Information processing methods, information processing devices and computer programs

Abstract

This invention provides an information processing method, an information processing apparatus, and a computer program for outputting an information processing response related to an analysis device and corresponding to a user's expectations. In the method processing apparatus, a query related to the analysis device is obtained, a language generation model is used to obtain a response related to the control of the analysis device corresponding to the query, and the response is output.

Description

Technical Field

[0001] This invention relates to information processing methods, information processing devices, and computer programs for sample analysis. Background Technology

[0002] There are various analytical apparatuses used for analyzing samples. For example, analytical apparatuses may include X-ray analysis apparatuses that analyze X-rays generated from a sample to determine the components contained in the sample, or Raman analysis apparatuses that analyze Raman light generated from a sample. An example of an X-ray analysis apparatus is disclosed in Patent Document 1. When an analytical apparatus performs certain analyses, it is necessary to properly control the analytical apparatus. Existing technical documents

[0003] Patent Document 1: Japanese Patent Publication No. 2023-74559

[0004] In situations where the operation method of the analysis device is unclear but analysis is desired, it can be difficult to control the analysis device according to the user's expectations. Therefore, a technology is needed that can obtain information related to the control of the analysis device based on the user's expectations. Summary of the Invention

[0005] The object of the present invention is to provide an information processing method, an information processing apparatus, and a computer program for obtaining information related to the control of an analysis device according to the user's expectations.

[0006] One aspect of the information processing method of the present invention is characterized by obtaining a query related to an analysis device, using a language generation model to obtain a response related to the control of the analysis device corresponding to the query, and outputting the response.

[0007] In one aspect of the invention, a query containing user expectations related to the analysis device is obtained; a language generation model is used to obtain a response related to the control of the analysis device corresponding to the query; and the response is output. Thus, a response corresponding to the user's expectations is output. For example, based on a query containing content about processes that the analysis device is desired to perform, a response is output to cause the analysis device to perform the user's desired processes.

[0008] One aspect of the information processing method of the present invention is characterized by storing command data that records the specifications of commands for causing the analysis device to perform processing, using the language generation model, obtaining the response containing the command for causing the analysis device to perform a specific processing based on the query and the command data, and inputting the command contained in the response into the analysis device.

[0009] In one embodiment of the invention, a response containing the command is obtained based on command data and a query that records specifications for commands used to cause the analysis device to perform processing, and the command is input to the analysis device. The analysis device then performs the processing desired by the user by executing the processing corresponding to the command.

[0010] One aspect of the information processing method of the present invention is characterized in that, if the command contained in the response is an executable command in the analysis device, and if the command contained in the response is an unexecutable command in the analysis device, the language generation model is used to obtain a new response containing the new command based on the query, the command data, and the unexecutable command in the analysis device.

[0011] In one aspect of the invention, it is determined whether the command contained in the response can be executed in the analysis device. If the command can be executed in the analysis device, processing according to the command is performed in the analysis device. If the command is not executable in the analysis device, a language generation model is used to output a new response containing the new command. Thus, the analysis device can reliably execute processing according to the command.

[0012] One aspect of the information processing method of the present invention is characterized in that the command data is modified according to whether the command contained in the response is an executable command in the analysis device.

[0013] In one aspect of the invention, command data is modified based on whether the command contained in the response can be executed in the analysis device. The command data reflects whether the command can be executed in the analysis device. By using the modified command data, commands for controlling the analysis device are efficiently obtained.

[0014] One aspect of the information processing method of the present invention is characterized by obtaining the processing result in the analysis device according to the command input to the analysis device, and outputting the processing result.

[0015] In one aspect of the invention, the processing result from the analysis device according to the command is output. The user can then confirm whether the desired processing has been performed.

[0016] One aspect of the information processing method of the present invention is characterized in that, as the query, a query is obtained containing a request to create a report related to the analysis results of the analysis device; the analysis results of the analysis device and the format of the report are obtained; the language generation model is used to obtain a report related to the analysis results created according to the format based on the query, the analysis results and the format; and the report is output.

[0017] In one aspect of the invention, a language generation model is used to obtain a report based on a query containing a request to create a report related to the analysis results of the analysis device and the report format. The analysis results are output in the form of a report with the content organized according to the format, allowing the user to confirm the analysis results of the analysis device.

[0018] An information processing method according to one aspect of the present invention is characterized by obtaining multiple analysis results from multiple analysis devices as the query, obtaining a query containing a request to create a report that integrates the multiple analysis results, and using the language generation model to obtain a report that integrates the multiple analysis results according to the query, the obtained multiple analysis results, and the format.

[0019] In one aspect of the invention, multiple analysis results from multiple analysis devices are obtained. Using a language generation model, a report summarizing the multiple analysis results is generated based on a query including a report creation request and the report format. The content of the multiple analysis results is output in the form of a formatted report. Users can easily understand and verify the multiple analysis results from the multiple analysis devices.

[0020] One aspect of the information processing method of the present invention is characterized by obtaining the type of the analysis device, using the language generation model, and obtaining the response corresponding to the query and the type of the analysis device.

[0021] In one aspect of the invention, a language generation model is used to obtain a response corresponding to the type of analysis device. Even if the type of analysis device differs, the user-expected processing can be performed within the analysis device.

[0022] One aspect of the information processing method of the present invention is characterized by using the language generation model to obtain the response containing a notification to the user related to the control of the analysis device corresponding to the query, and outputting the notification.

[0023] In one aspect of the invention, a language generation model is used to obtain a response containing notifications to the user related to the control of the analysis device, and a notification is output. The user operates the analysis device according to the output notification, and the analysis device can perform the processing desired by the user by utilizing the notification.

[0024] One aspect of the information processing method of the present invention is characterized by acquiring measurement data consisting of measurement values ​​measured by sensors provided by the analysis device, using the language generation model, and obtaining the response based on the query and the measurement data.

[0025] In one aspect of the invention, a language generation model is used to obtain a response related to the control of the analysis device based on measurement data from sensors provided by the analysis device. This enables processing corresponding to the measurement results obtained by the sensors to be performed within the analysis device.

[0026] One aspect of the information processing method of the present invention is characterized by acquiring an image, using the language generation model, and obtaining the response corresponding to the query and the image.

[0027] In one aspect of the invention, a language generation model is used to obtain a response related to the control of the analysis device based on an image. Thus, the analysis device can be controlled according to the user's expectations expressed not only using natural language but also using images.

[0028] One aspect of the information processing method of the present invention is characterized in that, as the image, an image of a sample to be analyzed in the analytical apparatus is acquired, and the language generation model is used to obtain, based on the query and the image, a response containing information for instructing the analytical apparatus to analyze the sample.

[0029] In one aspect of the invention, a language generation model is used to obtain a response related to the control of the analytical apparatus based on captured images of the sample. By using the captured images to designate specific points on the sample as measurement points, the analytical apparatus can be controlled according to the user's expectations based on the captured images.

[0030] One aspect of the information processing method of the present invention is characterized by obtaining the processing result in the analysis device, using the language generation model, and obtaining the response based on the query and the processing result.

[0031] In one aspect of the invention, the processing result in the analysis device is obtained, and a language generation model is used to obtain a response related to the control of the analysis device based on the processing result. Thus, the analysis device can be made to perform an appropriate next process based on the processing result in the analysis device.

[0032] One aspect of the information processing method of the present invention is characterized by obtaining user evaluations of the output response and enabling the language generation model to learn based on training data including the query, the response, and the evaluations.

[0033] In one aspect of the invention, user evaluations of responses obtained using a language generation model are acquired, and these evaluations are used to enable the language generation model to learn. By using the learned language generation model, more appropriate responses are obtained.

[0034] One aspect of the information processing method of the present invention is characterized by pre-storing text data related to the analysis device, obtaining association information associated with the query from the text data, and using the language generation model to obtain the response based on the query and the extracted association information.

