Information processing device and information processing program

Abstract

To provide an information processing device capable of generating text with a reduced risk of information leakage compared to conventional methods, and an information processing program.SOLUTION: An information processing device of an embodiment comprises a first acquisition unit, a removal unit, and a second acquisition unit. The first acquisition unit acquires first text information from first audio data. The removal unit removes elements satisfying predetermined conditions as confidential information from the first text information. The second acquisition unit inputs the first text information from which the confidential information has been removed into a large language model configured to instruct the summary of the first text information, and acquires the summary information generated by the large language model.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] Embodiments of the present invention relate to an information processing apparatus and an information processing program.

Background Art

[0002] In recent years, a technique for automatically generating a sentence using an AI model called a large language model (LLM) has been proposed. A large language model is an AI model pre-trained with a large corpus in the field of natural language processing. For example, when text describing an inquiry content is input together with instruction information (also referred to as a prompt) that indicates an output form, it is configured to function so as to generate and output a response sentence (hereinafter, also referred to as summary information) regarding the inquiry content in the indicated form.

[0003] By the way, in operations such as a support center, an inquiry from a customer is received by phone and responded to. In such an operation, it is assumed that voice data of a phone call is converted into text and input into the above-described large language model to generate summary information regarding the inquiry content.

[0004] However, voice data may contain information that should be kept confidential. For example, it may contain customer information regarding an individual customer or information regarding an organization (company, etc.) to which the customer belongs. In addition, the handling of the text input into the large language model after input is often unclear, and for example, it may be used for re-learning.

[0005] Therefore, inadvertently inputting text containing confidential information into a large language model may lead to a risk of information leakage of the confidential information, and there is room for further improvement from the viewpoint of security.

Summary of the Invention

Problems to be Solved by the Invention

[0006] The problem that this invention aims to solve is to provide an information processing device and an information processing program that can generate documents with a reduced risk of information leakage compared to conventional methods. [Means for solving the problem]

[0007] The information processing device of the embodiment is Includes inquiries regarding the specified target. A first acquisition unit that acquires first text information from first audio data, The acquisition by the first acquisition unit A removal unit removes elements that meet predetermined conditions from the first text information as confidential information, and a second acquisition unit inputs the first text information from which the removal unit has removed the confidential information into a large-scale language model that is configured to generate a summary of the first text information, and acquires the summary information generated by the large-scale language model. An extraction unit extracts similar information from a storage device that stores information related to the subject, which is similar in content to the summary information acquired by the second acquisition unit; a third acquisition unit combines the summary information acquired by the second acquisition unit and the similar information extracted by the extraction unit to obtain checklist information in which the contents of both pieces of information are listed in bullet points; and an output unit outputs the checklist information acquired by the third acquisition unit. It is equipped with. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a schematic diagram showing the general configuration of the information processing system according to this embodiment. [Figure 2] Figure 2 is a block diagram showing an example of the hardware configuration of the information processing device according to this embodiment. [Figure 3] Figure 3 is a block diagram showing an example of the functional configuration of the information processing device according to this embodiment. [Figure 4] Figure 4 is a schematic diagram showing an example of checklist information according to this embodiment. [Figure 5] Figure 5 is a schematic diagram showing an example of reporting information according to this embodiment. [Figure 6] Figure 6 is a flowchart showing an example of the control process of the information processing apparatus according to this embodiment. [Figure 7] Figure 7 is a flowchart showing an example of the control process of the information processing apparatus according to this embodiment. [Figure 8] Figure 8 is a flowchart showing an example of the control process of the information processing apparatus according to this embodiment. [Modes for carrying out the invention]

[0009] The embodiments of the information processing system will be described below with reference to the drawings. Furthermore, the present invention is not limited by the following embodiments, and the components in the following embodiments include those easily conceivable by those skilled in the art, those substantially identical, and those within the scope of equivalents. Moreover, various omissions, substitutions, modifications, and combinations of components can be made without departing from the spirit of the following embodiments.

[0010] Figure 1 is a schematic diagram showing the general configuration of an information processing system 10 according to an embodiment. The information processing system 10 is a system comprising a voice input device 1, an information processing device 2, and a display device 3.

