Information processing method and device based on large model, electronic equipment and storage medium

Abstract

The invention provides an information processing method and device based on a large model, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence such as data security, large models, machine learning and task reasoning. According to the specific implementation scheme, input information of a pre-input large model is obtained; using a pre-trained sensitive information identification model to identify whether the input information includes sensitive information; responding to sensitive information included in the input information, and based on a preset desensitization rule, desensitizing the input information to obtain desensitized input information. According to the technology disclosed by the invention, the data input into the large model can be desensitized, so that the security of the data is effectively improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and more particularly to the fields of data security, large models, and artificial intelligence technologies such as machine learning and task reasoning, and more particularly to an information processing method, device, electronic device, and storage medium based on a large model. Background Art

[0002] In the field of big data and artificial intelligence, data privacy protection has become an issue that cannot be ignored. Especially in the training process of large models, how to protect data privacy is one of the key issues.

[0003] Currently, the main privacy protection technologies include data encryption, homomorphic encryption, etc. The implementation of these technologies usually requires encrypting the data and retraining the model based on the encrypted data. Summary of the invention

[0004] The present disclosure provides a large model-based information processing method, device, electronic device and storage medium.

[0005] According to one aspect of the present disclosure, there is provided an information processing method based on a large model, comprising:

[0006] Get input information for pre-input large model;

[0007] Using a pre-trained sensitive information recognition model to identify whether the input information includes sensitive information;

[0008] In response to the input information including sensitive information, the input information is desensitized based on a preset desensitization rule to obtain desensitized input information.

[0009] According to another aspect of the present disclosure, there is provided an information processing device based on a large model, comprising:

[0010] An acquisition module is used to obtain input information pre-input into the large model;

[0011] An identification module, used to use a pre-trained sensitive information identification model to identify whether the input information includes sensitive information;

[0012] The desensitization processing module is used to perform desensitization processing on the input information based on preset desensitization rules in response to the input information including sensitive information, so as to obtain desensitized input information.

[0013] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any possible implementation manner and the aspects described above.

[0017] According to yet another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method of the above-mentioned aspect and any possible implementation manner.

[0018] According to yet another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the computer program implements the above-mentioned aspects and any possible implementation method.

[0019] According to the technology disclosed in the present invention, the data input into the large model can be desensitized, which effectively improves the security of the data.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0022] Figure 1 is a schematic diagram according to a first embodiment of the present disclosure;

[0023] Figure 2 is a schematic diagram according to a second embodiment of the present disclosure;

[0024] Figure 3 is a schematic diagram according to a third embodiment of the present disclosure;

[0025] Figure 4 is a schematic diagram according to a fourth embodiment of the present disclosure;

[0026] Figure 5 The block diagram is a block diagram of an electronic device for implementing the method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0028] Obviously, the described embodiments are only part of the embodiments of the present disclosure, but not all of them. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present disclosure.

[0029] It should be noted that the terminal devices involved in the embodiments of the present disclosure may include but are not limited to mobile phones, personal digital assistants (PDAs), wireless handheld devices, tablet computers and other smart devices; display devices may include but are not limited to personal computers, televisions and other devices with display functions.

[0030] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0031] Figure 1 is a schematic diagram according to the first embodiment of the present disclosure; Figure 1 As shown, this embodiment provides an information processing method based on a large model, which may specifically include the following steps:

[0032] S101, obtaining input information of a pre-input large model;

[0033] S102, using a pre-trained sensitive information recognition model to identify whether the input information includes sensitive information;

[0034] S103: In response to the input information including sensitive information, desensitize the input information based on a preset desensitization rule to obtain desensitized input information.

[0035] The execution subject of the information processing method based on the big model of this embodiment is an information processing device based on the big model, which can be an electronic entity or a software integrated application. The information processing device based on the big model of this embodiment is set on the user side of the application of the big model, and performs sensitive information identification and desensitization processing on the local input information of the user side.

[0036] The input information of this embodiment refers to the information input into the big model when the big model is used, and may include information such as the user's query and prompt.

[0037] The large model of this embodiment can be called a large language model (LLM), which is a complex model that uses deep learning technology to train a large amount of text data to understand and generate natural language.

