Data protection method, device, apparatus and computer storage medium

Abstract

An embodiment of the present invention provides a data protection method, device, apparatus and computer storage medium, the method comprising: obtaining n privacy sub-models; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1; obtaining data to be processed, and determining the privacy sub-model corresponding to the data to be processed; when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, generating warning information to prompt that the data to be processed needs to be privacy protected.

Description

Technical Field

[0001] The embodiments of the present invention relate to, but are not limited to, privacy data protection technology, and in particular, to a data protection method, device, apparatus, and computer storage medium. Background Art

[0002] With the rapid development of mobile Internet, various applications on mobile terminals have become important tools for users to understand the world; due to the openness and interoperability of the Internet, users are increasingly concerned about their personal online privacy. Although user privacy is sensitive information, it is still exposed at any time. For example, users' various behavior trajectories such as search, browsing, downloading, payment, location, exercise volume, etc. are collected, stored, and analyzed by various websites, apps, and terminals, and then used for precision marketing or other commercial purposes, and even bring about information leakage, identity theft, malicious attacks and other hazards.

[0003] In the related art, the method for protecting user privacy data is not flexible enough and cannot determine whether user data needs to be protected according to actual needs. Summary of the invention

[0004] The embodiments of the present invention provide a data protection method, device, apparatus and computer storage medium, which can flexibly protect and manage privacy data.

[0005] To achieve the above object, the technical solution of the embodiment of the present invention is implemented as follows:

[0006] An embodiment of the present invention provides a data protection method, the method comprising:

[0007] Obtain n privacy sub-models; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1;

[0008] Obtain the data to be processed and determine the privacy sub-model corresponding to the data to be processed;

[0009] When the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, an early warning message is generated to prompt that the data to be processed needs to be privacy protected.

[0010] An embodiment of the present invention further provides a data protection device, the device comprising a processor and a memory for storing a computer program that can be run on the processor; wherein:

[0011] The processor is used to execute the steps of any one of the above-mentioned data protection methods when running the computer program.

[0012] The embodiment of the present invention further provides a data protection device, which includes: an acquisition module and a decision module, wherein:

[0013] An acquisition module, used to acquire n privacy sub-models; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1;

[0014] The decision module is used to obtain the data to be processed and determine the privacy sub-model corresponding to the data to be processed; when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, generate warning information to prompt that the data to be processed needs privacy protection.

[0015] An embodiment of the present invention further provides a computer storage medium, and when the computer program is executed by a processor, the steps of any one of the above-mentioned data protection methods are implemented.

[0016] In a data protection method, device, apparatus and computer storage medium provided by an embodiment of the present invention, first, n privacy sub-models are obtained; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1; then, the data to be processed is obtained, and the privacy sub-model corresponding to the data to be processed is determined; finally, when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, an early warning message is generated to prompt that the data to be processed needs to be privacy protected.

[0017] When adopting the technical solution recorded above, since the n privacy attributes corresponding to the n privacy sub-models can be flexibly set according to the user's actual needs, n privacy sub-models that meet the actual needs can be obtained. Furthermore, when it is determined to generate warning information based on the n privacy sub-models that meet the actual needs, it means that the generation of warning information meets the actual needs; that is, by flexibly and autonomously setting n privacy attributes in advance, early warning reminders for user privacy data can be achieved, which has a certain degree of flexibility and autonomy, and can prevent the leakage of privacy data that requires privacy protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of a data protection method according to an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of clustering results for training data in an embodiment of the present invention;

[0020] Figure 3 is a flow chart of another data protection method according to an embodiment of the present invention;

[0021] Figure 4 A schematic diagram of the structure of a data protection device according to an embodiment of the present invention;

[0022] Figure 5 The figure is a schematic diagram of the hardware structure of a data protection device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] In the related art, there are two main methods for protecting user privacy data: (1) Protecting applications. Specifically, creating a program disguise trigger to disguise the application icon and name, so that when the terminal lock screen is released, the application can still be guaranteed to have strong privacy for the user, preventing the application from being used by a second person other than the terminal user, thereby protecting the privacy security of user data in the application. (2) Protecting user data. Specifically, changing user privacy data to generate anonymized data hiding that protects personal privacy.

