methods of summarizing articles

Abstract

This invention relates to a technique for summarizing an article. A method executed by an information processing device includes the following steps: if a first article is obtained, generating one or more identification information corresponding to one or more object information contained in the first article; obtaining an article obtained by replacing each object information contained in the first article with each identification information as a second article; inputting the second article into an external generative AI and obtaining a summary of the second article output from the generative AI; and outputting an article obtained by replacing each identification information contained in the summary of the second article with each object information as a summary of the first article.

Description

Technical Field

[0001] This invention relates to a method for summarizing an article. Background Technology

[0002] Previously, techniques for summarizing articles were known. For example, Patent Document 1 discloses a technique for using a third-party generative AI to provide a concise summary of a long article.

[0003] Patent Document 1: Japanese Patent Application Publication No. 2024-150099 Summary of the Invention

[0004] When articles are input into external generative AI, such as third-party cloud services, to generate summaries, if the article contains information that should be kept confidential (e.g., personal information), this information may be unintentionally used by the generative AI for learning. As a solution to this problem, one approach is to mask the article before inputting it into the generative AI. However, due to the information loss caused by masking, the generative AI may not be able to correctly understand the content of the article, potentially reducing the accuracy of the output summary. In view of this, the object of the present invention is to improve the technique for summarizing articles.

[0005] An embodiment of the present invention relates to a method executed by an information processing device, which includes the following steps:

[0006] If the first article is obtained, then generate one or more identification information corresponding to one or more object information contained in the first article;

[0007] The article obtained by replacing each of the object information contained in the first article with each of the identification information is used as the second article;

[0008] The second article is input into an external generative AI, and a summary of the second article output by the generative AI is obtained; and

[0009] The output is the article obtained by replacing each of the identification information contained in the summary of the second article with each of the object information, and it is used as the summary of the first article.

[0010] Invention Effects

[0011] According to one embodiment of the present invention, the technique for summarizing an article is improved. Attached Figure Description

[0012] Figure 1 This is a block diagram illustrating the general structure of a system according to an embodiment of the present invention.

[0013] Figure 2 It is a flowchart illustrating the operation of an information processing device. Detailed Implementation

[0014] (Summary of this implementation method)

[0015] refer to Figure 1 The following is a summary description of a system 1 according to an embodiment of the present invention. System 1 includes an information processing device 10 and a terminal device 20. The information processing device 10 and the terminal device 20 are communicatively connected via a network 30, including the Internet or a mobile communication network.

[0016] The information processing device 10 includes one or more computers, such as a server device, capable of communicating with each other. The information processing device 10 generates a summary based on the article.

[0017] Terminal device 20 includes any information processing terminal such as a laptop computer, tablet computer, or smartphone. Terminal device 20 is operated by a user.

[0018] First, an overview of this embodiment will be given; detailed descriptions will follow. The method described in this embodiment is performed by an information processing device 10. Upon acquiring a first article, the information processing device 10 generates at least one identification information corresponding to each of the one or more object information items contained in the first article. The object information is information that should be kept confidential. The identification information is, for example, information indicating the type of object information. The information processing device 10 acquires an article obtained by replacing each object information item contained in the first article with the corresponding identification information as a second article. The information processing device 10 inputs the second article into an external generative AI and acquires a summary of the second article output from the generative AI. Furthermore, the information processing device 10 outputs an article obtained by replacing each identification information item contained in the summary of the second article with the corresponding object information as a summary of the first article.

[0019] The method described in this embodiment can obtain an article summary and prevent object information (personal information, etc.) from being input into external generative AI. This prevents object information from being unintentionally used for the learning of generative AI.

[0020] (Composition of information processing device 10)

[0021] like Figure 1 As shown, the information processing device 10 includes a control unit 11, a communication unit 12, and a storage unit 13.

