CN120470628B - Privacy protection method, device, equipment and storage medium - Google Patents
Privacy protection method, device, equipment and storage mediumInfo
- Publication number
- CN120470628B CN120470628B CN202510969432.1A CN202510969432A CN120470628B CN 120470628 B CN120470628 B CN 120470628B CN 202510969432 A CN202510969432 A CN 202510969432A CN 120470628 B CN120470628 B CN 120470628B
- Authority
- CN
- China
- Prior art keywords
- privacy
- page
- user
- strategy
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Storage Device Security (AREA)
Abstract
本申请公开了一种隐私保护方法、装置、设备及存储介质，涉及隐私保护技术领域，包括：获取用户在前端交互页面上的操作行为数据构建用户操作特征向量；执行隐私风险评估操作，得到前端交互页面的多个页面区域的隐私风险等级；从隐私保护策略集中选取与隐私风险等级匹配的隐私保护策略，生成多个策略任务；预测多个页面区域中除用户当前交互页面区域外生育页面区域的后续交互概率并结合对应的隐私风险等级，对多个策略任务进行排序，生成策略任务执行序列并依次执行。本申请能够融合用户行为建模进行页面区域级的隐私风险评估，实现风险的自动化动态识别和策略的可信执行，解决现有技术中动态防护不足、多维隐私覆盖欠缺的问题。
The present application discloses a privacy protection method, apparatus, device and storage medium, which relates to the field of privacy protection technology, including: obtaining the user's operation behavior data on the front-end interactive page to construct a user operation feature vector; performing a privacy risk assessment operation to obtain the privacy risk level of multiple page areas of the front-end interactive page; selecting a privacy protection policy that matches the privacy risk level from the privacy protection policy set to generate multiple policy tasks; predicting the subsequent interaction probability of the fertility page area other than the user's current interactive page area in multiple page areas and combining the corresponding privacy risk levels to sort the multiple policy tasks, generate a policy task execution sequence and execute them in sequence. The present application can integrate user behavior modeling to perform page area-level privacy risk assessment, realize automatic dynamic identification of risks and trusted execution of policies, and solve the problems of insufficient dynamic protection and lack of multi-dimensional privacy coverage in the existing technology.
Description
Technical Field
The present application relates to the field of privacy protection technologies, and in particular, to a privacy protection method, device, apparatus, and storage medium.
Background
With the development of the internet service and the continuous enhancement of the user data dependency, the operation behavior data (such as click event, page residence time, rolling behavior, input content and the like) of the user in the front-end interaction page are increasingly important resources for supporting personalized recommendation, behavior prediction and service optimization, however, the operation behaviors often carry privacy preference, intention and even sensitive information of the user, and especially in page areas such as identity authentication, financial information, medical records and the like, the user can expose privacy risks without being aware.
In the prior art, basic protection means such as image blurring, third party Cookie shielding, permission popup window prompting and the like are provided, but the technical problems still need to be solved, namely, firstly, protection logic mainly depends on static configuration rules and is difficult to adapt to real-time changes of user behaviors, so that privacy protection is insufficient in a dynamic scene, secondly, a user needs to actively configure a privacy strategy or conduct manual authorization, operation burden is increased, use experience is influenced, and thirdly, the prior art means are mostly single-point protection and are difficult to cover multi-dimensional privacy risks such as page area identification, operation track analysis, behavior prediction and the like.
Disclosure of Invention
The application provides a privacy protection method which can integrate user behavior modeling to carry out privacy risk assessment at page area level, so as to accurately carry out task scheduling and execution of privacy protection strategies, realize automatic dynamic identification of risks and trusted execution of strategies, and at least solve the problems of insufficient dynamic protection, strong configuration dependence, insufficient multidimensional privacy coverage, non-traceability of records and the like in the prior art.
The application provides a privacy protection method, which comprises the following steps:
Acquiring operation behavior data of a user on a front-end interaction page, and constructing a user operation feature vector based on the operation behavior data;
according to the user operation feature vector, privacy risk assessment operation is carried out, and privacy risk levels of a plurality of page areas of the front-end interaction page are obtained;
based on privacy risk levels of a plurality of page areas, respectively selecting privacy protection strategies matched with the privacy risk levels from a pre-stored privacy protection strategy set to generate a plurality of strategy tasks;
executing corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probability of the rest page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk levels to generate a strategy task execution sequence;
Based on the execution sequence, the execution parameters of a plurality of strategy tasks are pre-cached and sequentially executed.
The application also provides a privacy protection device, which comprises:
the operation recording module is used for acquiring operation behavior data of a user on the front-end interaction page and constructing a user operation feature vector based on the operation behavior data;
the risk assessment module is used for executing privacy risk assessment operation according to the user operation feature vector to obtain privacy risk levels of a plurality of page areas of the front-end interaction page;
The policy task generation module is used for respectively selecting privacy protection policies matched with the privacy risk levels from a pre-stored privacy protection policy set based on the privacy risk levels of the plurality of page areas to generate a plurality of policy tasks;
The task scheduling module is used for executing corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probability of the rest page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk levels to generate a strategy task execution sequence;
and the task execution module is used for pre-caching the execution parameters of the plurality of strategy tasks based on the execution sequence and sequentially executing the strategy tasks.