[0025] When method (1) is used to protect user privacy data, since only the identification information of the application is protected, the scope of protection is very small and the protection effect on user privacy data is not great; when method (2) is used to protect user privacy data, since the user data needs to be anonymized, it is necessary to establish an association relationship between the anonymous information and the real data. The mapping degree of the association relationship directly affects the use of the privacy data, that is, the protection scheme for privacy data is implemented by changing the user privacy data, which will affect the use of the privacy data.

[0026] It can be seen that in the related art, the protection method for user privacy data is not flexible enough and cannot determine whether user data needs to be protected according to actual needs.

[0027] The embodiments of the present invention can be applied to any scenario where privacy protection is required. For example, when privacy protection is performed on user data generated when an application is running on a terminal, it can be implemented based on the technical solution provided by the embodiments of the present invention. The embodiments of the present invention do not limit the types of applications running on the terminal.

[0028] The embodiments of the present invention may be applied to a terminal or other devices. The terminal or other devices mentioned above may include components such as a processor and a memory.

[0029] Based on the above-described contents, the following specific embodiments are proposed.

[0030] First embodiment

[0031] The first embodiment of the present invention describes a data protection method. Figure 1 is a flow chart of a data protection method according to an embodiment of the present invention, such as Figure 1 As shown, the process may include:

[0032] Step 101: Obtain n privacy sub-models; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1.

[0033] As for the implementation method of this step, illustratively, training data can be first obtained, where the training data is used to represent user data generated when the application is running; then, the training data is clustered with the preset n privacy attributes as the central object to obtain n privacy sub-models.

[0034] In actual implementation, the user original data generated when the application is running can be obtained, and the user original data recorded above can be preprocessed to obtain training data; illustratively, at least one of the following can be performed on the user original data recorded above to obtain training data: word segmentation processing, filtering useless words processing, useless words can include punctuation, single words, symbols, and other meaningless words; it should be noted that the content recorded above only provides an implementation method of preprocessing in an illustrative manner, and preprocessing can also have other implementation methods, which are not limited by the embodiments of the present invention.

[0035] Regarding the implementation method of the user original data recorded above, exemplarily, the user original data recorded above may be the user data generated when the application (Application, App) of the mobile terminal is running, and may include various data generated when the user uses various applications of the mobile terminal, such as login information, reading, consumption, preference details, etc.

[0036] In practical applications, n different privacy attributes can be pre-set according to the actual needs of protecting privacy data. Each of the n privacy attributes represents a privacy point that the user determines to be protected (i.e., the privacy point that the user cares about most). For example, the n privacy attributes can include "identity", "interest", etc.; n can be considered as a pre-set protection coefficient. The larger the value of n, the more privacy points the user determines to be protected. Furthermore, after setting the n privacy attributes, the user can change the n privacy attributes according to actual needs, and then re-cluster the training data based on the revised privacy attributes to obtain the corresponding privacy sub-model.

[0037] It can be seen that by setting the protection coefficient n, users can flexibly determine the privacy protection strategy of their personal data, and the protection coefficient n directly affects the size of the privacy protection scope. Users set the corresponding protection coefficient according to the degree of privacy protection they want.

[0038] During specific implementation, the user may input the protection coefficient n and n privacy attributes through the user interface (UI) of the terminal, which is convenient for user operation.

[0039] In an embodiment of the present invention, after obtaining n privacy attributes and user original data, the n privacy attributes and user original data can be used as input data for constructing n privacy sub-models. Then, the input data of the n privacy sub-models can be processed to obtain n privacy sub-models.

[0040] For the implementation method of obtaining n privacy sub-models, a clustering-based natural language processing method commonly used in machine learning can be used to automatically cluster the input data of the n privacy sub-models, and through an iterative method, the central object of each clustering is updated successively until the final clustering result is obtained; here, the final clustering result may include n clusters, the privacy attributes of the n clusters in the final clustering result are different from each other, and each cluster in the final clustering result represents a privacy sub-model. It should be noted that the embodiments of the present invention do not limit the machine learning model structure and learning method.