[0035] In one aspect of the present invention, relevant information associated with a query is obtained from text data of an analysis device such as a guide, and a language generation model is used to obtain a response corresponding to the query and the relevant information. The response obtained using the language generation model becomes a response generated based on the relevant information, and therefore becomes a response associated with the analysis device. For example, the notification to the user included in the response may be a notification related to the analysis device, such as instructions on how to operate the analysis device.

[0036] An information processing apparatus according to one aspect of the present invention is characterized by comprising a computation unit that acquires a query related to an analysis device, the computation unit uses a language generation model to acquire a response related to the control of the analysis device corresponding to the query, and the computation unit outputs the response.

[0037] One aspect of the computer program of the present invention is characterized in that it causes the computer to perform the following processes: obtaining a query related to the analysis device; using a language generation model, obtaining a response related to the control of the analysis device corresponding to the query; and outputting the response.

[0038] In one aspect of the invention, an information processing device obtains a query containing the user's expectations related to an analysis device, uses a language generation model to obtain a response corresponding to the query and related to the control of the analysis device, and outputs the response. The information processing device executes processing according to a computer program. Based on the query containing content about the processing that the analysis device is desired to perform, a response is output to cause the analysis device to perform the processing expected by the user. Using the output response, the analysis device can perform the processing expected by the user.

[0039] According to the present invention, excellent effects can be obtained, such as the ability to obtain information related to the control of the analysis device that corresponds to the user's expectations. Attached Figure Description

[0040] Figure 1 This is a schematic diagram illustrating an example of the configuration of an information processing system for Implementation 1 of the control and analysis device. Figure 2 This is a block diagram illustrating an example of the functional configuration of an analytical apparatus. Figure 3 This is a block diagram illustrating an example of the internal structure of a terminal device. Figure 4 This is a block diagram illustrating an example of the internal structure of the information processing apparatus according to Embodiment 1. Figure 5 This is a diagram representing an example of the content of command data. Figure 6 This is a flowchart illustrating a first example of the steps of the processing performed by the information processing device of Embodiment 1 in order to control the analysis device. Figure 7 This is a flowchart illustrating a second example of the steps involved in the processing performed by the information processing device in order to control the analysis device. Figure 8 This is a flowchart illustrating a third example of the steps involved in the processing performed by the information processing device in order to control the analysis device. Figure 9 This is a flowchart illustrating the fourth example of the steps involved in the processing performed by the information processing device in order to control the analysis device. Figure 10 This is a concept map representing an example of the analysis results. Figure 11 This is a conceptual diagram representing an example of report format content. Figure 12 This is a flowchart illustrating an example of the processing steps performed by the information processing system 100 to generate a report related to the analysis results in the analysis device. Figure 13 This is a diagram illustrating an example of a report related to the analysis results. Figure 14 This is a flowchart illustrating an example of the steps involved in the learning process performed by an information processing device. Figure 15 This is a schematic diagram illustrating an example of the configuration of the information processing system in Implementation Method 2. Figure 16 This is a block diagram illustrating an example of the internal structure of the information processing device in Embodiment 2. Figure 17 This is a flowchart illustrating an example of the steps of processing performed by the information processing device of Embodiment 2 in order to control the analysis device. Detailed Implementation

[0041] The present invention will now be described in detail with reference to the accompanying drawings illustrating this embodiment. <Implementation Method 1> In this embodiment, a language generation model is used to control an analysis device that corresponds to the user's expectations. Figure 1This is a schematic diagram illustrating an example configuration of an information processing system 100 used to control an analytical apparatus according to Embodiment 1. The information processing system 100 includes a terminal device 2 used by a user 5, an information processing device 1, and an analytical apparatus 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 LAN (Local Area Network) or the Internet. Furthermore, the terminal device 2 communicates with the analytical apparatus 3 via a second communication network 42. The analytical apparatus 3 is a device that measures phenomena occurring in a sample and analyzes the measurement results. The analytical apparatus 3 is, for example, an X-ray analysis device.

[0042] Figure 2 This is a block diagram illustrating an example of the functional configuration of the analysis device 3. Figure 1 The diagram illustrates an example of an X-ray analysis apparatus, denoted as Analytical Apparatus 3. Analytical Apparatus 3 includes: a sample stage 36 for holding a sample 6, an irradiation unit 35 for irradiating the sample 6 with X-rays, an imaging unit 37 for imaging 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 imaging elements. When radiation is irradiated from the irradiation unit 35 onto the sample 6, fluorescent X-rays are generated on the sample 6. These fluorescent X-rays are incident on the radiation detector 34, which detects the incident fluorescent X-rays. In the diagram, 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 a semiconductor-based element. In the event of incident fluorescent X-rays, the radiation detection element 341 outputs a signal corresponding to the energy of the incident fluorescent X-rays. The radiation detector 34 detects fluorescent X-rays by the incident fluorescent X-rays from the sample 6 and by outputting a signal from the radiation detection element 341 based on the incident fluorescent X-rays. The temperature sensor 342 measures the internal temperature of the radiation detector 34. The temperature sensor 342 is constructed, for example, using a thermistor or a thermocouple.

[0044] The analytical apparatus 3 includes a sample chamber 38. The sample chamber 38 is box-shaped and has an openable and closable lid 381. Closing the lid 381 seals the sample chamber 38. A sample stage 36, an irradiation unit 35, an imaging unit 37, and a radiation detector 34 are disposed inside the sample chamber 38. A sample 6 is placed inside the sample chamber 38 by placing it on the sample stage 36. With the lid 381 open, the user 5 places or removes the sample 6 from the sample stage 36. With the lid 381 closed, the pressure inside the sample chamber 38 is reduced, and X-ray irradiation of the sample 6 and detection of fluorescence X-rays are performed. The irradiation unit 35 can also irradiate the sample 6 disposed inside the sample chamber 38 from the outside of the sample chamber 38. The imaging unit 37 can also image the sample 6 disposed inside the sample chamber 38 from the outside of the sample chamber 38.

[0045] The signal processing unit 33 is connected to the radiation detector 34. The signal processing unit 33 receives the signal output from the radiation detector 34 and detects the signal intensity, thereby detecting the signal value corresponding to the energy of the radiation detected by the radiation detector 34. The signal processing unit 33 counts the signals according to their values ​​and outputs data representing the relationship between the signal values ​​and the count count.

[0046] The analysis unit 32 is connected to the signal processing unit 33. The analysis unit 32 is configured using a computer. The analysis unit 32 receives data from the signal processing unit 33 representing the relationship between signal values ​​and counts. Based on the data from the signal processing unit 33, the analysis unit 32 generates a spectrum of fluorescent X-rays incident on the radiation detector 34. The spectrum represents the relationship between the energy and intensity of the fluorescent X-rays. The processing of counting the signal output from the radiation detector 34 according to signal values ​​can also be performed by the analysis unit 32, not the signal processing unit 33. The generation of the fluorescent X-ray spectrum can also be performed by the signal processing unit 33. The analysis unit 32 stores spectral data representing the fluorescent X-ray spectrum. Thus, the analysis device 3 measures fluorescent X-rays. The analysis unit 32 can also perform spectral information processing based on radiation. For example, the analysis unit 32 performs qualitative or quantitative analysis of the elements contained in the sample 6. The analysis unit 32 stores the analysis results.

[0047] A drive unit 361 is connected to the sample stage 36. The drive unit 361 is, for example, constructed using a stepper motor. The sample stage 36 is moved by driving the sample stage 36 through the drive unit 361. For example, the sample stage 36 moves in the horizontal direction. By moving the sample stage 36, 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 operation 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. Furthermore, the control unit 31 communicates with the terminal device 2 via a second communication network 42. The control unit 31 and the analysis unit 32 may also be integrated. Figure 2 The structure of the analytical device 3 shown is one example; the structure of the analytical device 3 can also be other structures.