The application also provides electronic equipment, which comprises a memory and a processor, wherein the memory is used for storing a computer program, and the processor is used for realizing any privacy protection method when executing the computer program.
The present application also provides a computer readable storage medium having a computer program stored therein, wherein the computer program when executed by a processor implements the steps of any of the privacy preserving methods described above.
According to the method and the device, the operation behavior data of the user in the front-end page are collected, the user operation characteristic vector is constructed, multidimensional modeling of the user interaction intention is achieved, therefore, the depicting capability of page behavior risks is improved, the limitation that the user behavior cannot be perceived dynamically in the prior art is effectively broken through, meanwhile, privacy risk levels of different page areas can be accurately identified due to the fact that risk assessment operation is executed, the existing problem that static rules are relied on and real-time judgment is lacking is solved, furthermore, due to the fact that the current interaction area of the user and the user historical behavior data are combined, probability of a subsequent interaction page area is predicted, strategy tasks can be ordered, prospective and priority control of strategy response is achieved, protection efficiency under a dynamic interaction scene is improved, pre-caching management is conducted on strategy tasks based on ordering results, high-priority strategy tasks can be loaded in advance, the fact that the strategy does not need manual configuration of the user is guaranteed, response delay is remarkably reduced, and the problem of protection triggering delay in the prior art is solved.
Drawings
For a clearer description of embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described, it being apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for those skilled in the art.
Fig. 1 is a schematic view of an application environment of a privacy preserving architecture according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a privacy preserving engine according to an embodiment of the present application;
FIG. 3 is a schematic diagram of task scheduler operation according to an embodiment of the present application;
FIG. 4 is a schematic diagram of user interaction according to an embodiment of the present application;
FIG. 5 is a privacy preserving step diagram provided by an embodiment of the present application;
FIG. 6 is a flowchart of a privacy preserving method according to an embodiment of the present application;
fig. 7 is a block diagram of a privacy preserving apparatus according to an embodiment of the present application;
Fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without making any inventive effort are within the scope of the present application.
It should be noted that in the description of the present application, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. The terms "first," "second," and the like in this specification are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order.
The present application will be further described in detail below with reference to the drawings and detailed description for the purpose of enabling those skilled in the art to better understand the aspects of the present application.
The privacy protection method provided by the application can be applied to an application environment shown in figure 1. The client 102 communicates with the server 104 via a network, and a privacy protection architecture is deployed in the server 104, where the architecture includes an event monitor, a privacy protection engine, a resource monitor, a task scheduler, and a task recorder to perform privacy protection on a front-end interactive interface currently operated by a user on the client 102. The client 102 may be, but not limited to, various personal computers, notebook computers, smartphones, tablet computers, and portable wearable devices, and the server 104 may be implemented as a stand-alone server or a server cluster composed of multiple servers.
It should be noted that, as shown in fig. 2, the event monitor adopts user operation data on the client and sends the user operation data to the privacy protection engine, the privacy protection engine receives the user operation data sent by the event monitor, performs intelligent analysis and risk assessment on the user operation data, determines whether the privacy risk level of the front-end interaction page area currently operated by the user is medium risk or high risk, matches the corresponding privacy protection policy according to the privacy risk level if yes, and maintains the current privacy protection policy if not.
It should be noted that, as shown in fig. 3, the task scheduler receives a policy task, divides the policy task into a plurality of subtasks according to required resources, and allocates the subtasks to the CPU, the GPU and the cloud, so as to implement distributed task execution, and the result of executing the subtasks is summarized and returned to the client.
It should be noted that, as shown in fig. 4, a user operates a front-end interaction page on a client, an event monitor in a server monitors user operation in real time and sends data to a privacy protection engine, the privacy protection engine performs policy matching and task generation, and sends the generated policy task to a task scheduler, the task scheduler performs sorting, scheduling and execution of the policy task by receiving a system state sent by a resource monitor and the policy task sent by the privacy protection engine, and issues an execution result to the client, and the client loads a corresponding privacy protection page to realize privacy protection of the front-end interaction page.
As shown in fig. 5, an embodiment of the present application provides a privacy protection method, including:
Step 201, acquiring operation behavior data of a user on a front-end interaction page, and constructing a user operation feature vector based on the operation behavior data;
step 202, performing privacy risk assessment operation according to a user operation feature vector to obtain privacy risk levels of a plurality of page areas of a front-end interaction page;
Step 203, based on the privacy risk levels of the multiple page areas, respectively selecting privacy protection policies matching the privacy risk levels from a pre-stored privacy protection policy set to generate multiple policy tasks;
step 204, executing corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probabilities of the rest page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk levels to generate a strategy task execution sequence;
step 205, pre-caching the execution parameters of a plurality of strategy tasks based on the execution sequence and executing the strategy tasks in sequence.
The method and the device have the advantages that the operation behavior data of the user in the front-end page are collected, the user operation feature vector is constructed, multidimensional modeling of the user interaction intention is achieved, therefore, the depicting capability of page behavior risks is improved, the limitation that the user behavior cannot be dynamically perceived in the prior art is effectively overcome, meanwhile, privacy risk levels of different page areas can be accurately identified due to the fact that risk assessment operation is executed, the existing problem that static rules are relied on and real-time judgment is lacking is solved, moreover, due to the fact that the current interaction area of the user and the user historical behavior data are combined, probability of a subsequent interaction page area is predicted, strategy tasks can be ordered, look-ahead and priority control of strategy responses is achieved, protection efficiency in a dynamic interaction scene is improved, pre-caching management is conducted on strategy tasks based on ordering results, high-priority strategy tasks can be loaded in advance, the strategy is guaranteed to be ready in advance before the user is about to interact with the sensitive area, and the problem of protection trigger lag in the prior art is solved.