[0041] That is to say, after obtaining the training data and the preset n privacy attributes, the preset n privacy attributes are used as the central object, and an iterative clustering method is adopted to perform multiple clustering processes on the training data to obtain n privacy sub-models; wherein, when performing non-first clustering processing, the central object of the clustering is updated so that the preset evaluation index of the current clustering result is higher than the preset evaluation index of the previous clustering result.

[0042] Here, the preset evaluation index of the clustering result can be used to represent: the proximity of each record in the same cluster in the clustering result, and the distance between records in different clusters in the clustering result; the closer the records in the same cluster in the clustering result are, and the farther the records in different clusters are, the higher the preset evaluation index of the clustering result is.

[0043] Specifically, performing a first clustering process on the training data to obtain a first clustering result;

[0044] Let m represent the total number of iterations of the iterative clustering method. When i is 2 to m, on the basis of the i-1th clustering result, with the preset evaluation index of the clustering result of the i-1th clustering being higher than the preset evaluation index of the i-1th clustering result as the goal, the central object of the i-1th clustering is updated to obtain the central object of the i-th cluster; according to the central object of the i-th cluster, the training data is clustered for the i-th time to obtain the i-th clustering result.

[0045] Here, m can be a preset integer greater than 1, or can be determined by a preset iteration termination condition; in one example, the preset iteration termination condition can be: it is impossible to update the central object of the previous clustering with the goal of achieving a higher preset evaluation index of the clustering result.

[0046] Below through Figure 2 The clustering results of the training data in the embodiment of the present invention are described. For ease of understanding, refer to Figure 2 , when i is 1 to m, when the training data is clustered for the i-th time, the n central objects of the i-th clustering are abstracted into n points. With these n points as the core, the objects in the data set that are semantically closest to them are attracted and aggregated to form clusters centered on these n attributes. Figure 2 In the figure, four circles K1, K2, K3 and Kn are used to identify four clusters. Each cluster represents various attributes related to a privacy attribute. For example, the cluster centered on "identity" gathers various attributes related to "identity", such as "name", "..." If the user wants a wider range of privacy protection, it can be achieved by adjusting the protection coefficient: the larger the protection coefficient, the more privacy attributes need to be set.

[0047] Step 102: Obtain the data to be processed and determine the privacy sub-model corresponding to the data to be processed; when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, determine that the data to be processed needs to be privacy protected.

[0048] In actual applications, user data generated when the application is running can be monitored, and when the user data generated when the application is running is monitored, the monitored user data generated when the application is running is used as the data to be processed. Exemplarily, the data to be processed is data to be uploaded by the terminal or data to be saved by the terminal.

[0049] After obtaining the data to be processed, the n privacy sub-models recorded above are taken as the central objects, and sufficient calculations are performed through the machine learning algorithm to determine the privacy sub-model to which the data to be processed belongs (the privacy sub-model corresponding to the data to be processed). Exemplarily, the correlation between the data to be processed and the n privacy sub-models can be determined respectively, and the privacy sub-model with the greatest correlation with the data to be processed is used as the privacy sub-model corresponding to the data to be processed.

[0050] Here, for the implementation method of determining the correlation between the data to be processed and each privacy sub-model, in one example, the semantic distance between the data to be processed and each privacy sub-model can be calculated, and the correlation between the data to be processed and each privacy sub-model can be determined based on the semantic distance between the data to be processed and each privacy sub-model; the smaller the semantic distance between the data to be processed and the privacy sub-model, the greater the privacy sensitivity of the data to be processed, and the greater the correlation between the data to be processed and the privacy sub-model.

[0051] In specific implementation, after determining the privacy sub-model corresponding to the data to be processed, the relationship between the correlation between the data to be processed and the corresponding privacy sub-model and the preset correlation threshold can be determined. When the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to the preset correlation threshold, it is determined that the data to be processed needs to be privacy protected. At this time, an early warning message can be generated to prompt that the data to be processed needs to be privacy protected; when the correlation between the data to be processed and the corresponding privacy sub-model is less than the preset correlation threshold, it is determined that the data to be processed does not need to be privacy protected, and the process can be directly terminated.