[0049] Figure 3 This is a block diagram illustrating an example of the internal configuration of terminal device 2. Terminal device 2 is a computer used by user 5, such as a smartphone, tablet computer, or personal computer. Terminal device 2 includes an arithmetic 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 arithmetic unit 21 is a processor, such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a multi-core CPU. The arithmetic unit 21 may also be constructed using a quantum computer. The memory 22 stores temporary data generated during computation. The memory 22 is, for example, RAM (Random Access Memory). The reading unit 24 reads information from a recording medium 20, such as an optical disc or removable memory. The storage unit 23 is non-volatile, such as a hard disk or non-volatile semiconductor memory.

[0050] The operation unit 25 accepts input of information such as text by receiving operations from the user 5. The operation unit 25 may be, for example, a touch panel, keyboard, or pointing device. The display unit 26 displays images. The display unit 26 may be, for example, a liquid crystal display (LCD) or an electroluminescent display (EL display). The operation unit 25 and the display unit 26 may also be integrated. The first communication unit 27 communicates with the information processing device 1 via a first communication network 41 via 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 via wired or wireless communication.

[0051] The arithmetic unit 21 causes the reading unit 24 to read the computer program (program product) 231 recorded on the recording medium 20, and stores the read computer program 231 in the storage unit 23. The arithmetic unit 21 executes processing to implement the functions of the terminal device 2 according to the computer program 231. The computer program 231 may also be pre-stored in the storage unit 23 or downloaded from outside the terminal device 2. In this case, the terminal device 2 may not need to have a reading unit 24.

[0052] Computer program 231 can be executed on a single computer, or configured on one site or distributed across multiple sites and executed on multiple computers interconnected via a communication network. That is, terminal device 2 can also be composed of multiple computers, and computer program 231 can also be executed on multiple computers connected via a communication network.

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

[0054] The arithmetic unit 11 causes the reading unit 14 to read the computer program (program product) 131 recorded on the recording medium 10, and stores the read computer program 131 in the storage unit 13. The arithmetic unit 11 executes processing to implement the functions of the information processing device 1 according to the computer program 131. The computer program 131 may also be pre-stored in the storage unit 13 or downloaded from outside the information processing device 1. In this case, the information processing device 1 may not need to have a reading unit 14.

[0055] Computer program 131 can be executed on a single computer, or configured on one site or distributed across multiple sites and executed on multiple computers interconnected via a communication network. That is, information processing apparatus 1 can also be composed of multiple computers, and computer program 131 can also be executed on multiple computers connected via a communication network. Information processing apparatus 1 can also be configured using a cloud server.

[0056] The processing of each step in the information processing method described later can be performed by multiple computers. The processing of each step can also be performed by different computers. The processing of each step can also be performed using a virtual machine. The processing of each step can also be performed by multiple processing units. The processing of each step can also be performed by different processing units.

[0057] Information processing device 1 includes a language generation model 132. The language generation model 132 is a learned model that outputs text corresponding to the input prompt when a prompt consisting of an instruction requesting a certain response is input. Alternatively, the language generation model 132 is a learned model that outputs text corresponding to both the input prompt and the input information when a prompt and accompanying information 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, or LLaMA, etc. The language generation model 132 has been pre-learned.

[0058] The language generation model 132 is implemented by the arithmetic unit 11 executing information processing according to the computer program 131. The storage unit 13 stores data used to implement the language generation model 132. Alternatively, the language generation model 132 can also be constructed in hardware. The language generation model 132 can also be implemented using a quantum computer. Alternatively, the language generation model 132 can be disposed outside the information processing device 1, and the information processing device 1 can perform processing using the external language generation model 132. The language generation model 132 can be implemented using multiple computers connected via a communication network, or it can be implemented using the cloud.

[0059] Storage unit 13 stores command data 133. Command data 133 records the specifications of commands used to cause analysis device 3 to perform processing. Figure 5 This is a diagram illustrating an example of the contents of command data 133. Command data 133 describes, in text, the specifications of commands used to instruct the analysis device 3 to perform specific processes. Command data 133 is determined based on the specifications of the analysis device 3. When a command is input, the analysis device 3 executes the process corresponding to that command. Information related to multiple commands is recorded in command data 133. Figure 5 The text illustrates an example where command MECR is recorded in command data 133 for adding or removing measurement points from the analytical apparatus 3, and command STSR is recorded for obtaining the status of the analytical apparatus 3. The measurement point is a point on the sample 6 that has been irradiated with X-rays; the fluorescence X-rays generated from the measurement point are measured. Command data 133 also records information related to other commands.

[0060] Storage unit 13 stores text data 134. Text data 134 records text related to the analysis apparatus 3. For example, text data 134 includes text describing the operation method of the analysis apparatus 3, instructions on the information output from the analysis apparatus 3, methods for processing the sample 6 used in the analysis apparatus 3, or methods for maintaining the analysis apparatus 3. For example, text data 134 includes a guide to the analysis apparatus 3 or a troubleshooting manual related to the analysis apparatus 3.

[0061] The information processing system 100 uses the language generation model 132 to control the analysis device 3. Figure 6 This is a flowchart illustrating a first example of the processing steps performed by the information processing device 1 of Embodiment 1 in order to control the analysis device 3. Hereinafter, the steps will be abbreviated as S. The arithmetic unit 11 performs information processing according to computer program 131, thereby the information processing device 1 performs the following processing. Furthermore, the arithmetic unit 21 performs information processing according to computer program 231, thereby the terminal device 2 performs processing.

[0062] Information processing device 1 obtains a query related to analysis device 3 (S101). In S101, the user 5 inputs a query to terminal device 2 by operating operation unit 25. Computation unit 21 causes first communication unit 27 to send the input query to information processing device 1 via first communication network 41. Information processing device 1 obtains the query by receiving the query sent from terminal device 2 via communication unit 15.

[0063] The query contains the content that user 5 wants the analysis device 3 to perform. For example, the query is data composed of natural language. The following are examples of queries used to perform specific processes in the analysis device 3. "Generate Python code for adding measurement points. Measure as quickly as possible. Automatic current control. Other settings are optional. Standby until the device becomes idle. If it becomes idle, call print("finished.") and terminate." The query requests the generation of code to enable the analysis device 3 to perform processing for additional measurement points.

[0064] The query can also include a request for notification to user 5. For example, if the method of operation for instructing the analysis device 3 to perform a process is unclear, a query can be entered requesting notification of the operation method. The following is an example of a query requesting notification to user 5 in natural language. What is the choice of spot size? What are the advantages of each spot size? Which one should be chosen? The query requests information including a method for selecting the size of the X-ray spot used to irradiate sample 6. Thus, a query containing the user 5's expectations of the analysis apparatus 3 is obtained.

[0065] Information processing device 1 obtains association information related to the obtained query from text data 134 (S102). In S102, the calculation unit 11 retrieves association information related to the obtained query from text data 134, thereby obtaining the retrieved association information. For example, if text data 134 is divided into multiple parts by page, paragraph, or item, the calculation unit 11 extracts keywords contained in the query, retrieves the parts containing the keywords from text data 134, and obtains the parts containing the keywords as association information. For example, if text data 134 is vectorized, the calculation unit 11 calculates the similarity, such as the cosine similarity, between the vectorized keywords and each part of the vectorized text data 134, and obtains the part with the highest similarity as association information.

[0066] Next, the information processing device 1 generates a prompt containing the obtained query, the obtained related information, and command data 133 (S103). In S103, the arithmetic unit 11 reads the command data 133 from the storage unit 13 and generates a prompt containing the query, related information, and command data 133. The arithmetic unit 11 generates a prompt requesting reference to the related information and command data 133 to generate a response, which includes a command for the analysis device 3 to perform processing corresponding to the query and a notification to the user 5. For example, the storage unit 13 stores a fixed statement contained in the prompt, and the arithmetic unit 11 generates a prompt containing the fixed statement, query, related information, and command data 133. For example, the fixed statement contains information specifying the function of the language generation model 132 and the output form of the command.