[0038] The large model of this embodiment can be considered as a public model, which can be set on the remote server side, and different users can share the large model for information processing. The sensitive information identification model can be considered as a private model set locally in the information processing device, which only performs sensitive information identification on local input information. Compared with the large model, the sensitive information identification model has a simpler structure and a relatively single function.

[0039] The sensitive information identification model of this embodiment can be implemented using a pre-trained neural network model, which can detect sensitive information in the input information to be input into the large model. Furthermore, if it is detected that the input information includes sensitive information, the input information is desensitized based on the preset desensitization rules to obtain desensitized input information for input into the large model, which can avoid sensitive information in the input information from being input into the large model, causing information leakage.

[0040] The information processing method based on the large model of this embodiment can effectively desensitize sensitive information in the input information by desensitizing the data before the data is input into the large model, thereby avoiding information leakage, effectively protecting the privacy of the user, and effectively improving the security of the information. Moreover, compared with the prior art, the technical solution of this embodiment does not require retraining of the large model, the implementation cost is very low, and the efficiency of privacy protection is very high.

[0041] Figure 2 is a schematic diagram according to the second embodiment of the present disclosure; the information processing method based on the large model of this embodiment, in the above Figure 1 Based on the technical solutions of the embodiments shown in the figure, the technical solutions of the present disclosure are further described in more detail. Figure 2 As shown, the information processing method based on the large model of this embodiment may specifically include the following steps:

[0042] S201, obtaining input information of a pre-input large model;

[0043] S202, using a pre-trained sensitive information recognition model to identify whether the input information includes sensitive information;

[0044] S203: In response to the input information including sensitive information, obtaining the type of sensitive information identified by the sensitive information identification model;

[0045] The sensitive information identification model of this embodiment can not only identify whether the input information includes sensitive information, but also identify the type of sensitive information when sensitive information is identified, so that different desensitization processes can be performed for different types of sensitive information in the future, which can effectively improve the effect of desensitization. For example, in this embodiment, the type of sensitive information includes numbers or subjects. The subject in this embodiment refers to a name-like word that can be used as a subject or object. In other words, the subject needs to meet two conditions: first, the subject appears as a name, which can be a person's name, or a company, group or organization name; second, the subject appears as a subject or object in the sentence. In practical applications, the type of sensitive information can also include other extensible types, which are not limited here.

[0046] For example, the input information includes the revenue of Company A of X yuan, the revenue of Company B of Y yuan, and the revenue of Company C of Z yuan. For confidentiality, Company A, Company B, and Company C are all sensitive information, specifically subject-type sensitive information. The revenue of X yuan, Y yuan, and Z yuan are also sensitive information, specifically digital-type sensitive information.

[0047] S204, detecting whether the type of sensitive information is a number or a subject, if it is a number, executing step S205; if the type of sensitive information is a subject, executing step S206;

[0048] S205, performing differential privacy processing or Gaussian noise processing on the digital sensitive information in the input information to obtain desensitized input information; executing step S209;

[0049] Optionally, step S205 may also be replaced by the following step S205'.

[0050] S205', performing linear transformation processing on the digital sensitive information in the input information to obtain desensitized input information.

[0051] Further optionally, in order to further enhance the security of the data, after step S205', differential privacy processing or Gaussian noise processing can be further performed on the digital type information after the linear transformation processing in the input information to obtain desensitized input information, that is, the processing method of step S205 is adopted and then processed to prevent the desensitized input information from being obtained by an attacker and then reversely inferring the linear transformation, and then obtaining all digital type information in the input information.

[0052] S206, obtaining a replacement subject that does not include sensitive information;

[0053] The replacement subject of this embodiment may be a subject collected in advance and not including sensitive information. Alternatively, some replacement subjects not including sensitive information may be generated by the large model.

[0054] S207, establishing a mapping relationship between sensitive information and replacement subject;

[0055] S208, based on the mapping relationship, desensitizing the sensitive information in the input information to obtain desensitized input information; executing step S209;

[0056] For example, the replacement entities of Company A, Company B, and Company C are obtained as Category 1, Category 2, and Category 3, respectively. Based on this, a mapping relationship of Company A→Category 1, a mapping relationship of Company B→Category 2, and a mapping relationship of Company C→Category 3 can be established. Based on the above mapping relationships, Company A, Company B, and / or Company C in the input information can be replaced with Category 1, Category 2, and / or Category 3, respectively.