[0011] Voice input device 1 is, for example, a telephone used by customers or maintenance personnel. Voice input device 1 acquires voice data from customers or maintenance personnel and outputs the acquired voice data to information processing device 2. Customers use voice input device 1, for example, when their equipment malfunctions and they make inquiries to the management company that manages the malfunctioning equipment, such as requesting repairs for the malfunctioning equipment.

[0012] A maintenance worker uses the voice input device 1, for example, when contacting the company to which the maintenance worker belongs regarding the maintenance of a malfunctioning piece of equipment. The voice input device 1 is connected to the information processing device 2 via a communication line, such as a mobile network.

[0013] Information Processing Unit 2 is a device used by a management company that manages equipment used by customers. Information Processing Unit 2 also includes, for example, a Large Language Model (LLM) and consists of one or more computers. Information Processing Unit 2 is, for example, a device for receiving voice data from customers and maintenance personnel and responding to inquiries about that voice data.

[0014] The information processing device 2 is connected to the voice input device 1 via a communication line, such as a mobile network. The information processing device 2 is also connected to the display device 3 via a communication line, such as a mobile network.

[0015] Display device 3 is a device used by a management company that maintains or manages equipment used by customers. Display device 3 consists of one or more computers. Display device 3 is a device that, for example, allows a maintenance worker responding to a customer inquiry about a faulty piece of equipment to display information about the equipment to be repaired, as well as repair manuals, etc. Display device 3 is connected to information processing device 2 via a communication line, such as a mobile network.

[0016] Next, the hardware configuration of the information processing device 2 will be described using Figure 2. Figure 2 is a block diagram showing an example of the hardware configuration of the information processing device 2 according to this embodiment. As shown in Figure 2, the information processing device 2 includes a CPU (Central Processing Unit) 21, RAM (Random Access Memory) 22, ROM (Read Only Memory) 23, auxiliary storage device 24, and communication I / F (interface) 25.

[0017] The CPU 21 is the primary control unit. The RAM 22 stores programs and various data. The ROM 23 stores various programs. The auxiliary storage device 24 stores various programs and data. The communication interface 25 is an interface for data communication with the voice input device 1 and the display device 3.

[0018] The CPU 21, RAM 22, ROM 23, auxiliary storage device 24, and communication interface 25 are connected to each other via a bus 26. The CPU 21, RAM 22, and ROM 23 constitute the control unit. That is, the control unit executes the control processing of the information processing device 2, which will be described later, by having the CPU 21 operate according to the control program stored in the ROM 23 and auxiliary storage device 24 and loaded into the RAM 22.

[0019] The auxiliary storage device 24 is composed of a non-volatile memory such as a HDD (Hard Disc Drive) or a flash memory that retains stored information even when the power is turned off. The auxiliary storage device 24 stores the summary information, similarity information, checklist information, maintenance information, report response information, and progress management information generated by the information processing device 2. Details of the summary information, similarity information, checklist information, maintenance information, report response information, and progress management information will be described later.

[0020] The information processing device 2 is installed in a secure environment. A secure environment is a configuration for placement and use in a local environment. With this configuration, the input text information for performing processing of the large language model included in the information processing device 2, which will be described later, includes customer information of customers, etc. Therefore, the information processing device 2 can contribute to the protection of personal information by being placed and used in a local environment.

[0021] FIG. 3 is a block diagram showing an example of the functional configuration of the information processing device 2 according to the present embodiment. As shown in FIG. 3, the information processing device 2 includes a first reception unit 211, a first acquisition unit 212, a removal unit 213, a second acquisition unit 214, an extraction unit 215, a third acquisition unit 216, an output unit 217, a second reception unit 218, a fourth acquisition unit 219, a fifth acquisition unit 220, a sixth acquisition unit 221, a seventh acquisition unit 222, and an eighth acquisition unit 223. Note that the functional configuration of the information processing device 2 is not limited to this.

[0022] The first reception unit 211 receives the first audio data. Specifically, the first reception unit 211 receives the first audio data from a customer output by the audio input device 1. For example, the first audio data includes the name, phone number, email address of the customer who made the inquiry, the company name of the company to which the customer belongs, etc. Also, for example, the first audio data includes the content of the inquiry regarding the device used by the customer. Examples of the inquiry content include, for example, the failure, usage method, claim, etc. of the device used by the customer. Further, the first reception unit 211 stores the received first audio data in the auxiliary storage device 24.