[0052] In an embodiment of the present invention, for an implementation method of determining the relationship between the correlation between the data to be processed and the corresponding privacy sub-model and a preset correlation threshold, illustratively, the semantic distance between the data to be processed and the corresponding privacy sub-model can be calculated. When the semantic distance between the data to be processed and the corresponding privacy sub-model is less than or equal to the preset semantic distance threshold, it is determined that the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to the preset correlation threshold. Conversely, when the semantic distance between the data to be processed and the corresponding privacy sub-model is greater than the preset semantic distance threshold, it is determined that the correlation between the data to be processed and the corresponding privacy sub-model is less than the preset correlation threshold.

[0053] Since the n privacy attributes corresponding to the n privacy sub-models can be flexibly set according to actual needs, n privacy sub-models that meet actual needs can be obtained. Furthermore, when the generation of warning information is determined based on the n privacy sub-models that meet actual needs, it means that the generation of warning information meets actual needs; that is, by flexibly and autonomously setting n privacy attributes in advance, early warning reminders for user privacy data can be achieved, which has certain flexibility and autonomy, and can prevent the leakage of privacy data that needs privacy protection. In addition, when the correlation between the data to be processed and the corresponding privacy sub-model is less than the preset correlation threshold, the data to be processed can be ignored, so that a safe channel and strong protection can be provided for data that does not need privacy protection.

[0054] In practical applications, the data protection method of the first embodiment of the present invention can be implemented based on a processor of a terminal, etc.

[0055] In the related art, there are two main methods for protecting user privacy data: (1) Protecting applications. Specifically, creating a program disguise trigger to disguise the application icon and name, so that when the terminal lock screen is released, the application can still be guaranteed to have strong privacy for the user, preventing the application from being used by a second person other than the terminal user, thereby protecting the privacy security of user data in the application. (2) Protecting user data. Specifically, changing user privacy data to generate anonymized data hiding that protects personal privacy.

[0056] When method (1) is used to protect user privacy data, since only the identification information of the application is protected, the scope of protection is very small and the protection effect on user privacy data is not great; when method (2) is used to protect user privacy data, since the user data needs to be anonymized, it is necessary to establish an association relationship between the anonymous information and the real data. The mapping degree of the association relationship directly affects the use of the privacy data, that is, the protection scheme for privacy data is implemented by changing the user privacy data, which will affect the use of the privacy data.

[0057] In the related technologies, when users install an App on a mobile terminal, they must agree to all authorizations declared by the App, otherwise the application cannot be installed and used. Research in related fields also mainly focuses on malware detection, cloud, server assistance, and privacy protection methods based on information flow control. In the App ecosystem, there is no agreement or policy for users to decide for themselves which information to make public and which information to hide.

[0058] It can be seen that in the related art, the protection method for user privacy data is not flexible enough and cannot determine whether user data needs to be protected according to actual needs.

[0059] In the embodiment of the present invention, the original user data on the terminal is automatically refined and aggregated according to the preset n privacy attributes by using the machine learning method, and a privacy protection scheme that meets the actual needs of a single user is generated. On this basis, it is detected whether the data generated during the use of the application meets the user's expected degree of privacy information openness, and then corresponding measures are taken. It can be seen that by setting n privacy attributes, the data generated during the use of the application can be filtered and screened, which can protect the privacy data from abuse or even attack, and make effective use of the data. That is to say, in the embodiment of the present invention, from the user's perspective, adhering to the purpose of "letting users decide for their own data", a privacy data protection scheme that meets the user's needs can be automatically constructed by using the machine learning method, and then decisions and management of those privacy data that may be of concern to the user can be made, so that the user can protect his or her privacy while providing information to enjoy the service.

[0060] Second embodiment

[0061] Based on the data protection method proposed in the aforementioned embodiment of the present invention, further examples are given.

[0062] Figure 3 FIG. 1 is a flow chart of another data protection method according to an embodiment of the present invention. Figure 3 As shown, the process may include:

[0063] Step 301: Obtain data to be processed and n privacy sub-models.

[0064] The implementation method of this step has been explained in the first embodiment and will not be repeated here.

[0065] Step 302: Determine the privacy sub-model corresponding to the data to be processed.

[0066] The implementation method of this step has been explained in step 102 and will not be repeated here.

[0067] Step 303: Determine whether the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold. If yes, execute step 304; if not, end the process.