[0057] By adopting the above steps S206-S208, the sensitive information of the subject type in the input information can be effectively desensitized, the risk of leakage of the sensitive information of the subject type can be reduced, and the security of the sensitive information of the subject type in the input information can be effectively improved.

[0058] S209, using a pre-trained large model to optimize the desensitized input information to obtain optimized input information;

[0059] S210, inputting optimized input information into the pre-trained large model, so that the large model can perform reasoning based on the optimized input information to obtain a reasoning result;

[0060] Since the sentences of the desensitized input information may not be smooth after desensitization, the direct reasoning results may not be ideal. In order to further improve the accuracy of the reasoning results, in this embodiment, the desensitized input information is input into the large model, and the large model is allowed to optimize the desensitized input information to obtain optimized input information with higher quality. Furthermore, using the large model to perform reasoning based on the optimized input information can further improve the accuracy of the reasoning results.

[0061] S211, obtaining the inference result output by the large model;

[0062] S212: Based on the preset desensitization rules, the desensitized information in the inference result is restored.

[0063] For example, the following methods may be included:

[0064] (1) If sensitive information of the subject type in the input information is desensitized based on the mapping relationship, the replacement subject in the mapping relationship included in the inference result is restored to the sensitive information in the mapping relationship based on the mapping relationship.

[0065] (2) If a linear transformation is performed on the digital sensitive information in the specified field of the input information to achieve desensitization, the digital information in the inference result is restored based on the inverse transformation of the linear transformation.

[0066] Further optionally, in one embodiment of the present disclosure, the inference results output by the large model may be recorded, and all inference results of a certain period of time may be reviewed periodically in the future to prevent certain sensitive subjects or data from not being identified, thereby providing an opportunity for attackers to exploit.

[0067] The information processing method based on the big model in this embodiment effectively protects user privacy by desensitizing the data before inputting it into the big model and restoring the data after the big model returns the inference result. This solution not only protects the privacy of the company, but also provides very high data accuracy. This solution is very suitable for scenarios where the company uses an external big model for data analysis.

[0068] The information processing method based on the big model in this embodiment can accurately and effectively desensitize digital sensitive information and subject sensitive information, respectively, thereby effectively protecting digital sensitive information and subject sensitive information in the input information. Moreover, in this embodiment, the big model can also be used to optimize the desensitized user information, further improving the quality of the input information input to the big model, thereby effectively improving the effect of the big model's reasoning based on the optimized desensitized user information.

[0069] The information processing method based on the large model of this embodiment mainly prevents data leakage when the company's internal data is used for model reasoning in the data analysis scenario. Compared with the data encryption solution, this solution has a low threshold for use, is easy to maintain, and is controllable in the future. Using the technical solution of this embodiment, the privacy and security of the company's internal access to external models or interfaces can be greatly protected.

[0070] Figure 3 is a schematic diagram according to the third embodiment of the present disclosure; this embodiment provides an information processing device 300 based on a large model, including:

[0071] The acquisition module 301 is used to acquire input information pre-input into the large model;

[0072] An identification module 302 is used to use a pre-trained sensitive information identification model to identify whether the input information includes sensitive information;

[0073] The desensitization processing module 303 is used to perform desensitization processing on the input information based on preset desensitization rules in response to the input information including sensitive information, so as to obtain desensitized input information.

[0074] The large model-based information processing device 300 of this embodiment realizes the implementation principle and technical effect of information processing by adopting the above-mentioned modules, which is the same as the implementation of the above-mentioned related method embodiments. For details, please refer to the records of the above-mentioned related method embodiments, which will not be repeated here.

[0075] Figure 4 is a schematic diagram according to the fourth embodiment of the present disclosure; the information processing device 400 based on the large model of this embodiment, in the above Figure 3 Based on the technical solutions of the embodiments shown in the figure, the technical solutions of the present disclosure are further described in more detail. Figure 4 As shown, the information processing device 400 based on the large model of this embodiment includes the above Figure 3 The modules with the same name and function are shown as follows: an acquisition module 401 , an identification module 402 and a desensitization processing module 403 .