[0023] The first acquisition unit 212 acquires first text information. Specifically, the first acquisition unit 212 acquires first text information from the first audio data received by the first receiving unit 211. The first acquisition unit 212 converts the first audio data into text by applying speech recognition processing or natural language processing (NLP) to the first audio data using known methods, and acquires the converted first text information. The first acquisition unit 212 also stores the acquired first text information in the auxiliary storage device 24.

[0024] The removal unit 213 removes elements from the text information that meet predetermined conditions as confidential information. Specifically, the removal unit 213 removes confidential information, which are elements that meet predetermined conditions, from the first text information acquired by the first acquisition unit 212.

[0025] The elements that meet the predetermined conditions are those related to confidential information. These are elements that have been predetermined to be kept confidential. For example, confidential information includes at least customer information relating to the customer as an individual. Specifically, confidential information includes customer information such as the name, telephone number, email address, and trade name of the company to which the customer belongs. The removal unit 213 also stores the first text information from which the confidential information has been removed in the auxiliary storage device 24.

[0026] The second acquisition unit 214 acquires summary information. Specifically, the second acquisition unit 214 inputs the first text information, from which the removal unit 213 has removed confidential information, into a large-scale language model that is configured to instruct the model to summarize the first text information, and acquires the summary information generated by the large-scale language model.

[0027] For example, the second acquisition unit 214 inputs the first text information and instruction information (also called a prompt) that instructs the first text information to be summarized into a large-scale language model using a known method, causing the large-scale language model to generate summary information that summarizes the first text information, and acquires the summary information generated by the large-scale language model. The summary information includes the case number of the summary information, the date and time the summary information was created, and the query requirements included in the first text information.

[0028] The second acquisition unit 214 generates a first prompt that includes, for example, assigning a case number, adding the creation date and time, assigning an inquiry title corresponding to the inquiry requirements, extracting inquiry content such as concerns, and listing the information in bullet points, in order to manage inquiries from multiple customers. The first prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the second acquisition unit 214. The second acquisition unit 214 stores the acquired summary information in the auxiliary storage device 24.

[0029] The extraction unit 215 extracts similar information. Specifically, the extraction unit 215 extracts similar information from an auxiliary storage device different from the auxiliary storage device 24 that is similar to the summary information acquired by the second acquisition unit 214. The similar information is similar to the content of the customer inquiry, and is information that contains similar content to, for example, the details of the equipment failure used by the customer, the method of using the equipment, the details of the failure, or the details of the equipment complaint, and is extracted. The extraction unit 215 then stores the extracted similar information in the auxiliary storage device 24.

[0030] The third acquisition unit 216 acquires checklist information. Specifically, the third acquisition unit 216 acquires checklist information corresponding to the summary information acquired by the second acquisition unit 214 and the similar information extracted by the extraction unit 215. The third acquisition unit 216 inputs the summary information, the similar information, and instruction information that instructs the combination of the summary information and the similar information into a large-scale language model using a known method, causing the large-scale language model to generate checklist information by combining the summary information and the similar information, and acquires the checklist information generated by the large-scale language model. The checklist information is a checklist that lists the contents of the summary information and the similar information in bullet points.

[0031] The third acquisition unit 216 generates a second prompt that, for example, lists the similar information extracted by the extraction unit 215 below the summary information acquired by the second acquisition unit 214, generates check items, adds a checkbox to the first sentence of the list, and, if there are duplicate cases where the requirements of the summary information and the requirements of the similar information overlap, lists the requirements of the remaining similar information excluding the overlapping requirements.

[0032] The second prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the third acquisition unit 216. The third acquisition unit 216 may also combine summary information and similar information for the checklist information without using a large-scale language model, and acquire the combined information as checklist information. The third acquisition unit 216 then stores the acquired checklist information in the auxiliary storage device 24.

[0033] The output unit 217 outputs checklist information. Specifically, the output unit 217 outputs the checklist information acquired by the third acquisition unit 216 to the display device 3. The checklist information output by the output unit 217 will now be explained using Figure 4. Figure 4 is a schematic diagram showing an example of checklist information according to this embodiment.