[0068] Step 304: Perform early warning or other processing on the data to be processed.

[0069] Specifically, when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to the preset correlation threshold, it can be considered that the data to be processed is likely to belong to the privacy category that the user cares about. Warning information can also be generated to indicate the risk of privacy information leakage, or the data to be processed can be directly protected for privacy. As for the display method of the warning information, for example, the warning information can be displayed in the UI or other forms of the terminal; as for the implementation method of privacy protection of the data to be processed, for example, when it is determined that the data to be processed is saved, uploaded, or other operations that may cause privacy leakage are performed, the corresponding operations on the data to be processed are prevented, and the user is warned or reminded.

[0070] Furthermore, when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, the data to be processed may be added to the corresponding privacy sub-model to make the privacy sub-model more complete.

[0071] Third embodiment

[0072] Based on the data protection method proposed in the above-mentioned embodiment, a fourth embodiment of the present invention provides a data protection device.

[0073] Figure 4 FIG. 1 is a schematic diagram of a data protection device according to an embodiment of the present invention. Figure 4 As shown, the device includes an acquisition module 401 and a decision module 402, wherein:

[0074] An acquisition module 401 is used to acquire n privacy sub-models; wherein each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are different from each other, and n is an integer greater than 1;

[0075] The decision module 402 is used to obtain the data to be processed and determine the privacy sub-model corresponding to the data to be processed; when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, generate warning information to prompt that the data to be processed needs to be privacy protected.

[0076] In one embodiment, the acquisition module 401 is specifically used to acquire training data, where the training data is used to represent user data generated when the application is running; the training data is clustered with the preset n privacy attributes as the central object to obtain n privacy sub-models.

[0077] In one embodiment, the acquisition module 401 is specifically used to use the preset n privacy attributes as the central object, adopt an iterative clustering method, perform multiple clustering processes on the training data, and obtain n privacy sub-models; wherein, when performing non-first clustering processing, the central object of the clustering is updated so that the preset evaluation index of the current clustering result is higher than the preset evaluation index of the previous clustering result.

[0078] In one embodiment, the preset evaluation index of the clustering result can be used to represent: the proximity of each record in the same cluster in the clustering result, and the distance between records in different clusters in the clustering result.

[0079] In one implementation, the data to be processed is data to be uploaded by the terminal or data to be saved by the terminal.

[0080] In one implementation, the decision module 402 is specifically configured to use, among the n privacy sub-models, a privacy sub-model with the greatest correlation with the data to be processed as the privacy sub-model corresponding to the data to be processed.

[0081] In one implementation, the decision module 402 is further configured to perform privacy protection on the data to be processed when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold.

[0082] In one embodiment, the decision module 402 is further used to add the data to be processed to the corresponding privacy sub-model when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, so as to achieve supplementary expansion of the corresponding privacy sub-model.

[0083] In one implementation, the decision module 402 is further configured to determine that the data to be processed does not need to be privacy protected when the correlation between the data to be processed and the corresponding privacy sub-model is less than a preset correlation threshold.

[0084] In practical applications, the acquisition module 401 and the decision module 402 can be implemented by a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP), or a field programmable gate array (FPGA) located in the terminal.

[0085] In addition, each functional module in this embodiment can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or software functional modules.

[0086] If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform all or part of the steps of the method described in this embodiment. The aforementioned storage medium includes: U disk, mobile hard disk, read only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0087] Specifically, a computer program instruction corresponding to a data protection method in this embodiment can be stored on a storage medium such as a CD, a hard disk, or a USB flash drive. When the computer program instruction corresponding to a data protection method in the storage medium is read or executed by an electronic device, the steps of any data protection method in the aforementioned embodiments are implemented.

[0088] Based on the same technical concept as the above embodiments, see Figure 5 , which shows a data protection device 50 provided by an embodiment of the present invention, the device may include: a memory 51, a processor 52 and a bus 53; wherein,

[0089] The bus 53 is used to connect the memory 51, the processor 52 and the mutual communication between these devices;

[0090] The memory 51 is used to store computer programs and data;

[0091] The processor 52 is used to execute the computer program stored in the memory to implement the steps of any one of the data protection methods in the aforementioned embodiments.