[0067] The following are examples of fixed statements included in the prompt. "You are a generator that produces Python scripts based on what the user expresses in natural language." The script needs to be generated as simply as possible. It shouldn't generate Python methods, but rather simply generate executable code based on existing code. No explanation is required for the generated code. Please only answer with the code. No explanation of the independent or variable is required. Please only answer with the code. Commands begin with the command name in ASCII form, followed by the entire set of parameters, with words separated by commas ",". For example, the complete command for "ABC" is "ABC,1,2,3,4". Furthermore, the response command (where defined) is "ABC,2,3,4,5,...". Since this is a simple string, it needs to be split to retrieve the parameters, with the command type always at the beginning. Other parameters follow the command type. In Python, as you answered, it requires a command-line style. The method "SendCommand(string_)" sends the generated command. Only one instance of this method is needed for a complete command text. The method "SendAndGetCommand(string_)" sends the generated command and, if a response value exists, retrieves the command as a string. In this fixed statement, the code that causes the analysis device 3 to perform certain processes is instructed to be generated, and the format of the generated code is specified.

[0068] Information processing device 1 obtains a response related to the control of analysis device 3 corresponding to the query, command data 133, and associated information from language generation model 132 (S104). In S104, arithmetic unit 11 inputs the generated prompt to language generation model 132. Language generation model 132 performs calculations based on the prompt input and outputs a response. Since the prompt contains a query related to the control of analysis device 3, language generation model 132 outputs a response related to the control of analysis device 3. Arithmetic unit 11 obtains the response output by language generation model 132. Arithmetic unit 11 stores the obtained response in memory 12 or storage unit 13.

[0069] Alternatively, the information processing device 1 can also perform the following processing: in S103, a prompt character without command data 133 and associated information is created; in S104, the prompt character with command data 133 and associated information is input into the language generation model 132. In S103, the information processing device 1 can also create a prompt character that includes information such as the storage address of command data 133 and associated information for referencing command data 133 and associated information, but does not contain command data 133 and associated information. Alternatively, in S104, the information processing device 1 can use command data 133 and associated information to enable the language generation model 132 to learn; after learning, a prompt character without command data 133 and associated information is input into the language generation model 132 and a response is obtained.

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

[0071] The notification to user 5 is included in the response generated based on the associated information, and therefore becomes a notification associated with the analytical apparatus 3. The notification to user 5 is an instruction to user 5 to perform operations on the analytical apparatus 3 for controlling the apparatus, such as "Please turn on the analytical apparatus" or "Please open the cover, place the sample on the sample stage, and close the cover." The notification to user 5 may also be a request for input of further information required to control the analytical apparatus 3.

[0072] The notification to user 5 is the notification corresponding to the query. If the query includes a request for a notification to user 5, the requested notification is included in the response. For example, based on a query requesting a notification explaining the operation method of the analysis device 3, a notification explaining the operation method of the analysis device 3 is output. The following is an example of a notification corresponding to a query requesting a notification including a method for selecting the spot size. "There are two options for the selection of spot size: 1. 100μm, 2. 7mm. The advantages of each spot size are described below." 1. With a spot size of 100 μm, the spatial resolution is improved, and the detection limit for small particles and features on the sample surface is increased. However, the measurement time is longer, and the absorption effect of X-rays may become more pronounced. 2. With a 7mm spot size, X-ray intensity is high, and measurement time is shortened. This is suitable for large samples and specimens with high X-ray absorption. However, spatial resolution and detection limit are sacrificed. The selection of the spot size needs to be based on a trade-off between the sample size, sample characteristics, spatial resolution, and measurement speed. In this example, a notification was received explaining in natural language how to select the spot size.

[0073] User 5, having confirmed the output notification, operates the analysis device 3 accordingly. By outputting a notification to user 5, control of the analysis device 3 becomes possible, even if it cannot be performed simply by inputting commands to the analysis device 3. If the response does not include a notification to user 5, the information processing device 1 omits the processing in step S105.

[0074] Information processing device 1 determines whether the received response contains a command (S106). In S106, the arithmetic unit 11 determines whether the response contains a command based on command data 133. Here, a command refers to a command that causes the analysis device 3 to perform certain processes. The command can be a single command or a command statement containing multiple commands. The command can also be a program or a script. Binary code can also be included in the command. If the response does not contain a command (S106: No), information processing device 1 proceeds the processing to S114, which will be described later.

[0075] If the response contains a command (S106: Yes), the information processing device 1 determines whether the command contained in the response can be executed in the analysis device 3 (S107). In S107, the arithmetic unit 11 extracts the command contained in the response and causes the communication unit 15 to send the command to the terminal device 2. The terminal device 2 receives the command via the first communication unit 27. The arithmetic unit 21 causes the second communication unit 28 to send 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 sent from the terminal device 2 and determines whether the processing according to the received command can be executed. For example, the control unit 31 verifies whether the command can be executed by trial execution. The control unit 31 sends the determination result of whether the command can be executed to the terminal device 2, and the terminal device 2 sends the determination result to the information processing device 1. The information processing device 1 receives the determination result via the communication unit 15. Alternatively, the information processing device 1 may also perform the processing of determining whether the command can be executed in the analysis device 3 itself.

[0076] If the command contained in the response cannot be executed in the analysis device 3 (S107: No), the information processing device 1 generates a prompt corresponding to the command that cannot be executed in the analysis device 3 (S108). In S108, the arithmetic unit 11 generates a prompt containing the query, related information, command data 133, and the command that cannot be executed. The arithmetic unit 11 generates a prompt requesting the generation of a response containing a command different from the command that cannot be executed in the analysis device 3 in order for the analysis device 3 to perform the processing corresponding to the query. For example, the arithmetic unit 11 specifies the command that cannot be executed and generates a prompt requesting the generation of a new command that does not contain the specified command. Alternatively, the arithmetic unit 11 obtains an error message generated by the control unit 31 attempting to execute the command that cannot be executed and generates a prompt requesting the generation of a new command that does not generate the obtained error message.

[0077] Next, the information processing device 1 returns the processing to S104. The arithmetic unit 11 inputs the prompt created in S108 into the language generation model 132, and the language generation model 132 outputs a response corresponding to the prompt. The arithmetic unit 11 obtains the response. The command contained in the response is different from the command that cannot be executed in the analysis device 3. In this way, the information processing device 1 obtains a new response containing the new command based on the query, the associated information, the command data 133, and the command that cannot be executed in the analysis device 3.

[0078] If the command contained in the response can be executed in the analysis device 3 (S107: Yes), the information processing device 1 corrects the command data 133 based on the determination result of whether the command can be executed in the analysis device 3 (S109). In S109, for example, the arithmetic unit 11 adds information indicating that the commands recorded in the command data 133 can be executed to commands determined to be executable in the analysis device 3. For example, the arithmetic unit 11 adds information indicating that the commands cannot be executed in the analysis device 3 to commands recorded in the command data 133. For example, the arithmetic unit 11 deletes commands determined to be unexecutable in the analysis device 3 from the command data 133.

[0079] After the command data 133 is corrected, the information processing device 1 uses the language generation model 132 to obtain a response containing the command corresponding to the corrected command data 133. The corrected command data 133 reflects whether the command can be executed in the analysis device 3; therefore, commands that can be executed in the analysis device 3 are easily included in the response. Thus, commands for controlling the analysis device 3 can be obtained effectively. Alternatively, if a command included in the response is determined to be unexecutable in the analysis device 3, the processing in step S109 can be executed. The processing in step S109 can also be omitted.

[0080] Next, the information processing unit 1 inputs the command contained in the response to the analysis unit 3 (S110). In S110, the arithmetic unit 11 causes the communication unit 15 to send the command contained in the response to the terminal device 2. The terminal device 2 receives the command and sends it to the analysis unit 3. In the analysis unit 3, the control unit 31 receives the command sent from the terminal device 2. In this way, the information processing unit 1 outputs the response obtained from the language generation model 132 by inputting the command contained in the response to the analysis unit 3. Alternatively, in S110, the command may not be input to the analysis unit 3 again, and the analysis unit 3 may perform the processing according to the command received in S107.