[0034] Figure 4 shows the checklist information 41. The checklist information 41 includes the case number, creation date, inquiry title, and inquiry requirements listed in bullet points with checkboxes, as well as similar requirements listed in bullet points with checkboxes. This allows maintenance personnel who maintain the equipment used by customers to improve the efficiency of their work by checking the checklist information 41.

[0035] Returning to Figure 3, the second receiving unit 218 receives the second audio data. Specifically, the second receiving unit 218 receives the second audio data from the maintenance personnel, which is output by the audio input device 1. For example, the second audio data includes the name of the customer who was present for the maintenance, the trade name of the company to which the customer belongs, the name of the maintenance personnel corresponding to the case in the checklist information 41, the case number, a reading of the check items, the details of the maintenance corresponding to the check items, and the results of the maintenance.

[0036] The fourth acquisition unit 219 acquires the second text information. Specifically, the fourth acquisition unit 219 acquires the second text information corresponding to the second audio data received by the second receiving unit 218. The fourth acquisition unit 219 converts the second audio data into text by performing speech recognition processing or natural language processing on the second audio data using known methods, and acquires the converted second text information.

[0037] The fifth acquisition unit 220 acquires second summary information. Specifically, the fifth acquisition unit 220 acquires second summary information (hereinafter also referred to as report information) that shows a summary of the checklist information 41 acquired by the third acquisition unit 216 and the second text information acquired by the fourth acquisition unit 219. The fifth acquisition unit 220 inputs the checklist information 41, the second text information, and instruction information that instructs the summarization of the checklist information 41 and the second text information into a large-scale language model using a known method, causing the large-scale language model to generate second summary information that shows a summary of the checklist information 41 and the second text information, and acquires the second summary information generated by the large-scale language model.

[0038] The fifth acquisition unit 220 generates a third prompt that, for example, states that the case number listed in the checklist information 41 matches, that the creation date and time should be added, that check results should be added to the check items corresponding to the inquiry requirements and similarity requirements, and that confidential information should not be included, as part of the maintenance response from the maintenance personnel. Confidential information in the third prompt includes customer information, such as the name of the customer who was present during the maintenance, the trade name of the company to which the customer belongs, etc.

[0039] The third prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the fifth acquisition unit 220. The fifth acquisition unit 220 may also combine the checklist information 41 and the second text information for the second summary information without using a large-scale language model, and acquire the combined information as the second summary information. The fifth acquisition unit 220 also stores the acquired summary information in the auxiliary storage device 24.

[0040] Here, the reporting information will be explained using Figure 5. Figure 5 is a schematic diagram showing an example of the reporting information according to this embodiment. The reporting information 51 shown in Figure 5 includes the contents described in Figure 4, as well as the check results of the check items corresponding to the inquiry requirements and similarity requirements. As a result, the fifth acquisition unit 220 can acquire the reporting information 51 from the second voice data regarding the maintenance response of the maintenance personnel, thereby reducing the effort required for maintenance personnel to create reports and improving the efficiency of the maintenance personnel's work. Furthermore, the fifth acquisition unit 220 stores the acquired reporting information 51 in the auxiliary storage device 24.

[0041] Returning to Figure 3, the sixth acquisition unit 221 acquires the first progress management information. Specifically, the sixth acquisition unit 221 acquires the first text information acquired by the first acquisition unit 212 and the first progress management information that manages the progress of the checklist, based on the checklist information. The sixth acquisition unit 221 inputs the first text information and instruction information that instructs the progress management of the first text information into a large-scale language model using a known method, causing the large-scale language model to generate the first progress management information corresponding to the first text information, and acquires the first progress management information generated by the large-scale language model.

[0042] Here, the first progress management information is, for example, a Gantt chart with a time axis on the horizontal axis and columns for recording maintenance check items on the vertical axis. The first progress management information is, for example, information for managing the progress of maintenance responses to customer inquiries.

[0043] The sixth acquisition unit 221 generates a fourth prompt that, for example, includes instructions for a customer inquiry, such as assigning a case number, adding a creation date and time, and adding information for progress management.