[0076] like Figure 4 As shown, in this embodiment, the desensitization processing module 403 includes:

[0077] A type identification unit 4031 is used for obtaining the type of sensitive information identified by the sensitive information identification model in response to the input information including sensitive information;

[0078] The first desensitization processing unit 4032 is used to:

[0079] If the type of the sensitive information is digital, differential privacy processing or Gaussian noise processing is performed on the digital type sensitive information in the input information to obtain desensitized input information; or linear transformation processing is performed on the digital type sensitive information in the input information to obtain desensitized input information.

[0080] Further optionally, in one embodiment of the present disclosure, the first desensitization processing unit 4032 is further configured to:

[0081] Differential privacy processing or Gaussian noise processing is performed on the digital information after linear transformation in the input information to obtain desensitized input information.

[0082] Further optionally, if Figure 4 As shown, in one embodiment of the present disclosure, the desensitization processing module 403 further includes a second desensitization processing unit 4033, which is used to:

[0083] If the type of the sensitive information is a subject, obtaining a replacement subject that does not include the sensitive information;

[0084] Establishing a mapping relationship between the sensitive information and the replacement subject;

[0085] Based on the mapping relationship, the sensitive information in the input information is desensitized to obtain desensitized input information.

[0086] Further optionally, if Figure 4 As shown, in one embodiment of the present disclosure, the information processing device 400 further includes:

[0087] An input module 404 is used to input the desensitized input information into a pre-trained large model, so that the large model can perform reasoning based on the desensitized input information to obtain a reasoning result;

[0088] The acquisition module 401 is also used to obtain the inference result output by the large model;

[0089] The restoration module 405 is used to restore the desensitized information in the inference result based on a preset desensitization rule.

[0090] Further optionally, if Figure 4 As shown, in one embodiment of the present disclosure, the information processing device 400 further includes:

[0091] The optimization module 406 is used to optimize the desensitized input information using a pre-trained large model.

[0092] Further optionally, in one embodiment of the present disclosure, the restoration module 405 is used to:

[0093] If the sensitive information of the subject type in the input information is desensitized based on the mapping relationship, the replacement subject in the mapping relationship included in the inference result is restored to the sensitive information in the mapping relationship based on the mapping relationship.

[0094] Further optionally, in one embodiment of the present disclosure, the restoration module 405 is used to:

[0095] If a linear transformation is performed on the sensitive information of the numeric type in the specified field in the input information to achieve desensitization, the numeric type information in the inference result is restored based on the inverse transformation of the linear transformation.

[0096] The large model-based information processing device 400 of this embodiment realizes the implementation principle and technical effect of information processing by adopting the above-mentioned modules, which is the same as the implementation of the above-mentioned related method embodiments. For details, please refer to the records of the above-mentioned related method embodiments, which will not be repeated here.

[0097] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0098] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0099] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0100] like Figure 5 As shown, the device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0101] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0102] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as the above-mentioned methods of the present disclosure. For example, in some embodiments, the above-mentioned methods of the present disclosure may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the above-mentioned methods of the present disclosure described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the above-mentioned methods of the present disclosure in any other appropriate manner (e.g., by means of firmware).

[0103] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0104] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0105] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0106] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0107] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0108] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0109] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0110] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. An information processing method based on a large model, comprising: Get input information for pre-input large model; Using a pre-trained sensitive information recognition model to identify whether the input information includes sensitive information; In response to the input information including sensitive information, the input information is desensitized based on a preset desensitization rule to obtain desensitized input information.

2. The method according to claim 1, wherein: In response to the input information including sensitive information, desensitizing the input information based on a preset desensitization rule to obtain desensitized input information, including: In response to the input information including sensitive information, obtaining the type of sensitive information identified by the sensitive information identification model; If the type of the sensitive information is digital, perform differential privacy processing or Gaussian noise processing on the digital sensitive information in the input information to obtain desensitized input information; or Linear transformation is performed on the digital sensitive information in the input information to obtain desensitized input information.