In one embodiment, as shown in fig. 6, the privacy protection method is realized based on the following logic that firstly, operation behavior data are acquired to obtain operation data of a user on a front-end interaction page, then, the operation behavior data are analyzed to realize privacy risk assessment of a page area of the front-end interaction page, furthermore, visual feedback is generated to the front-end interaction page according to a privacy risk assessment result to remind a user of the current risk state, meanwhile, matching of privacy protection strategies and generation of corresponding strategy tasks are carried out to conduct targeted privacy protection on the page area, afterwards, the generated strategy tasks are reordered to enable execution of the strategy tasks to be more suitable for the user, finally, the strategy tasks are executed to achieve privacy protection, when the strategy tasks are executed, pre-caching is carried out on the strategy tasks to be executed to improve deployment efficiency, the executed strategy tasks are generated to complete records and written into a blockchain to serve as verifiable certificates, and operation of the user is monitored to update the privacy risk level and the corresponding strategy tasks at any time.
In one embodiment, acquiring operation behavior data of a user on a front-end interaction page and constructing a user operation feature vector based on the operation behavior data comprises:
invoking a performance interface provided by a front-end interaction page, collecting operation records of a user in a plurality of page areas, and obtaining operation behavior data, wherein the operation behavior data at least comprises a screen coordinate value sequence of a clicking event, entering time and leaving time of the page areas, rolling speed and rolling acceleration of a page rolling event and an input character sequence of an input content change event;
obtaining a standard deviation of the distribution of the clicking positions according to the screen coordinate value sequence of the clicking events;
obtaining the average residence time of the page area according to the entering time and the leaving time of the page area;
obtaining a rolling acceleration integral value according to the rolling speed and the rolling acceleration of the page rolling event;
calculating a weight value of an input text sensitive word according to an input character sequence of the input content change event;
calling a sensor of equipment where the front-end interaction page is located, acquiring a screen inclination angle and ambient light intensity, and generating a sensing auxiliary characteristic value;
And splicing the click position distribution standard deviation, the average residence time of the page area, the rolling acceleration integral value, the weight value of the input text sensitive word and the sensing auxiliary characteristic value to obtain the user operation characteristic vector.
Specifically, in the embodiment, by defining the click position distribution standard deviation, the page average residence time, the rolling acceleration integral value, the input sensitive word weight, the sensing auxiliary characteristic value and other dimensional behavior indexes and combining and constructing the user operation characteristic vector, the fine granularity is facilitated to describe the behavior characteristics of the user in different page areas, compared with the traditional mode of only collecting simple interaction data, more comprehensive and more accurate user behavior modeling is realized, a stable data base is provided for subsequent privacy risk dynamic evaluation, and the adaptability and the precision of a protection strategy are enhanced.
In a preferred embodiment, the standard deviation of the distribution of the click positions is obtained according to the sequence of screen coordinate values of the click events, and is calculated based on the following formula:
;
Wherein, the Represents the standard deviation of the distribution of the clicking positions operated by the user, is used for measuring the concentration degree of the clicking behaviors of the user, n represents the total number of clicking events, x i represents the screen abscissa of the ith clicking event, y i represents the screen ordinate of the ith clicking event,The screen average abscissa representing all click events,The screen average ordinate representing all click events.
In a preferred embodiment, the rolling acceleration integral value is derived from the rolling speed and the rolling acceleration of the page rolling event, based on the following formula:
;
Wherein S a represents a rolling acceleration integral value of a user operation, which is used to reflect continuity and jerk of page browsing, a k is a kth rolling acceleration, and Δt is a rolling event sampling interval.
In a preferred embodiment, calculating a weight value for an input text-sensitive word from a sequence of input characters of an input content alteration event, comprises:
matching a predefined sensitive dictionary according to the input character sequence, wherein the content of the sensitive dictionary record comprises, but is not limited to, an identity card, a mobile phone number and an address;
Each sensitive word w i in the dictionary is given a weight If the input contains a plurality of sensitive words, the overall sensitive weight S sw is obtained based on the following formula:
;
Wherein S sw represents the overall weight value of the sensitive words of the input text operated by the user, so as to reflect the sensitivity of the text input by the user, and M represents the set of sensitive words identified in the input text.
In one embodiment, according to a user operation feature vector, performing a privacy risk assessment operation to obtain privacy risk levels of a plurality of page areas, including:
Dividing the user operation feature vector according to the page area to obtain a plurality of behavior segment vectors;
Constructing a privacy risk assessment model, wherein the privacy risk assessment model comprises a time sequence feature extraction sub-model constructed based on a long-term and short-term memory network and a reinforcement scoring sub-model constructed based on a reinforcement learning algorithm;
extracting time-dependent features in a plurality of behavior segment vectors through a time sequence feature extraction sub-model, and outputting an intermediate hidden state sequence;
Inputting the intermediate hidden state sequence into an enhanced scoring sub-model, and outputting a behavior privacy risk scoring value through a pre-trained state action cost function;
Comparing the behavior privacy risk score value with a preset privacy risk class interval to determine the privacy risk class of the corresponding page area, preferably, when the behavior privacy risk score value is greater than or equal to 0.7, judging that the privacy risk class of the corresponding page area is high risk, when the behavior privacy risk score value is greater than or equal to 0.4 and less than 0.7, judging that the privacy risk class of the corresponding page area is medium risk, and when the behavior privacy risk score value is less than 0.4, judging that the privacy risk class of the corresponding page area is low risk.