[0081] The control unit 31 controls each part of the analytical apparatus 3 according to commands. Under the control of the control unit 31, each part of the analytical apparatus 3 operates, thereby the analytical apparatus 3 performs processing according to commands. For example, the analytical apparatus 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 the fluorescent X-ray spectrum as a processing result. For example, the analytical apparatus 3 continuously or intermittently measures physical quantities such as the intensity of fluorescent X-rays, and the processing result is time-series data representing the results of continuous or intermittent measurements of the physical quantities.

[0082] The commands contained in the response output by language generation model 132 are the commands corresponding to the query. These commands are used to execute the processes requested in the query to be performed by analysis device 3. The following are examples of commands corresponding to queries requesting analysis device 3 to perform processes to add measurement points. #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 received to cause the analysis device 3 to perform the process of adding measurement points.

[0083] Information processing device 1 obtains the processing result from analysis device 3 (S111). Control unit 31 sends the processing result from analysis device 3 to terminal device 2. Terminal device 2 receives the processing result sent from analysis device 3 and sends it to information processing device 1. Information processing device 1 receives the processing result through communication unit 15. Calculation unit 11 stores the received processing result in storage unit 13. In this way, information processing device 1 obtains the processing result from analysis device 3.

[0084] Information processing device 1 outputs the processing result from analysis device 3 (S112). In S112, the arithmetic unit 11 causes the communication unit 15 to send the processing result to terminal device 2. Terminal device 2 receives the processing result, and the arithmetic unit 21 outputs the processing result by displaying the received processing result on display unit 26. At this time, the arithmetic unit 21 generates an image containing the processing result and displays the generated image on display unit 26. For example, as a processing result, a fluorescence X-ray spectrum is displayed on display unit 26. For example, as a processing result, a graph representing time series data is displayed. User 5 confirms the output processing result from analysis device 3. By confirming the processing result, user 5 can know whether the processing that user 5 wanted analysis device 3 to perform was actually executed.

[0085] Information processing device 1 obtains an evaluation from user 5 (S113). In S113, terminal device 2 obtains an evaluation as a result of processing the information processing device 1 to output a response from language generation model 132, by having user 5 operate operation unit 25. For example, calculation unit 11 causes communication unit 15 to send data of an input image for inputting the evaluation to terminal device 2, terminal device 2 receives the input image data, and calculation unit 21 displays the input image based on the received data on display unit 26. User 5 inputs the evaluation to terminal device 2 using the input image, and terminal device 2 obtains the evaluation. Terminal device 2 sends the obtained evaluation to information processing device 1, and information processing device 1 receives the evaluation. Calculation unit 11 stores the received evaluation from user 5 in storage unit 13.

[0086] The evaluation from user 5 may be, for example, information indicating whether the processing of information processing device 1 is good or bad, in a binary choice. The evaluation from user 5 may also be information indicating whether the processing is good or bad, in multiple levels than two. Processing step S113 may also be omitted.

[0087] Information processing device 1 determines whether to terminate the processing for control analysis device 3 (S114). In S114, for example, by user 5 operating operation unit 25, user 5 inputs an end instruction to terminal device 2, terminal device 2 sends the end instruction to information processing device 1, and if information processing device 1 receives the end instruction, calculation unit 11 determines that the processing has ended. If no end instruction is received, calculation unit 11 determines that the processing should not end.

[0088] If it is determined that the processing for the control analysis device 3 should not be terminated (S114: No), the information processing device 1 returns the processing to S101. In S101, the user 5 inputs a query based on the processing result. The information processing device 1 obtains the query and repeats the processing after S102. If it is determined that the processing for the control analysis device 3 should be terminated (S114: Yes), the information processing device 1 terminates the processing for the control analysis device 3.

[0089] S101 to S114 include processing that utilizes the associated information obtained from the text data 134, but the information processing device 1 may also perform processing that does not utilize the associated information. That is, the processing in S102 may be omitted. S101 to S114 include processing that obtains a response including a notification to the user 5, but the information processing device 1 may also perform processing that obtains a response not including a notification to the user 5. That is, the processing in S105 may be omitted.

[0090] In steps S101 to S114, the information processing device 1 obtains a response from the language generation model 132 corresponding to the query, command data 133, and associated information. The information processing device 1 may also perform processing to obtain a response corresponding to the processing result in the analysis device 3. Figure 7 This is a flowchart illustrating a second example of the steps of processing performed by the information processing device 1 in order to control the analysis device 3.

[0091] Information processing device 1 acquires the processing result from analysis device 3 (S21). In S21, the arithmetic unit 11 acquires the processing result by reading the processing result from analysis device 3 stored in storage unit 13. At this time, information processing device 1 can also cause terminal device 2 to output the processing result. Information processing device 1 can also execute the processing in S21 by having user 5 operate operation unit 25 to input an instruction to acquire the processing result from analysis device 3 to terminal device 2, and then executing the processing in S21 according to the instruction received from terminal device 2. For example, the processing result from analysis device 3 is a fluorescence X-ray spectrum. For example, the processing result from analysis device 3 is time series data.

[0092] Next, the information processing device 1 executes the processing in S101. In S101, the information processing device 1 obtains a query corresponding to the processing result in the analysis device 3. For example, the user 5 inputs a query to the terminal device 2 to cause the analysis device 3 to perform processing that changes the conditions relative to the obtained processing result. For example, if the processing result is a spectrum of fluorescence X-rays generated from a specific point on the sample 6, a query is obtained instructing the user to obtain a spectrum of fluorescence X-rays generated from a point near that specific point. For example, if the processing result is time series data, a query is obtained instructing the user to perform analysis based on the time series data.

[0093] Next, the information processing device 1 executes processes S102 to S114. In S103, the information processing device 1 generates a prompt containing the query, related information, command data 133, and the obtained processing result. Through the processing in S104, the information processing device 1 obtains a response from the language generation model 132 corresponding to the processing result in the query and analysis device 3. In this way, the information processing device 1 enables the analysis device 3 to execute the appropriate next processing based on the processing result. The information processing device 1 may also perform the following processing: in S103, generate a prompt without processing result, and in S104, input a prompt with processing result into the language generation model 132.

[0094] The information processing device 1 can also perform processing to obtain a response corresponding to the image. Figure 8 This is a flowchart illustrating a third example of the processing steps performed by the information processing unit 1 to control the analysis device 3. The information processing unit 1 acquires an image (S22). In S22, the arithmetic unit 11 acquires an image by reading the image stored in the storage unit 13. At this time, the information processing unit 1 can also cause the terminal device 2 to display the image. Alternatively, the information processing unit 1 can execute the processing in S22 by having the user 5 operate the operation unit 25, inputting an instruction to acquire the image to the terminal device 2, and receiving the instruction from the terminal device 2. For example, the image may be an image representing the spectrum of fluorescence X-rays, or a mapping image representing the distribution of elements within the sample 6 obtained through fluorescence X-ray analysis.

[0095] Next, the information processing device 1 executes the processing in S101. In S101, the information processing device 1 obtains a query corresponding to the image. For example, the user 5 inputs a query with reference to the image into the terminal device 2. For example, a query is obtained asking why a specific peak is contained in the spectrum of fluorescence X-rays. For example, a query is obtained requesting the location of a substance containing a specific element contained in the mapped image within the sample 6.

[0096] Next, the information processing device 1 executes processes S102 to S114. In S103, the information processing device 1 generates a prompt containing the query, related information, command data 133, and the obtained image. Through the processing in S104, the information processing device 1 obtains a response corresponding to the query and the image from the language generation model 132. In this processing, the information processing device 1 uses a multimodal model such as GPT-4, which can utilize images, as the language generation model 132. In this way, the information processing device 1 enables the analysis device 3 to perform processing corresponding to the image. The information processing device 1 may also perform the following processing: in S103, a prompt without an image is generated, and in S104, a prompt with an image is input into the language generation model 132.