3. The method according to claim 2, wherein: After performing linear transformation processing on the digital sensitive information in the input information and before obtaining the desensitized input information, the method further includes: Perform differential privacy processing or Gaussian noise processing on the digital information after linear transformation in the input information.

4. The method according to claim 2, wherein: In response to the input information including sensitive information, desensitizing the input information based on a preset desensitization rule to obtain desensitized input information, including: If the type of the sensitive information is a subject, obtaining a replacement subject that does not include the sensitive information; Establishing a mapping relationship between the sensitive information and the replacement subject; Based on the mapping relationship, the sensitive information in the input information is desensitized to obtain desensitized input information.

5. The method according to any one of claims 1 to 4, wherein: After performing desensitization processing on the input information based on a preset desensitization rule to obtain the desensitized input information, the method further includes: Inputting the desensitized input information into a pre-trained large model, so that the large model can perform reasoning based on the desensitized input information to obtain a reasoning result; Obtaining the inference result output by the large model; Based on the preset desensitization rules, the desensitized information in the inference result is restored.

6. The method according to claim 5, wherein: Based on a preset desensitization rule, after desensitizing the input information to obtain the desensitized input information, before inputting the desensitized input information into the pre-trained large model, the method further includes: The desensitized input information is optimized using a pre-trained large model.

7. The method according to claim 5, wherein: Based on the preset desensitization rules, the desensitized information in the inference result is restored, including: If the sensitive information of the subject type in the input information is desensitized based on the mapping relationship, the replacement subject in the mapping relationship included in the inference result is restored to the sensitive information in the mapping relationship based on the mapping relationship.

8. The method according to claim 5, wherein: Based on the preset desensitization rules, the desensitized information in the inference result is restored, including: If a linear transformation is performed on the sensitive information of the numeric type in the specified field in the input information to achieve desensitization, the numeric type information in the inference result is restored based on the inverse transformation of the linear transformation.

9. An information processing device based on a large model, comprising: An acquisition module is used to obtain input information pre-input into the large model; An identification module, used to use a pre-trained sensitive information identification model to identify whether the input information includes sensitive information; The desensitization processing module is used to perform desensitization processing on the input information based on preset desensitization rules in response to the input information including sensitive information, so as to obtain desensitized input information.

10. The device according to claim 9, wherein: The desensitization processing module comprises: a type identification unit, configured to obtain the type of sensitive information identified by the sensitive information identification model in response to the input information including sensitive information; The first desensitization processing unit is used for: If the type of the sensitive information is digital, differential privacy processing or Gaussian noise processing is performed on the digital type sensitive information in the input information to obtain desensitized input information; or linear transformation processing is performed on the digital type sensitive information in the input information to obtain desensitized input information.

11. The device according to claim 10, wherein: The first desensitization processing unit is further used for: Perform differential privacy processing or Gaussian noise processing on the digital information after linear transformation in the input information.

12. The device according to claim 10, wherein: The desensitization processing module further includes a second desensitization processing unit, which is used to: If the type of the sensitive information is a subject, obtaining a replacement subject that does not include the sensitive information; Establishing a mapping relationship between the sensitive information and the replacement subject; Based on the mapping relationship, the sensitive information in the input information is desensitized to obtain desensitized input information.

13. The device according to any one of claims 9 to 12, wherein: The device also includes: An input module, used for inputting the desensitized input information into a pre-trained large model, so that the large model can perform reasoning based on the desensitized input information to obtain a reasoning result; The acquisition module is further used to acquire the inference result output by the large model; The restoration module is used to restore the desensitized information in the inference result based on a preset desensitization rule.

14. The device according to claim 13, wherein: The device also includes: The optimization module is used to optimize the desensitized input information using a pre-trained large model.

15. The device according to claim 13, wherein: The restoration module is used to: If the sensitive information of the subject type in the input information is desensitized based on the mapping relationship, the replacement subject in the mapping relationship included in the inference result is restored to the sensitive information in the mapping relationship based on the mapping relationship.

16. The device according to claim 13, wherein: The restoration module is used to: If a linear transformation is performed on the sensitive information of the numeric type in the specified field in the input information to achieve desensitization, the numeric type information in the inference result is restored based on the inverse transformation of the linear transformation.

17. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 4.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-4.

19. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 4.

Images