and the command data; and inputting the command included in the response to the analysis device.

3. The information processing method according to claim 2, further comprising: determining whether a command included in the response is a command that can be executed by the analysis device; and, if the command included in the response is a command that cannot be executed by the analysis device, using the language generation model to obtain a new response that includes a new command in accordance with the query, the command data, and the command that cannot be executed by the analysis device.

4. The information processing method according to claim 3, further comprising the step of: correcting the command data depending on whether a command included in the response is executable by the analytical device.

5. The information processing method according to any one of claims 2 to 4, further comprising the steps of: acquiring a processing result of said analysis device in accordance with a command input to said analysis device; and outputting said processing result.

6. An information processing method as claimed in any one of claims 1 to 4, characterized in that the method comprises: acquiring, as the query, a query including a request for creation of a report on the analysis results by the analysis device; acquiring the analysis results by the analysis device and a format of the report; using the language generation model, acquiring a report on the analysis results created in accordance with the format in response to the query, the analysis results and the format; and outputting the report.

7. The information processing method according to claim 6, further comprising the steps of: acquiring a plurality of analysis results obtained by a plurality of analysis devices; acquiring as the query a query including a request for the creation of a report integrating the plurality of analysis results; and using the language generation model to acquire a report integrating the plurality of analysis results created in accordance with the format in accordance with the query, the acquired plurality of analysis results, and the format.

8. The information processing method according to any one of claims 1 to 7, further comprising: acquiring a type of the analysis device; and using the language generation model to acquire the response according to the query and the type of the analysis device.

9. An information processing method according to any one of claims 1 to 8, characterized in that the response includes a notification to a user regarding control of the analysis device in response to the query, using the language generation model, and outputting the notification.

10. An information processing method according to any one of claims 1 to 9, characterized in that: measurement data consisting of measurements taken by a sensor provided in the analysis device is acquired; and the response is acquired according to the query and the measurement data using the language generation model.

11. The information processing method according to any one of claims 1 to 10, further comprising: acquiring an image; and using the language generation model to acquire the response corresponding to the query and the image.

12. The information processing method according to claim 11, further comprising obtaining, as the image, a photographed image of a sample to be analyzed by the analytical device, and using the language generation model, obtaining the response including information for causing the analytical device to analyze the sample in response to the query and the photographed image.

13. The information processing method according to any one of claims 1 to 12, further comprising: acquiring a processing result from the analysis device; and acquiring the response according to the query and the processing result using the language generation model.

14. An information processing method according to any one of claims 1 to 13, further comprising obtaining an evaluation from a user of the output response, and training the language generation model based on training data including the query, the response, and the evaluation.

15. An information processing method as claimed in any one of claims 1 to 14, characterized in that: text data relating to the analysis device is stored; related information relating to the query is obtained from the text data; and the response is obtained according to the query and the extracted related information using the language generation model.

16. An information processing device comprising a calculation unit, which obtains a query related to an analysis device, uses a language generation model to obtain a response related to control of the analysis device in response to the query, and outputs the response.

17. A computer program that causes a computer to execute the following processes: acquire a query related to an analysis device; acquire a response related to control of the analysis device in response to the query using a language generation model; and output the response.

Images