In a preferred embodiment, the intermediate hidden state sequence is input into the reinforcement scoring sub-model, and the behavioral privacy risk score value is output through a pre-trained state action cost function, based on the following formula:
;
wherein Q (s, a) represents taking action in state s I.e., the behavioral privacy risk score value corresponding to the behavioral segment vector under the current page area, alpha is the learning rate, r is the current reward value, gamma is the discount factor,Representing in future stateAll possible actions belowIs the maximum score of (2).
Specifically, in the embodiment, the user operation feature vector is divided into a plurality of behavior segment vectors, the time dependency feature is extracted by utilizing the long-term and short-term memory network, the privacy risk score is output by combining the reinforcement learning model and is matched to the risk level interval, so that the privacy risk assessment not only has time sequence sensing capability, but also can adapt to the optimization judgment strategy, the problem that the static scoring model cannot adapt to behavior change is solved, the potential privacy risk of the user in different areas can be more accurately identified, and the dynamic and intelligent levels of front-end risk identification are improved.
In one embodiment, after performing the privacy risk assessment operation according to the user operation feature vector to obtain the privacy risk levels of the plurality of page areas, the method further includes:
Based on privacy risk levels of the plurality of page areas, corresponding privacy protection visual feedback is generated to the front-end interaction page, wherein the high-risk page area preferably adopts a red particle icon, the middle-risk page area preferably adopts a red particle icon, and the low-risk page area preferably adopts a green particle icon.
In one embodiment, based on privacy risk levels of a plurality of page areas, privacy protection policies matching the privacy risk levels are respectively selected from a pre-stored privacy protection policy set, and a plurality of policy tasks are generated, including:
Setting a privacy risk level at least corresponding to one or more privacy protection strategies, generating a gradient strategy table, wherein preferably, when the privacy risk level is high risk, the corresponding privacy protection strategy at least comprises the steps of starting homomorphic encryption, displaying a virtual keyboard and preventing screen capturing, when the privacy risk level is medium risk, the corresponding privacy protection strategy at least comprises the steps of using format to keep encryption and displaying an input frame watermark, and when the privacy risk level is low risk, the corresponding privacy protection strategy at least comprises the step of performing data desensitization;
Acquiring the privacy risk level of the page area, inquiring a gradient policy table, and determining a privacy protection policy corresponding to the privacy risk level of the page area as a target policy;
and combining the page area identification of the page, the type identification of the target strategy and the execution parameters to construct a corresponding strategy task.
In particular, in the embodiment, the privacy protection policy can be dynamically matched according to the risk level of the page area by establishing the mapping relation between the privacy risk level and the multi-level protection policy, namely the gradient policy table, meanwhile, the construction of the policy task not only considers the area identification and the policy type, but also comprises specific execution parameters, and each policy is ensured to have a definite execution target and context suitability, so that individuation and execution effectiveness of policy configuration are improved, and the defects of single policy response and lack of flexibility in the prior art are overcome.
In one embodiment, predicting a subsequent interaction probability for remaining page areas of the plurality of page areas other than the user's current interaction page area based on the user historical behavior data comprises:
extracting historical behavior data of a user to obtain a time sequence record of a user interaction event;
Constructing a behavior path sequence according to the time sequence record of the user interaction event;
extracting a page area identifier of a user interaction event in the behavior path sequence, obtaining an interaction area node sequence, coding according to time sequence, and constructing a state set;
on the state set, counting the actual jump times between two states in sequence, setting nodes as states, taking edges as user jump behaviors and side weights as jump frequencies, constructing a state jump frequency diagram and generating a state transition probability matrix;
and taking the current interaction page area of the user as input, performing multi-step jump simulation on the state transition probability matrix, and respectively outputting the subsequent interaction probabilities of the residual page areas.
Specifically, in the embodiment, the user interaction behavior path sequence is constructed, the state transition probability matrix is generated based on the state jump frequency diagram, multi-step prediction can be carried out on the subsequent interaction page area of the user, so that the high-probability interaction area is obtained in advance, meanwhile, the privacy protection strategy can be prepared in advance before the user actually touches the page, the protection prepositivity and the resource scheduling efficiency are improved, and the protection time lag is effectively avoided.
In one embodiment, in combination with the corresponding privacy risk level, the plurality of policy tasks are ordered to generate a policy task execution sequence, including:
according to the privacy risk level and the subsequent interaction probability of the residual page area, weighting calculation is carried out to obtain the ranking score of the residual page area;
and according to the sorting scores of the residual page areas, performing descending order arrangement on a plurality of corresponding strategy tasks to obtain a strategy task execution sequence.