[0022] The control unit 11 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or combinations thereof. The processor may be, for example, a general-purpose processor such as a Central Processing Unit (CPU) or a Graphics Processing Unit (GPU), or a dedicated processor specifically designed for particular processing, but is not limited to these. The programmable circuit may be, for example, a Field-Programmable Gate Array (FPGA), but is not limited to these. The dedicated circuit may be, for example, an Application Specific Integrated Circuit (ASIC), but is not limited to these. The control unit 11 controls the various parts of the information processing device 10 while performing processing related to the operation of the information processing device 10.

[0023] The communication unit 12 includes at least one communication interface connected to the network 30. The communication interface may correspond to mobile communication standards such as 5G, wired local area network (LAN) communication standards, or wireless LAN communication standards, but is not limited to these and may correspond to any communication standard.

[0024] Storage unit 13 includes one or more memory units. Each memory unit included in storage unit 13 can function as a main storage device, an auxiliary storage device, or a cache memory, for example. Storage unit 13 stores any information used for the operation of information processing device 10. Furthermore, storage unit 13 can store, for example, system programs, application programs, and embedded software. Storage unit 13 can store any information related to an article's summary. The information stored in storage unit 13 can be updated, for example, based on information obtained from network 30 via communication unit 12.

[0025] (Structure of terminal device 20)

[0026] like Figure 1 As shown, the terminal device 20 includes a control unit 21, a communication unit 22, an input unit 23, and an output unit 24.

[0027] The control unit 21 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or combinations thereof. The processor may be, for example, a general-purpose processor such as a CPU or GPU, or a dedicated processor specifically designed for particular processing, but is not limited to these. The programmable circuit may be, for example, an FPGA, but is not limited to these. The dedicated circuit may be, for example, an ASIC, but is not limited to these. The control unit 21 controls the various parts of the terminal device 20 while performing processing related to the operation of the terminal device 20.

[0028] The communication unit 22 includes at least one communication interface connected to the network 30. The communication interface may correspond to mobile communication standards such as 5G, wired LAN communication standards, or wireless LAN communication standards, but is not limited to these and may correspond to any communication standard.

[0029] The input unit 23 includes one or more input interfaces. The input unit 23 accepts input information for the operation of the terminal device 20. The input interface can be, for example, a physical key, a capacitive key, a pointing device, a touchscreen integrated with the display of the output unit 24, or a microphone for accepting voice input. The input unit 23 can be connected to the terminal device 20 as an external input device, replacing the device itself. As a connection method, it can use any method such as Universal Serial Bus (USB), High-Definition Multimedia Interface (HDMI) (registered trademark), or Bluetooth (registered trademark).

[0030] The output unit 24 includes one or more output interfaces. The output interface may be, for example, a display that shows information as an image or a speaker that outputs information as sound. The output unit 24 outputs information obtained through the operation of the terminal device 20. The output unit 24 may not be provided on the terminal device 20, but may be connected to the terminal device 20 as an external display device. As a connection method, any method such as USB, HDMI (registered trademark), or Bluetooth (registered trademark) can be used.

[0031] (Operation flow of information processing device 10)

[0032] refer to Figure 2 The operation of the information processing device 10 will be explained below. Communication between the information processing device 10 and the terminal device 20 will be conducted via communication units 12 and 22 and network 30.

[0033] S1: The control unit 11 of the information processing device 10 obtains the first article from the terminal device 20. For example, the control unit 11 can receive the first article input into the input unit 23 of the terminal device 20 from the terminal device 20.

[0034] S2: The control unit 11 generates one or more identification information corresponding to one or more object information contained in the first article.

[0035] The object information is information that should be kept confidential. Object information includes, for example, personal information, technical information, or product information. Personal information may include, for example, name, age, address, date of birth, family structure, identification document number (e.g., personal identification number card, passport, driver's license, student ID, or employee ID), education, work experience, or medical history. Technical information may include, for example, the components, steps, or parameters required to implement the specified technology. Product information may include, for example, the name, price, or nature of the product. Control unit 11 may, for example, use natural language processing to extract the object information from the first article. The object information may be pre-stored in storage unit 13 and generated by searching storage unit 13 when the first article is obtained. Alternatively, the object information may also be generated from the object information using natural language processing.