[0097] The image acquired by the information processing device 1 in S22 can also be a photographed image of the sample 6. The user 5 operates the operation unit 25, inputting an instruction to acquire the photographed image into the terminal device 2, which then sends the instruction to the analysis device 3. Under the control of the control unit 31 according to the photographed instruction, the imaging unit 37 photographs the sample 6, creating a photographed image. The control unit 31 sends the photographed image to the terminal device 2, which in turn sends it to the information processing device 1. The information processing device 1 acquires the photographed image by receiving it.

[0098] In S101, the information processing device 1 obtains a query corresponding to the captured image. For example, user 5 inputs a query referring to the captured image into terminal device 2. For example, by specifying any point on the sample 6 reflected in the captured image, a query is obtained requesting X-ray irradiation of the specified point for elemental analysis. For example, by specifying any part of the captured image, a query is obtained requesting magnification of the specified part. For example, a query is obtained requesting the location of any part contained in the captured image within the sample 6. Through the processing in S104, the information processing device 1 obtains a response from the language generation model 132 containing information for causing the analysis device 3 to perform processing for analyzing the sample 6 based on the query and the captured image. For example, a response is obtained containing a command for causing the analysis device 3 to perform processing by irradiating the specified point on the sample 6 with X-rays for elemental analysis, or a notification guiding the operations required for the analysis device 3 to perform such processing. In this way, the information processing device 1 enables the analysis device 3 to perform processing corresponding to the captured image of the sample 6.

[0099] As described above, the information processing device 1 obtains a query containing the user 5's expectations related to the analysis device 3, obtains a response related to the control of the analysis device 3 corresponding to the query from the language generation model 132, and outputs the obtained response. Thus, information related to the control of the analysis device 3 corresponding to the user's expectations is output. For example, based on a query containing content about a process that the analysis device 3 wants to perform, a response is output to cause the analysis device 3 to perform the process expected by the user 5. If the response contains a command to cause the analysis device 3 to perform a process, the analysis device 3 performs the process according to the command, thereby performing the process expected by the user 5. If the response contains a notification to the user 5, the user 5 can perform actions such as operating the analysis device 3 according to the output notification. Even if the operation method of the analysis device 3 is unclear, a response related to the control of the analysis device 3 corresponding to the user 5's expectations is output. By outputting a command to cause the analysis device 3 to perform the process expected by the user 5 or a notification to cause the analysis device 3 to perform a process, the analysis device 3 can perform the process expected by the user 5.

[0100] In cases where analysis of special samples, samples being analyzed for the first time, or unknown samples requires analysis using methods different from the usual, the operating method of the analytical apparatus 3 may sometimes be unclear. Alternatively, the operating method may differ depending on the type of analytical apparatus 3. For example, even among X-ray analytical apparatuses, the operating method may vary depending on the manufacturer or manufacturing period. Even if the user 5 is proficient in operating one type of analytical apparatus 3, there may be situations where the operating method of other types of analytical apparatus 3 is unknown. Even in such cases, by inputting a query containing the processing that the user 5 wants the analytical apparatus 3 to perform, a response can be obtained to cause the analytical apparatus 3 to perform the processing desired by the user 5. In this way, processing corresponding to the user 5's expectations can be performed within the analytical apparatus 3.

[0101] Even if the operation or control method of the analysis device 3 is clearly defined, sometimes a response related to the control of the analysis device 3 is required, corresponding to the expectations of the user 5. For example, there may be an expectation to automate the analysis process, which consists 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 corresponding to the output. Typically, it is difficult to combine multiple commands to create a program for automation. In this embodiment, the user 5 inputs a query containing the analysis steps and a request to automate the analysis steps. The information processing device 1 can output a response containing a combination of commands to automate the analysis steps based on the query. By causing the analysis device 3 to perform processing according to the output response, the analysis steps can be automated.

[0102] The information processing device 1 can also perform processing to obtain a response corresponding to the measurement data of the sensor. Figure 9 This is a flowchart illustrating a fourth example of the processing steps performed by the information processing unit 1 to control the analysis unit 3. The information processing unit 1 acquires measurement data generated using the temperature sensor 342 (S23). The user 5 operates the operation unit 25, inputting an instruction to acquire the measurement data into the terminal device 2, which then sends the instruction to acquire the measurement data to the analysis unit 3. The control unit 31, according to the instruction, generates measurement data consisting of the measured values ​​of the temperature of the radiation detector 34 measured using the temperature sensor 342. The control unit 31 may also store pre-generated measurement data. The control unit 31 sends the measurement data to the terminal device 2, which in turn sends the measurement data to the information processing unit 1. The information processing unit 1 acquires the measurement data by receiving it.

[0103] Next, the information processing device 1 executes the processing in S101. In S101, the information processing device 1 obtains a query corresponding to the measurement data. For example, the user 5 inputs a query requesting control of the temperature of the radiation detector 34 or a query requesting a notification suggesting operation corresponding to the temperature into the terminal device 2.

[0104] Next, the information processing device 1 executes processes S102 to S114. In S103, the information processing device 1 generates a prompt containing query, related information, command data 133, and measurement data. Through the processing in S104, the information processing device 1 obtains a response corresponding to the query and measurement data from the language generation model 132. In this way, the information processing device 1 enables the analysis device 3 to perform processing corresponding to the measured temperature of the radiation detector 34. The information processing device 1 may also perform the following processing: in S103, generate a prompt without measurement data, and in S104, input a prompt with measurement data into the language generation model 132.

[0105] The information processing system 100 can generate reports related to the analysis results in the analysis device 3 using the language generation model 132. The analysis device 3 stores analysis results in the control unit 31 or the analysis unit 32, which record the results of the analysis performed so far. The analysis results can be individual processing results or results from analysis performed by the analysis unit 32 based on multiple processing results.

[0106] Figure 10 This is a concept map representing an example of the analysis results. Figure 10 In the example shown, the analysis results include a title, fluorescence X-ray spectra, quantitative results, notes, experimental date, experimenter, and measurement conditions. The fluorescence X-ray spectra are recorded as image data. Figure 10 In the example shown, the filename of the image data is recorded.

[0107] Storage unit 13 stores report format 135. Report format 135 is information that specifies the format of reports related to analysis results. Figure 11 This is a conceptual diagram representing an example of the content of Report Format 135. For example, Report Format 135 specifies the data format or text format within the report. For example, Report Format 135 specifies various items included in the report, such as title, date, experimenter, or experimental conditions. For example, Report Format 135 specifies the arrangement of various items included in the report within the report. For example, Report Format 135 specifies figures included in the report, such as images in a spectrum or tables of quantitative results. For example, Report Format 135 specifies the display format or arrangement of images or figures included in the report.

[0108] Figure 12This is a flowchart illustrating an example of the processing steps performed by the information processing system 100 to create a report related to the analysis results in the analysis device 3. Information processing device 1 receives a query containing a report creation request (S31). In S31, the user 5 operates the operation unit 25 to input a query into the terminal device 2, which then sends the query to information processing device 1. Information processing device 1 obtains the query by receiving the query sent from the terminal device 2. The query contains a request to create a report related to the analysis results, such as "Please create a report that applies the analysis results ***.** to the report format." In the example query, "***.**" represents the filename of the analysis results.

[0109] Information processing device 1 obtains analysis results from analysis device 3 (S32). In S32, the arithmetic unit 11 causes the communication unit 15 to send a request for analysis results to terminal device 2. At this time, the arithmetic unit 11 may also send a request for a specific analysis result specified in the query. Terminal device 2 receives the request for analysis results and sends it to analysis device 3. Control unit 31 reads the analysis results according to the request for analysis results and sends them to terminal device 2. Terminal device 2 receives the analysis results and sends them to information processing device 1. Information processing device 1 obtains analysis results by receiving analysis results through communication unit 15. The arithmetic unit 11 stores the obtained analysis results in storage unit 13. Alternatively, information processing device 1 may pre-store the analysis results in storage unit 13 and obtain the analysis results by reading them from storage unit 13.