[0044] The fourth prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the sixth acquisition unit 221. The sixth acquisition unit 221 may also combine the first text information with the Gantt chart graph and acquire the combined information as the first progress management information without using a large-scale language model. The sixth acquisition unit 221 then stores the acquired first progress management information in the auxiliary storage device 24.

[0045] The seventh acquisition unit 222 acquires the second progress management information. Specifically, the seventh acquisition unit 222 acquires the second progress management information that corresponds to the first progress management information acquired by the sixth acquisition unit 221. In other words, the seventh acquisition unit 222 acquires the first progress management information and the second progress management information that manages the progress of the checklist based on the checklist information 41.

[0046] The seventh acquisition unit 222 inputs the first progress management information, the checklist information 41, and the instruction information that instructs the progress management of the first progress management information and the checklist information 41 into the large-scale language model using a known method, causing the large-scale language model to generate second progress management information corresponding to the second progress management information, and acquires the second progress management information generated by the large-scale language model.

[0047] Here, the second progress management information is, for example, a Gantt chart with a time axis on the horizontal axis and columns for recording maintenance check items on the vertical axis. The second progress management information is, for example, information for managing the progress of maintenance responses to customer inquiries. The second progress management information is information in which the contents of checklist information 41 are reflected in the first progress management information.

[0048] The seventh acquisition unit 222 generates a fifth prompt that, for example, includes instructions to add the check items included in the checklist information 41 acquired by the third acquisition unit 216 to the column for recording maintenance check items in the Gantt chart of the first progress management information.

[0049] The fifth prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the seventh acquisition unit 222. The seventh acquisition unit 222 may combine the first progress management information and the checklist information 41 with respect to the second progress management information without using a large-scale language model, and acquire the combined information as the second progress management information. The seventh acquisition unit 222 then stores the acquired second progress management information in the auxiliary storage device 24.

[0050] The eighth acquisition unit 223 acquires the third progress management information. Specifically, the eighth acquisition unit 223 acquires the third progress management information corresponding to the second progress management information acquired by the seventh acquisition unit 222. In other words, the eighth acquisition unit 223 acquires the third progress management information for managing the progress of the checklist, based on the second progress management information and the second summary information acquired by the fifth acquisition unit 220.

[0051] The eighth acquisition unit 223 inputs the second progress management information, the second summary information, and instruction information that instructs the progress management of the second progress management information and the second summary information into the large-scale language model using a known method, thereby causing the large-scale language model to generate second progress management information corresponding to the third progress management information, and acquires the third progress management information generated by the large-scale language model.

[0052] Here, the third progress management information is, for example, a Gantt chart with a time axis on the horizontal axis and columns for recording maintenance check items on the vertical axis. The third progress management information is, for example, information for managing the progress of maintenance responses to customer inquiries. The third progress management information is information in which the content of the second summary information is reflected in the second progress management information.

[0053] The eighth acquisition unit 223 generates a sixth prompt that includes instructions such as adding the maintenance response date and time described in the second summary information acquired by the fifth acquisition unit 220, which corresponds to the time axis of the maintenance check items in the Gantt chart of the second progress management information, and adding the maintenance response date and time described in the second text information acquired by the fourth acquisition unit 219.

[0054] The sixth prompt may be one that has been prepared in advance in the auxiliary storage device 24, or it may be dynamically generated by the eighth acquisition unit 223. The eighth acquisition unit 223 may combine the second progress management information and the second summary information with respect to the third progress management information without using a large-scale language model, and acquire the combined information as the third progress management information. The eighth acquisition unit 223 also stores the acquired third progress management information in the auxiliary storage device 24.

[0055] The 8th acquisition unit 223 acquires the 3rd progress management information, which is an updated version of the 2nd progress management information. This allows customers other than maintenance personnel to check the latest maintenance status without waiting for the maintenance personnel to update the progress management information. As a result, companies managing equipment can further strengthen their equipment support system.

[0056] Next, the control process of the information processing device 2 will be explained using Figures 6 to 8. Figures 6 to 8 are flowcharts showing an example of the control process of the information processing device 2 according to this embodiment.