[0092] In practical applications, the memory 51 may be a volatile memory, such as RAM; or a non-volatile memory, such as ROM, flash memory, hard disk drive (HDD) or solid-state drive (SSD); or a combination of the above types of memory, and provide instructions and data to the processor 52.

[0093] The processor 52 may be at least one of an application specific integrated circuit (ASIC), a DSP, a digital signal processing device (DSPD), a programmable logic device (PLD), an FPGA, a CPU, a controller, a microcontroller, and a microprocessor. It is understandable that for different devices, the electronic device used to implement the function of the processor may also be other, and the embodiment of the present invention does not specifically limit it.

[0094] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0095] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention.

Claims

1. A data protection method, characterized in that: The method comprises: Obtain n privacy sub-models; wherein the n privacy sub-models are obtained by clustering training data with n preset privacy attributes as central objects, the training data is application data collected by the server according to user operations when the application is running, each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are privacy points determined by the user to be protected, and are different from each other, n is an integer greater than 1, and the n privacy attributes are modified according to user input when the application is running; Obtaining data to be processed, and determining multiple privacy sub-models corresponding to the data to be processed; When the correlation between the data to be processed and the corresponding partial privacy sub-model is greater than or equal to a preset correlation threshold, a warning message is generated to prompt that the data to be processed corresponding to the partial privacy sub-model needs to be privacy protected; when the correlation between the data to be processed and the corresponding remaining partial privacy sub-model is less than the preset correlation threshold, it is determined that the data to be processed corresponding to the remaining partial privacy sub-model does not need to be privacy protected; Provide privacy protection for the data to be processed that requires privacy protection.

2. The method according to claim 1, characterized in that The clustering process of the training data with the preset n privacy attributes as the central object includes: Taking the preset n privacy attributes as the central object, an iterative clustering method is used to perform multiple clustering processes on the training data to obtain n privacy sub-models; wherein, when performing non-first clustering processing, the central object of the clustering is updated so that the preset evaluation index of the current clustering result is higher than the preset evaluation index of the previous clustering result.

3. The method according to claim 2, characterized in that The preset evaluation index of the clustering result is used to represent: the proximity degree of each record in the same cluster in the clustering result, and the distance degree between records in different clusters in the clustering result.

4. The method according to claim 1, characterized in that: The data to be processed is data to be uploaded by the terminal or data to be saved by the terminal.

5. The method according to claim 1, characterized in that The determining of multiple privacy sub-models corresponding to the data to be processed includes: A plurality of privacy sub-models corresponding to the data to be processed are determined according to the correlation between the data to be processed and the n privacy sub-models.

6. The method according to claim 1, characterized in that The method further includes: when the correlation between the data to be processed and the corresponding privacy sub-model is greater than or equal to a preset correlation threshold, adding the data to be processed to the corresponding privacy sub-model to achieve supplementary expansion of the corresponding privacy sub-model.

7. A data protection device, characterized in that: The device comprises a processor and a memory for storing a computer program that can be run on the processor; wherein, When the processor is used to run the computer program, the steps of the method according to any one of claims 1 to 6 are performed.

8. A data protection device, characterized in that: The device comprises: an acquisition module and a decision module, wherein: An acquisition module is used to acquire n privacy sub-models; wherein the n privacy sub-models are obtained by clustering training data with n preset privacy attributes as central objects, the training data is application data collected by the server according to user operations when the application is running, each privacy sub-model is a data set representing a privacy attribute, the privacy attributes represented by the n privacy sub-models are privacy points determined by the user to be protected, and are different from each other, n is an integer greater than 1, and the n privacy attributes are modified according to user input when the application is running; The decision module is used to obtain the data to be processed and determine the multiple privacy sub-models corresponding to the data to be processed; when the correlation between the data to be processed and the corresponding part of the privacy sub-models is greater than or equal to a preset correlation threshold, generate warning information to prompt that the data to be processed needs to be privacy protected; when the correlation between the data to be processed and the corresponding remaining part of the privacy sub-models is less than the preset correlation threshold, determine that the data to be processed corresponding to the remaining part of the privacy sub-models does not need to be privacy protected; perform privacy protection on the data to be processed that is prompted to need privacy protection.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images