Specifically, in the embodiment, the privacy risk level of the page area and the subsequent interaction probability are weighted and combined to construct the ordering score and order the policy tasks, so that the priority execution of the key protection tasks with high risk and high interaction probability under the limited system resources is facilitated, the dynamic scheduling and priority management of the policy resources can be realized, the timeliness and the practicability of privacy protection are ensured, and the execution efficiency and the risk relieving capability of the whole privacy protection system are improved.
In one embodiment, pre-caching and sequentially executing execution parameters of a plurality of policy tasks based on an execution sequence, comprising:
setting the number of cacheable tasks according to the current system resource state;
Selecting a plurality of strategy tasks which are ranked at the front from the strategy task execution sequence by taking the number of the cacheable tasks as the selection number, and writing a plurality of corresponding execution parameters into a front-end cache area;
scheduling and executing the cached strategy tasks in sequence according to the sequence of the strategy task execution sequence;
and responding to the completion of the execution of the strategy task, acquiring the execution parameters of the strategy task which is not executed yet and is sequenced to the front in the strategy task execution sequence so as to replace the cache data of the completed task in the front-end cache area.
Specifically, in the embodiment, the number of the cacheable tasks is set, the dynamic cache management is performed by combining with the execution sequence of the strategy tasks, meanwhile, the execution record is generated after the tasks are completed, and the storage is performed through the blockchain, so that the consistency, the integrity and the non-tamper property of the protection strategy in the execution process are ensured, the resource preloading and the scheduling optimization of the strategy tasks before the execution are realized, a reliable execution record chain is also established, the verifiability of a user on the privacy protection behavior of the system is enhanced, and the problem that the strategy execution process cannot be audited in the prior art is effectively solved.
In a preferred embodiment, setting the number of cacheable tasks based on the current system resource status includes:
The method comprises the steps of obtaining available cache space of a current equipment end system and average execution parameter size of strategy tasks, and calculating the number of the cacheable tasks by combining a preset maximum concurrent cache upper limit, wherein the number is specifically based on the following formula:
;
S avail represents the remaining available buffer space of the current system, S avg is the average size of a single policy parameter packet, K is the maximum number of parallel buffer tasks, and is preferably 10-20.
Specifically, in the embodiment, the number of the cacheable tasks is dynamically set according to the current resource state of the system, so that the caching behavior of the strategy tasks can be adaptively matched with the running environment of the equipment, more tasks with high priority are loaded in advance by fully utilizing the caching space when the resources are sufficient, the timeliness of strategy response is improved, the caching scale is automatically contracted when the resources are limited, the system is prevented from being blocked or abnormal in function due to excessive occupation of the memory, the overall resource utilization efficiency and running stability of the system are effectively improved, the adaptation capability of a privacy protection mechanism under various client environments is enhanced, and the problems of strategy loading delay or uneven system load caused by the fixed number of the caching tasks and the stiff resource scheduling in the prior art are solved.
In one embodiment, after pre-caching the execution parameters of the plurality of policy tasks and executing them in sequence based on the execution sequence, the method further comprises:
And constructing a hierarchical blockchain architecture formed by cooperation of a alliance chain and a public chain, wherein the alliance chain is set to store basic metadata such as user identity labels, privacy preference settings, policy configuration and the like and has node interaction authority control, the public chain is set to record sensitive operation log anchoring information subjected to hash processing and ensure verifiability and non-tamper property of key behaviors, and a data interface on the chain is set to support automatic hash uplink of task execution records.
In one embodiment, after building the hierarchical blockchain architecture that is formed by the federation chain in cooperation with the public chain, further comprising:
In response to the completion of the execution of the strategy task, acquiring the task identifier, the execution time and the execution result information of the completed strategy task, splicing the task identifier, the execution time and the execution result information into a strategy execution record data packet, carrying out SHA-256 hash on the strategy execution record data packet, and generating a summary value with a fixed length;
The abstract value is used as a sensitive strategy operation anchor value, and is combined with the task identifier and the user identifier to form an anchor record;
Writing the anchored record into the block through the alliance chain to form an on-chain evidence of the strategy task, and anchoring the record to the public chain to enhance the public verifiability to form a double trusted path in response to the task belonging to the operation of high sensitivity level, namely that the corresponding privacy risk level belongs to the high level;
Based on the state-parameter pair of task execution, a zero knowledge proof (such as zk-SNARK structure) is generated, and the zero knowledge proof is stored in the intelligent contract together with the task identification and the on-chain certificate for being called by an auditor or a user client to realize on-chain verification.
The embodiment realizes the trusted memory certificate and the verifiability protection of the privacy policy task execution record by introducing a hierarchical blockchain architecture formed by the cooperation of the alliance chain and the public chain. A complete set of 'controllable interaction', 'non-falsification', 'verification' trust mechanism is established by recording user identity and policy metadata in the coalition chain and anchoring the sensitive operation abstract to the public chain in the high-risk policy operation scene. Meanwhile, by combining with a zero knowledge proof protocol, an external auditor or a user client is allowed to verify the policy execution state on the premise of not exposing plaintext data, the technical problem that the policy execution process in the existing privacy protection scheme lacks transparency, credibility and auditability is effectively solved, and the security, compliance and user trust feeling of the system are further enhanced.
From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment.