[0057] In step S61 shown in Figure 6, the first receiving unit 211 receives first voice data from the customer output by the voice input device 1 (step S61). Subsequently, the first acquisition unit 212 acquires first text information from the first voice data received by the first receiving unit 211 (step S62).

[0058] Next, the removal unit 213 removes confidential information, which is an element that meets predetermined conditions, from the first text information acquired by the first acquisition unit 212 (step S63). Subsequently, the second acquisition unit 214 inputs the first text information, from which the removal unit 213 has removed the confidential information, into a large-scale language model that is configured to instruct the summarization of the first text information, and acquires the summary information generated by the large-scale language model (step S64).

[0059] Next, the extraction unit 215 extracts similar information from an auxiliary storage device different from the auxiliary storage device 24 that is similar to the summary information acquired by the second acquisition unit 214 (step S65). Subsequently, the third acquisition unit 216 acquires checklist information 41 corresponding to the summary information acquired by the second acquisition unit 214 and the similar information extracted by the extraction unit 215 (step S66).

[0060] Next, the output unit 217 outputs the checklist information 41 acquired by the third acquisition unit 216 to the display device 3 (step S67). Once this process is complete, the information processing device 2 finishes processing. This allows the maintenance personnel to perform their work based on the checklist information 41.

[0061] In step S71 shown in Figure 7, the second receiving unit 218 receives the second voice data from the maintenance worker output by the voice input device 1 (step S71). Subsequently, the fourth acquisition unit 219 acquires the second text information corresponding to the second voice data received by the second receiving unit 218 (step S72).

[0062] Next, the fifth acquisition unit 220 acquires report information 51 that shows a summary of the checklist information 41 acquired by the third acquisition unit 216 and the second text information acquired by the fourth acquisition unit 219 (step S73). Subsequently, the fifth acquisition unit 220 stores the acquired summary information in the auxiliary storage device 24 (step S74). When this process is completed, the processing of the information processing device 2 is terminated. As a result, maintenance personnel can reduce the effort required to create reports, thereby improving work efficiency.

[0063] In step S81 shown in Figure 8, the sixth acquisition unit 221 acquires the first text information acquired by the first acquisition unit 212 and first progress management information for managing the progress of the checklist based on the checklist (step S81). Subsequently, the seventh acquisition unit 222 acquires second progress management information corresponding to the first progress management information acquired by the sixth acquisition unit 221 (step S82).

[0064] Next, the eighth acquisition unit 223 acquires third progress management information corresponding to the second progress management information acquired by the seventh acquisition unit 222 (step S83). Subsequently, the eighth acquisition unit 223 stores the acquired third progress management information in the auxiliary storage device 24 (step S84). When this process is completed, the processing of the information processing device 2 is terminated.

[0065] This allows maintenance personnel to reduce the effort required for progress management, thereby improving work efficiency. Furthermore, since customers other than maintenance personnel can also check the progress, companies managing the equipment can strengthen their support system.

[0066] As described above, the information processing device 2 of this embodiment acquires first text information corresponding to audio data, performs natural language processing on the text information, and generates summary information corresponding to the text information that does not contain confidential information.

[0067] As a result, the summary information generated by the information processing device 2 does not contain confidential information, thus enabling the generation of documents with a reduced risk of information leakage.

[0068] The embodiments described above can also be modified and implemented as appropriate by changing some of the configurations or functions of each of the devices described above. Therefore, several modifications of the embodiments described above will be described below as other embodiments. In the following, we will mainly describe the differences from the embodiments described above, and will omit detailed explanations of points that are common with what has already been described. Furthermore, the modifications described below may be implemented individually or in combination as appropriate.

[0069] (First variation) The information processing device 2 stores the audio data in the auxiliary storage device 24, but it may delete it at any time. For example, the information processing device 2 may delete the data when the text information is generated or when the third progress management information is generated. This allows the information processing device 2 to contribute to the protection of personal information.

[0070] (Second variation) In the above-described embodiment, the information processing device 2 equipped with a large-scale language model was configured to be used in a local environment. However, the information processing device 2 equipped with a large-scale language model may also be configured to be used in a highly secure cloud environment. A highly secure cloud environment is one that is not exposed to the internet and is connected to the company's server environment via a highly confidential communication line.