[0110] Information processing device 1 acquires report format 135 (S33). In S33, the arithmetic unit 11 acquires report format 135 by reading it from storage unit 13. Alternatively, if multiple report formats 135 are stored in storage unit 13, and information specifying any report format 135 is included in a query, information processing device 1 can also acquire a specific report format 135 specified in the query. Alternatively, in S31, information processing device 1 can also acquire the report format by acquiring a query containing the report format. In this case, information processing device 1 can perform the processing in S31 without performing the processing in S33.

[0111] Next, the information processing device 1 generates a prompt (S34) containing the obtained query, the obtained analysis results, and report format 135. In S34, the arithmetic unit 11 generates a prompt requesting the generation of a report related to the analysis results according to report format 135 based on the query.

[0112] Information processing device 1 retrieves a report from language generation model 132 based on a prompt (S35). In S35, the calculation unit 11 inputs the generated prompt to language generation model 132. Language generation model 132 performs calculations based on the prompt input and outputs a response. The prompt requests the generation of a report related to the analysis results, therefore language generation model 132 outputs a report related to the analysis results as a response. The calculation unit 11 retrieves the report output by language generation model 132. The calculation unit 11 stores the retrieved report in storage unit 13.

[0113] Alternatively, the information processing device 1 can also perform the following processing: in S34, a prompt that does not include the analysis results and report format 135 is created; in S35, the prompt that includes the analysis results and report format 135 is input into the language generation model 132. Alternatively, in S34, the information processing device 1 can also create a prompt that includes information such as the storage address of the analysis results and report format 135 for reference, but does not include the analysis results and report format 135 itself.

[0114] Information processing device 1 outputs a report related to the analysis results (S36). In S36, communication unit 15 sends the obtained report to terminal device 2 via first communication network 41. Terminal device 2 receives the report, and arithmetic unit 21 outputs the report by displaying the received report on display unit 26. At this time, arithmetic unit 21 generates an image containing the report and displays the generated image on display unit 26.

[0115] Figure 13 This is a schematic diagram illustrating an example of a report related to the analysis results. The report may be formatted as HTML (HyperText Markup Language), PDF (Portable Document Format), or an image. The content of the analysis results is configured according to report format 135, displaying the results in an easily observable and organized manner. User 5 confirms the output report. By displaying the analysis results in an appropriately organized report format, user 5 can easily understand the analysis results from analysis device 3.

[0116] The information processing device 1 performs a process that uses the evaluations obtained from the user 5 to enable the language generation model 132 to learn. The information processing device 1 stores training data 136 containing the evaluations obtained in S113 in the storage unit 13. In the training data 136, the query, the response output by the language generation model 132 based on the prompt containing the query, and the evaluation for the response are recorded in relation to each other. Each time an evaluation is obtained in S113, the computing unit 11 stores the query, response, and evaluation in relation to each other in the storage unit 13, thereby generating the training data 136. The combination of query, response, and evaluation is recorded as a dataset, and multiple datasets are recorded in the training data 136.

[0117] The information processing device 1 uses training data to learn the language generation model 132. Figure 14 This is a flowchart illustrating an example of the steps involved in the learning process performed by the information processing device 1. The information processing device 1 acquires the training data 136 by reading the training data 136 stored in the storage unit 13 from the arithmetic unit 11 (S41). In S41, the information processing device 1 may also acquire the training data 136 from an external source.

[0118] Next, the information processing device 1 uses the training data 136 to train the language generation model 132 (S42). In S42, the computation unit 11 inputs a prompt containing the query included in the training data 136 into 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 computation unit 11 adjusts the parameters of the language generation model 132 in a manner that the difference is reduced when the evaluation associated with the query in the training data 136 is high and increased when the evaluation is low. The computation unit 11 repeatedly processes multiple datasets recorded in the training data 136 and adjusts the parameters of the language generation model 132's computation, thereby training the language generation model 132.

[0119] The arithmetic unit 11 stores the final parameters adjusted in S42 in the storage unit 13. After S42, the information processing device 1 ends the learning process. The processes of S41 to S42 can also be executed in other devices different from the information processing device 1. After the language generation model 132 is learned, the learned language generation model 132 is used to perform processes such as S101 to S114. By using the language generation model 132 learned using the evaluation from the user 5, a more appropriate response can be obtained.

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

[0121] In this embodiment, the analytical apparatus 3 is shown to include a temperature sensor 342, but the analytical apparatus 3 may also include sensors other than the temperature sensor 342. For example, the analytical apparatus 3 may also include a sensor that measures the pressure or vacuum level inside the sample chamber 38. Even if the analytical apparatus 3 includes a sensor other than the temperature sensor 342, the information processing unit 1 can still perform the processing steps S23 and S101 to S114 using the measurement data from the sensor.

[0122] The information processing device 1 can also process responses corresponding to the state of the analysis device 3, such as the consumption of consumables, the deterioration of various parts of the analysis device 3 (e.g., 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 can also process the acquisition of the state of the analysis device 3 by methods other than using sensors. For example, the control unit 31 measures the cumulative operating time of the analysis device 3, the information processing device 1 obtains the cumulative operating time from the control unit 31, and determines the consumption of consumables or the deterioration of various parts of the analysis device 3 based on the cumulative operating time. Alternatively, the control unit 31 determines the state of the analysis device 3 based on the cumulative operating time, and the information processing device 1 obtains the determined state of the analysis device 3 from the control unit 31. Alternatively, the control unit 31 determines the position of the sample stage 36 based on the control history of the drive unit 361, and the information processing device 1 obtains the position of the sample stage 36 from the control unit 31. After obtaining the state of the analysis device 3, the information processing device 1 executes processes S101 to S114 that utilize the obtained state of the analysis device 3. In S101, the information processing device 1 obtains a query corresponding to the state of the analysis device 3. Through the processes S101 to S114, the information processing device 1 obtains a response corresponding to the state of the analysis device 3.

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

[0124] <Implementation Method 2> Figure 15This is a schematic diagram illustrating an example configuration of the information processing system 100 in Embodiment 2. In Embodiment 2, the information processing system 100 includes multiple terminal devices 2 and multiple analysis devices 3. The multiple terminal devices 2 communicate with the information processing device 1 via a first communication network 41. Each terminal device 2 communicates with the multiple analysis devices 3 via a second communication network 42. The multiple analysis devices 3 include different types of analysis devices 3. For example, some analysis devices 3 are X-ray analysis devices, and others are Raman analysis devices. The structure of the terminal devices 2 is the same as in Embodiment 1.

[0125] Figure 16 This is a block diagram illustrating an example of the internal configuration of the information processing apparatus 1 in Embodiment 2. The information processing apparatus 1 is configured using a computer, such as a server device. Similar to Embodiment 1, the information processing apparatus 1 includes an arithmetic unit 11, a memory 12, a storage unit 13, a reading unit 14, and a communication unit 15. Also similar to Embodiment 1, the information processing apparatus 1 includes a language generation model 132, and the storage unit 13 stores a computer program 131, a report format 135, and training data 136. Furthermore, the storage unit 13 stores a command database (DB) 137 and a text database 138.

[0126] Command DB137 contains multiple command data, such as first command data and second command data. Each command data in Command DB137, like the command data 133 in Embodiment 1, records the specifications of the command used to cause the analysis device 3 to perform processing. Command DB137 contains command data according to the type of analysis device 3. That is, each command data in Command DB137 records the specifications of the command used to cause any one of the multiple analysis devices 3 to perform processing.

[0127] The text DB138 contains multiple text data, such as first text data and second text data. Each text data in the text DB138 records text related to the analysis device 3, such as instructions, similar to the text data 134 in Embodiment 1. The text data in the text DB138 is organized according to the type of analysis device 3. That is, each text data in the text DB138 records text related to any one of the multiple analysis devices 3.

[0128] In implementation 2, the information processing system 100 also uses the language generation model 132 to control each analysis device 3. Figure 17 This is a flowchart illustrating an example of the processing steps performed by the information processing device 1 of Embodiment 2 in order to control the analysis device 3.