The embodiment of the application also provides a privacy protection device, as shown in fig. 7, comprising:
the operation recording module is used for acquiring operation behavior data of a user on the front-end interaction page and constructing a user operation feature vector based on the operation behavior data;
the risk assessment module is used for executing privacy risk assessment operation according to the user operation feature vector to obtain privacy risk levels of a plurality of page areas of the front-end interaction page;
The policy task generation module is used for respectively selecting privacy protection policies matched with the privacy risk levels from a pre-stored privacy protection policy set based on the privacy risk levels of the plurality of page areas to generate a plurality of policy tasks;
The task scheduling module is used for executing corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probability of the rest page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk levels to generate a strategy task execution sequence;
and the task execution module is used for pre-caching the execution parameters of the plurality of strategy tasks based on the execution sequence and sequentially executing the strategy tasks.
The operation recording module is also used for calling performance interfaces provided by the front-end interaction page, collecting operation records of a user in a plurality of page areas to obtain operation behavior data, wherein the operation behavior data at least comprise a screen coordinate value sequence of a clicking event, entering time and leaving time of the page areas, rolling speed and rolling acceleration of the page rolling event and an input character sequence of an input content change event, obtaining a clicking position distribution standard deviation according to the screen coordinate value sequence of the clicking event, obtaining an average stay time of the page areas according to the entering time and the leaving time of the page areas, obtaining a rolling acceleration integral value according to the rolling speed and the rolling acceleration of the page rolling event, calculating a weight value of an input text sensitive word according to the input character sequence of the input content change event, calling a sensor of equipment in which the front-end interaction page is located, obtaining a screen inclination angle and environment light intensity, generating a sensing auxiliary characteristic value, and splicing the clicking position distribution standard deviation, the page area average stay time, the rolling acceleration integral value, the weight value of the input text sensitive word and the sensing auxiliary characteristic value to obtain a user operation characteristic vector.
The risk assessment module is further used for dividing user operation feature vectors according to page areas to obtain a plurality of behavior segment vectors, constructing a privacy risk assessment model, wherein the privacy risk assessment model comprises a time sequence feature extraction sub-model constructed based on a long-short-term memory network and a reinforcement grading sub-model constructed based on a reinforcement learning algorithm, extracting time dependent features in the plurality of behavior segment vectors through the time sequence feature extraction sub-model, outputting an intermediate hidden state sequence, inputting the intermediate hidden state sequence into the reinforcement grading sub-model, outputting a behavior privacy risk grading value through a pre-trained state action cost function, and comparing the behavior privacy risk grading value with a preset privacy risk grade interval to determine the privacy risk grade of the corresponding page area.
The policy task generating module is further used for setting a privacy risk level at least corresponding to one or more privacy protection policies to generate a gradient policy table, acquiring the privacy risk level of the page area, inquiring the gradient policy table, determining the privacy protection policy corresponding to the privacy risk level of the page area as a target policy, and combining the page area identification of the page, the type identification of the target policy and the execution parameters to construct a corresponding policy task.
The task scheduling module is also used for extracting historical behavior data of a user to obtain time sequence records of user interaction events, constructing a behavior path sequence according to the time sequence records of the user interaction events, extracting page area identifiers of the user interaction events in the behavior path sequence to obtain an interaction area node sequence, coding according to time sequence to construct a state set, sequentially counting actual jump times between two states on the state set, setting nodes as states, taking side weights as jump frequencies, constructing a state jump frequency chart and generating a state transfer probability matrix, taking a current interaction page area of the user as input, performing multi-step jump simulation on the state transfer probability matrix, respectively outputting subsequent interaction probabilities of a plurality of page areas, weighting and calculating according to privacy risk grades and the subsequent interaction probabilities of the page areas to obtain a sequencing score of the page areas, and performing descending order arrangement on a plurality of corresponding strategy tasks according to the sequencing scores of the page areas to obtain a strategy task execution sequence.
The task execution module is also used for setting the number of cacheable tasks according to the current system resource state, selecting a plurality of strategy tasks with front ordering from the strategy task execution sequence and writing a plurality of corresponding execution parameters into a front-end cache area by taking the number of cacheable tasks as the selection number, sequentially scheduling and executing the cached strategy tasks according to the sequence of the strategy task execution sequence, and acquiring the execution parameters of the next strategy task with front ordering which is not executed in the strategy task execution sequence to replace the cache data of the completed task in the front-end cache area in response to the completion of the strategy task execution.
The description of the features in the embodiment corresponding to the privacy protection apparatus may refer to the related description of the embodiment corresponding to the privacy protection method, which is not described in detail herein.
An embodiment of the application also provides an electronic device, as shown in fig. 8, comprising a memory in which a computer program is stored and a processor arranged to run the computer program to perform the steps of any of the privacy preserving method embodiments described above.
Embodiments of the present application also provide a computer readable storage medium having a computer program stored therein, wherein the computer program is arranged to perform the steps of any of the privacy preserving method embodiments described above when run.
In an exemplary embodiment, the computer readable storage medium may include, but is not limited to, a U disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a removable hard disk, a magnetic disk, or an optical disk, etc. various media in which a computer program may be stored.
Embodiments of the present application also provide a computer program product comprising a computer program which, when executed by a processor, implements the steps of any of the privacy preserving method embodiments described above.