[0071] Furthermore, the program executed by the information processing device 2 of this embodiment may be stored on a computer connected to a network such as the Internet and provided by downloading it via the network. Alternatively, the program executed by the information processing device 2 of this embodiment and its modified form may be provided or distributed via a network such as the Internet.

[0072] The programs executed by each device in the above-described embodiments are provided pre-installed in ROM or auxiliary storage devices. The programs executed by each device in the above-described embodiments may also be provided as files in an installable or executable format, recorded on a computer-readable recording medium such as a CD-ROM, flexible disk (FD), CD-R, or DVD (Digital Versatile Disk).

[0073] Furthermore, the programs executed by each of the above-described embodiments may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Alternatively, the programs executed by each of the above-described embodiments may be provided or distributed via a network such as the Internet.

[0074] Although embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments and their variations can be implemented in a variety of other forms, and various omissions, substitutions, changes, and combinations can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of Symbols]

[0075] 1. Voice input device 2. Information Processing Device 3 Display device 211 First Receiving Unit 212 First acquisition part 213 Removal section 214 Second Acquisition Department 215 Extraction part 216 Third Acquisition Department 217 Output section 218 Second Receiving Unit 219 4th Acquisition Department 220 Fifth Acquisition Department 221 6th Acquisition Department 222 7th Acquisition Department 223 8th Acquisition Department [Prior art documents] [Patent Documents]

[0076] [Patent Document 1] Japanese Patent Publication No. 2022-87521

Claims

1. A first acquisition unit that acquires first text information from first audio data including inquiry content relating to a predetermined object, A removal unit removes elements that meet predetermined conditions from the first text information acquired by the first acquisition unit as confidential information, The removal unit inputs the first text information from which the confidential information has been removed into a large-scale language model that is configured to generate a summary of the first text information, and the second acquisition unit acquires the summary information generated by the large-scale language model. An extraction unit extracts similar information from a storage device that stores information related to the subject, which is similar to the content of the summary information acquired by the second acquisition unit. A third acquisition unit combines the summary information acquired by the second acquisition unit with the similar information extracted by the extraction unit and acquires checklist information in which the contents of both pieces of information are listed in bullet points. The output unit outputs the checklist information acquired by the third acquisition unit, An information processing device equipped with the following features.

2. The aforementioned confidential information includes at least customer information, The aforementioned customer information includes the customer's name, telephone number, email address, and the trade name of the company to which the customer belongs. The summary information includes the case number of the summary information, the date and time the summary information was created, and the query requirements included in the first text information. The information processing apparatus according to claim 1.

3. A fourth acquisition unit that acquires second text information from second voice data spoken by the person in charge of responding to the inquiry, A fifth acquisition unit inputs the checklist information acquired by the third acquisition unit and the second text information acquired by the fourth acquisition unit into a large-scale language model that is configured to generate a summary of the checklist information and the second text information, and acquires the second summary information generated by the large-scale language model. The system further comprises a storage unit for storing the second summary information in a storage device, The information processing apparatus according to claim 1.

4. A sixth acquisition unit inputs the first text information acquired by the first acquisition unit and the checklist information acquired by the third acquisition unit into a large-scale language model that is configured to generate progress management information for managing the progress of check items listed in the checklist information from the first text information and the checklist information, and acquires the progress management information generated by the large-scale language model, The system further includes a storage unit for storing the aforementioned progress management information in a storage device. The information processing apparatus according to claim 1.

5. The computer of the information processing device, A first acquisition unit that acquires first text information from first audio data containing inquiry content related to a predetermined target, A removal unit removes elements that meet predetermined conditions from the first text information acquired by the first acquisition unit as confidential information, The removal unit inputs the first text information from which the confidential information has been removed into a large-scale language model that is configured to generate a summary of the first text information, and the second acquisition unit acquires the summary information generated by the large-scale language model. An extraction unit extracts similar information from a storage device that stores information related to the subject, which is similar to the content of the summary information acquired by the second acquisition unit. A third acquisition unit combines the summary information acquired by the second acquisition unit with the similar information extracted by the extraction unit and acquires checklist information in which the contents of both pieces of information are listed in bullet points. The output unit outputs the checklist information acquired by the third acquisition unit, An information processing program that enables a function to work.

Images