[0129] Information processing device 1 obtains the type of analysis device 3 of the controlled object and a query related to analysis device 3 (S51). In S51, the user 5 operates the operation unit 25 to input the query into the terminal device 2, and the terminal device 2 sends the query to information processing device 1. Furthermore, the terminal device 2 sends the type of analysis device 3 of the controlled object to 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 sends the stored information to information processing device 1. For example, the terminal device 2 obtains information indicating the type of analysis device 3 from the control unit 31 of the analysis device 3 of the controlled object and sends the obtained information to information processing device 1. Information processing device 1 receives the type of analysis device 3 and the query sent from the terminal device 2 via the communication unit 15, thereby obtaining 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 arithmetic unit 11 selects command data corresponding to the type of the acquired analysis device 3 from the multiple command data contained in the command DB137. That is, the arithmetic unit 11 selects command data that records the specifications of the commands used to cause the analysis device 3 of the controlled object to perform processing.

[0131] Information processing device 1 selects text data based on the type of analysis device 3 (S53). In S53, the arithmetic unit 11 selects text data corresponding to the type of analysis device 3 obtained from multiple text data contained in text DB138. That is, the arithmetic unit 11 selects text data that records text related to the analysis device 3 of the controlled object.

[0132] Next, the information processing device 1 performs the same processing steps S102 to S114 as in Embodiment 1. In S102, the information processing device 1 retrieves the associated information related to the obtained query from the text data selected in S53, and obtains the retrieved associated information. In S103, a prompt containing the query, associated information, and command data selected in S52 is created. Through the processing in S104, the information processing device 1 obtains a response corresponding to the query, associated information, and command data from the language generation model 132. In this way, the information processing device 1 controls each of the analysis devices 3.

[0133] The information processing device 1 can control multiple analysis devices 3 separately. Even when the multiple analysis devices 3 are of different types, the information processing device 1 controls each analysis device 3 in response to the same query output corresponding to the response of each analysis device 3. For example, when a query is obtained that includes a request to automate the analysis steps, the information processing device 1 can automate the analysis steps in each analysis device 3 in response to the same query output corresponding to the response of each analysis device 3. In Embodiment 2, similar to Embodiment 1, the information processing device 1 can also perform processing corresponding to the processing results in each analysis device 3, images such as images of the sample 6, or measurement data from the sensors equipped in each analysis device 3.

[0134] In Embodiment 2, the information processing system 100 uses the language generation model 132 to process the generation of reports related to multiple analysis results from multiple analysis devices 3. The information processing system 100 performs the same S31 to S36 processes as in Embodiment 1. The information processing device 1 stores a report format 135, which specifies the format of a report integrating multiple analysis results from multiple analysis devices 3, in the storage unit 13. In S31, the information processing device 1 obtains a query request for generating a report integrating multiple analysis results. In S32, the information processing device 1 obtains multiple analysis results from multiple analysis devices 3. In S33, the information processing device 1 obtains the report format 135, which specifies the format of a report integrating multiple analysis results. In S34, the information processing device 1 obtains a report integrating multiple analysis results according to the report format 135.

[0135] Through processing steps S31 to S36, a report integrating multiple analysis results from multiple analysis devices 3 is output. The content of the analysis results is output while integrating multiple analysis results according to report format 135. For example, the analysis results from multiple analysis devices 3, such as X-ray analysis and Raman analysis, are summarized and output. The user 5 confirms the output report. By outputting multiple analysis results in an appropriately integrated manner, the user 5 can easily understand the analysis results from multiple analysis devices 3. By combining and outputting multiple analysis results from multiple analysis devices 3, the user 5 can obtain insights that cannot be obtained from a single analysis result.

[0136] This invention is not limited to the embodiments described above, and various modifications can be made within the scope of the claims. That is, embodiments obtained by combining technical means appropriately modified within the scope of the claims are also included within the technical scope of this invention.

[0137] The items described in each embodiment can be combined with each other. Furthermore, the independent and dependent claims described in the claims can be combined with each other in all combinations, regardless of the form of reference. Moreover, the claims may adopt a claim form that references two or more other claims (a claim form with multiple references), but are not limited to this. It is also possible to use a form that references at least one claim with multiple references (a multiple-reference, multiple-claim form). Explanation of reference numerals in the attached figures

[0138] 1. Information processing device 11. Arithmetic Department 13 Storage Department 131 Computer Programs 132 Language Generation Model 133 Command Data 134 text data 135 Report Format 136 training data 2. Terminal device 3. Analytical apparatus 41 First Communication Network 42 Second Communication Network 5. Users.

Claims

1. An information processing method, characterized in that, Obtain queries related to the analytical device. Using a language generation model, obtain the response related to the control of the analysis device corresponding to the query. Output the response.

2. The information processing method according to claim 1, characterized in that, The system stores command data that records the specifications of commands used to instruct the analysis device to perform processing. Using the language generation model, based on the query and the command data, a response containing commands for instructing the analysis device to perform specific processing is obtained. The commands contained in the response are input into the analysis device.

3. The information processing method according to claim 2, characterized in that, Determine whether the command contained in the response is a command that can be executed in the analysis device. If the command included in the response is a command that cannot be executed in the analysis device, the language generation model is used to obtain a new response containing the new command based on the query, the command data, and the command that cannot be executed in the analysis device.

4. The information processing method according to claim 3, characterized in that, The command data is corrected based on whether the command contained in the response is a command that can be executed in the analysis device.

5. The information processing method according to any one of claims 2 to 4, characterized in that, Obtain the processing results from the analysis device according to the commands input to the analysis device. Output the processing result.

6. The information processing method according to any one of claims 1 to 4, characterized in that, The query retrieves a request to generate a report related to the analysis results of the analysis device. Obtain the analysis results and report format of the analytical device. Using the language generation model, based on the query, the analysis results, and the format, obtain a report related to the analysis results, generated according to the format. Output the report.

7. The information processing method according to claim 6, characterized in that, Obtain multiple analytical results from multiple analytical devices. The query retrieves a request to create a report that integrates multiple analysis results. Using the language generation model, based on the query, the obtained multiple analysis results, and the format, a report that synthesizes the multiple analysis results is generated according to the format.

8. The information processing method according to any one of claims 1 to 7, characterized in that, Obtain the type of the analytical device. Using the language generation model, obtain the response corresponding to the query and the type of the analysis device.

9. The information processing method according to any one of claims 1 to 8, characterized in that, Using the language generation model, a response is obtained that includes notifications to the user related to the control of the analysis device, corresponding to the query. Output the notification.

10. The information processing method according to any one of claims 1 to 9, characterized in that, Measurement data, consisting of measured values, is obtained from the sensors provided by the analytical device. The response is obtained using the language generation model, based on the query and the measurement data.

11. The information processing method according to any one of claims 1 to 10, characterized in that, Obtain the image. Using the language generation model, obtain the response corresponding to the query and the image.

12. The information processing method according to claim 11, characterized in that, As the image, an image of the sample to be analyzed in the analytical apparatus is obtained. Using the language generation model, based on the query and the captured image, a response containing information for instructing the analysis device to analyze the sample is obtained.

13. The information processing method according to any one of claims 1 to 12, characterized in that, Obtain the processing results from the analytical device. Using the language generation model, the response is obtained based on the query and the processing result.

14. The information processing method according to any one of claims 1 to 13, characterized in that, Obtain user feedback on the output response. The language generation model learns based on training data containing the query, the response, and the evaluation.

15. The information processing method according to any one of claims 1 to 14, characterized in that, Pre-store text data related to the analysis device. Obtain the association information related to the query from the text data. Using the language generation model, the response is obtained based on the query and the extracted association information.

16. An information processing device, characterized in that, Equipped with a computing unit, The computing unit obtains queries related to the analysis device. The computation unit uses a language generation model to obtain a response related to the control of the analysis device corresponding to the query. The arithmetic unit outputs the response.

17. A computer program, characterized in that, The computer will perform the following processing: Obtain queries related to the analysis device; Using a language generation model, obtain the response related to the control of the analysis device corresponding to the query; and Output the response.

Images