Embodiments of the present application also provide another computer program product comprising a non-volatile computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of any of the privacy preserving method embodiments described above.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The privacy protection method, the privacy protection device, the privacy protection equipment, the privacy protection storage medium and the privacy protection product provided by the application are described in detail. The principles and embodiments of the present application have been described herein with reference to specific examples, the description of which is intended only to facilitate an understanding of the method of the present application and its core ideas. It should be noted that it will be apparent to those skilled in the art that the present application may be modified and practiced without departing from the spirit of the present application.
Claims (9)
1. A privacy preserving method, comprising:
acquiring operation behavior data of a user on a front-end interaction page, and constructing a user operation feature vector based on the operation behavior data;
According to the user operation feature vector, privacy risk assessment operation is carried out, and privacy risk levels of a plurality of page areas of the front-end interaction page are obtained;
Based on privacy risk levels of a plurality of page areas, respectively selecting privacy protection strategies matched with the privacy risk levels from a pre-stored privacy protection strategy set to generate a plurality of strategy tasks;
Executing the corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probabilities of the residual page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk levels to generate a strategy task execution sequence, wherein the strategy task execution sequence is generated by weighting calculation according to the privacy risk levels and the subsequent interaction probabilities of the residual page areas;
and pre-caching the execution parameters of a plurality of strategy tasks based on the execution sequence and sequentially executing the strategy tasks.
2. The privacy preserving method of claim 1, wherein the obtaining operational behavior data of the user on the front-end interaction page and constructing the user operation feature vector based on the operational behavior data comprises:
invoking a performance interface provided by the front-end interaction page, and collecting operation records of a user in a plurality of page areas to obtain operation behavior data, wherein the operation behavior data at least comprises a screen coordinate value sequence of a clicking event, entering time and leaving time of the page area, rolling speed and rolling acceleration of a page rolling event and an input character sequence of an input content change event;
Obtaining a click position distribution standard deviation according to the screen coordinate value sequence of the click event;
obtaining the average residence time of the page area according to the entering time and the leaving time of the page area;
obtaining a rolling acceleration integral value according to the rolling speed and the rolling acceleration of the page rolling event;
calculating a weight value of an input text sensitive word according to the input character sequence of the input content change event;
calling a sensor of equipment where the front-end interaction page is located, acquiring a screen inclination angle and ambient light intensity, and generating a sensing auxiliary characteristic value;
and splicing the click position distribution standard deviation, the page area average residence time, the rolling acceleration integral value, the weight value of the input text sensitive word and the sensing auxiliary characteristic value to obtain the user operation characteristic vector.
3. The privacy preserving method of claim 1, wherein the performing the privacy risk assessment operation according to the user operation feature vector, to obtain privacy risk levels of a plurality of page areas, comprises:
Dividing the user operation feature vector according to a page area to obtain a plurality of behavior segment vectors;
constructing a privacy risk assessment model, wherein the privacy risk assessment model comprises a time sequence feature extraction sub-model constructed based on a long-short-term memory network and a reinforcement scoring sub-model constructed based on a reinforcement learning algorithm;
Extracting time-dependent features in a plurality of behavior segment vectors through the time sequence feature extraction submodel, and outputting an intermediate hidden state sequence;
Inputting the intermediate hidden state sequence into the strengthening scoring submodel, and outputting a behavior privacy risk scoring value through a pre-trained state action cost function;
And comparing the behavior privacy risk scoring value with a preset privacy risk class interval to determine the privacy risk class of the corresponding page area.
4. The privacy preserving method of claim 3, wherein the generating a plurality of policy tasks based on privacy risk levels of a plurality of the page areas by selecting privacy protection policies matching the privacy risk levels from a pre-stored privacy protection policy set, respectively, comprises:
setting a privacy risk level at least corresponding to one or more privacy protection strategies to generate a gradient strategy table;
Acquiring the privacy risk level of the page area, inquiring the gradient policy table, and determining a privacy protection policy corresponding to the privacy risk level of the page area as a target policy;
And combining the page area identification of the page, the type identification of the target strategy and the execution parameters to construct a corresponding strategy task.
5. The privacy preserving method of claim 1, wherein predicting the subsequent interaction probabilities of remaining page areas of the plurality of page areas other than the user's current interaction page area based on user historical behavior data comprises:
extracting the historical behavior data of the user to obtain a time sequence record of the user interaction event;
constructing a behavior path sequence according to the time sequence record of the user interaction event;
Extracting a page area identifier of a user interaction event in the behavior path sequence, obtaining an interaction area node sequence, coding according to time sequence, and constructing a state set;
On the state set, counting the actual jump times between two states in sequence, setting nodes as states, taking edges as user jump behaviors and side weights as jump frequencies, constructing a state jump frequency diagram and generating a state transition probability matrix;
and taking the current interaction page area of the user as input, performing multi-step jump simulation on the state transition probability matrix, and respectively outputting the subsequent interaction probability of the residual page area.
6. The privacy preserving method of claim 1, wherein the pre-caching and sequentially executing the execution parameters of the plurality of policy tasks based on the execution sequence comprises:
setting the number of cacheable tasks according to the current system resource state;
Taking the number of the cacheable tasks as the selection number, selecting a plurality of strategy tasks with front ordering from a strategy task execution sequence, and writing a plurality of corresponding execution parameters into a front-end cache area;
scheduling and executing the cached strategy tasks in sequence according to the sequence of the strategy task execution sequence;
And responding to the completion of the execution of the strategy task, acquiring the execution parameters of the strategy task which is not executed yet and is sequenced to the front in the strategy task execution sequence so as to replace the cache data of the completed task in the front-end cache region.