[0036] Identification information is used to keep the content of object information confidential. Identification information can be category information indicating the type of object information. The type of object information may include, for example, personal information, technical information, product information, etc. For example, if the object information is a name, the identification information could be "Personal Information 01" or "Name Information 01". For example, natural language processing can be used to generate more than one identification piece of information.

[0037] S3: The control unit 11 stores the correspondence between each object information and each identification information in the storage unit 13. The correspondence is used later when obtaining the abstract of the first article.

[0038] S4: The control unit 11 obtains the article obtained by replacing the object information contained in the first article with the identification information and uses it as the second article.

[0039] S5: Control unit 11 inputs the second article into an external generative AI and obtains a summary of the second article output by the generative AI. The external generative AI could be, for example, a generative AI based on a third-party cloud service. For example, the output of the summary could utilize a large-scale language model. In the second article, object information is replaced with recognition information, thus preventing object information from being input into the external generative AI and from being unintentionally used by the generative AI.

[0040] S6: The control unit 11 replaces each identification information contained in the summary of the second article with each object information, and outputs the resulting article as the summary of the first article. This prevents the object information from being learned by external generative AI while simultaneously obtaining the summary of the first article. The output processing can employ any method. For example, the control unit 11 can send the summary of the first article to the terminal device 20 and output it to the output unit 24 of the terminal device 20. The summary of the first article can, for example, be displayed on the display of the output unit 24, or output as sound to the speaker of the output unit 24.

[0041] The control unit 11 can replace the identification information contained in the second document with the object information by referring to the correspondence stored in the storage unit 13. From the point of view of protecting secrets, the control unit 11 can delete the correspondence from the storage unit 13 after the object information is replaced.

[0042] Although the present invention has been described with reference to the accompanying drawings and embodiments, those skilled in the art will understand that various modifications and variations can be made based on the present invention. Therefore, it should be noted that these modifications and variations are included within the scope of the present invention. For example, the functions included in each structural component or step can be reconfigured in a logically consistent manner, and multiple structural components or steps can be combined into one or divided. For example, in the above embodiments, it is also possible to implement an embodiment in which the structure and operation of the information processing device 10 or the terminal device 20 are distributed among multiple computers capable of communicating with each other. Furthermore, for example, in the above embodiments, the information processing device 10 can be the terminal device 20.

[0043] Identification information can be information belonging to the same category as the object information but with different content (virtual information). For example, if the object information is a name, the identification information could be "Yamada Taro". Some generative AIs sometimes understand articles more accurately by using virtual information representing the specific content of the object information (e.g., the pseudonym "Yamada Taro") than by using category information representing the object's type (e.g., "Personal Information 01"). In such cases, by utilizing virtual information, an accurate summary can be obtained.

[0044] Symbol Explanation

[0045] 1-System, 10-Information processing device, 11-Control unit, 12-Communication unit, 13-Storage unit, 20-Terminal device, 21-Control unit, 22-Communication unit, 23-Input unit, 24-Output unit, 30-Network.

Claims

1. A method for summarizing an article, performed by an information processing device, characterized by comprising the following steps: If the first article is obtained, then generate one or more identification information corresponding to one or more object information contained in the first article; The article obtained by replacing each of the object information contained in the first article with each of the identification information is used as the second article; The second article is input into an external generative AI, and a summary of the second article output from the generative AI is obtained. and The output is the article obtained by replacing each of the identification information contained in the summary of the second article with each of the object information, and it is used as the summary of the first article.

2. The method for summarizing an article according to claim 1, characterized in that, The object information includes at least one of the following: personal information, technical information, and product information.

3. The method for summarizing an article according to claim 1, characterized in that, It also includes the following steps: Store the correspondence between the information of each object and the identification information.

4. The method for summarizing an article according to claim 1, characterized in that, The identification information is stored in advance.

5. The method for summarizing an article according to claim 1, characterized in that, The identification information is information that belongs to the same category as the object information but has different content.

Images