7. A privacy preserving apparatus, comprising:
The operation recording module is used for acquiring operation behavior data of a user on the front-end interaction page and constructing a user operation feature vector based on the operation behavior data;
The risk assessment module is used for executing privacy risk assessment operation according to the user operation feature vector to obtain privacy risk levels of a plurality of page areas of the front-end interaction page;
The policy task generation module is used for respectively selecting privacy protection policies matched with the privacy risk levels from a prestored privacy protection policy set based on the privacy risk levels of the page areas to generate a plurality of policy tasks;
The task scheduling module is used for executing the corresponding strategy tasks on the current interaction page area of the user, predicting the subsequent interaction probabilities of the residual page areas except the current interaction page area of the user in the plurality of page areas based on the historical behavior data of the user, and sequencing the plurality of strategy tasks by combining the corresponding privacy risk grades to generate a strategy task execution sequence, wherein the strategy task execution sequence comprises the steps of weighting calculation to obtain the sequencing score of the residual page areas according to the privacy risk grade and the subsequent interaction probability of the residual page areas;
according to the sorting scores of the residual page areas, performing descending order arrangement on a plurality of corresponding strategy tasks to obtain the strategy task execution sequence;
and the task execution module is used for pre-caching the execution parameters of a plurality of strategy tasks based on the execution sequence, sequentially executing the strategy tasks, generating corresponding task execution records and writing the corresponding task execution records into the blockchain network.
8. An electronic device, comprising:
a memory for storing a computer program;
A processor for implementing the steps of the privacy preserving method as claimed in any of claims 1 to 6 when executing said computer program.
9. A computer readable storage medium, wherein a computer program is stored in the computer readable storage medium, wherein the computer program, when executed by a processor, implements the steps of the privacy preserving method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510969432.1A CN120470628B (en) | 2025-07-15 | 2025-07-15 | Privacy protection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510969432.1A CN120470628B (en) | 2025-07-15 | 2025-07-15 | Privacy protection method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN120470628A CN120470628A (en) | 2025-08-12 |
| CN120470628B true CN120470628B (en) | 2025-09-09 |
Family
ID=96628904
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510969432.1A Active CN120470628B (en) | 2025-07-15 | 2025-07-15 | Privacy protection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN120470628B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120951387B (en) * | 2025-09-25 | 2026-02-10 | 广州乐税信息科技有限公司 | Data sharing and privacy protection platform in response to the needs of tax digitalization |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115080143A (en) * | 2022-06-14 | 2022-09-20 | 北京字跳网络技术有限公司 | Page resource preloading method, device, equipment and storage medium |
| CN119557719A (en) * | 2023-08-31 | 2025-03-04 | 华为技术有限公司 | A strategy generation method, strategy application method, device and electronic device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12430465B2 (en) * | 2022-08-26 | 2025-09-30 | Samsung Electronics Co., Ltd. | Method and electronic device for personalizing user data privacy and utility |
| CN119746414A (en) * | 2024-12-10 | 2025-04-04 | 乌雀(深圳)信息科技有限公司 | Data processing method and device |
| CN119883442A (en) * | 2025-01-16 | 2025-04-25 | 中国工商银行股份有限公司 | Webpage task operation method and device, electronic equipment and computer program product |
-
2025
- 2025-07-15 CN CN202510969432.1A patent/CN120470628B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115080143A (en) * | 2022-06-14 | 2022-09-20 | 北京字跳网络技术有限公司 | Page resource preloading method, device, equipment and storage medium |
| CN119557719A (en) * | 2023-08-31 | 2025-03-04 | 华为技术有限公司 | A strategy generation method, strategy application method, device and electronic device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN120470628A (en) | 2025-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108229963A (en) | The Risk Identification Method and device of user's operation behavior | |
| CN114627330A (en) | Time sequence flow prediction method and device, storage medium and electronic equipment | |
| CN120470628B (en) | Privacy protection method, device, equipment and storage medium | |
| US20200250587A1 (en) | Framework for multi-tenant data science experiments at-scale | |
| CN109543891A (en) | Method for building up, equipment and the computer readable storage medium of capacity prediction model | |
| CN113515771B (en) | Data sensitivity determination method, electronic device and computer-readable storage medium | |
| CN119966681B (en) | A method for obtaining user data distribution privacy based on federated learning | |
| CN114218077A (en) | Software quality evaluation method, device, equipment and readable storage medium | |
| CN114461869B (en) | Service characteristic data processing method and device, electronic equipment and storage medium | |
| CN113705682B (en) | User behavior feature processing method and device | |
| CN118966786A (en) | Risk control method, device, equipment and storage medium based on rule engine | |
| CN117272113B (en) | Illegal behavior detection method and system based on virtual social network | |
| CN114565453A (en) | Model construction method, credit risk assessment device, credit risk assessment equipment and credit risk assessment medium | |
| CN115485662A (en) | Quota request resolution on a computing platform | |
| CN110162939A (en) | Man-machine recognition methods, equipment and medium | |
| CN112887408B (en) | System and method for solving data state sharing of multi-kernel browser | |
| CN114581130A (en) | Bank website number assigning method and device based on customer portrait and storage medium | |
| CN114372862A (en) | Data processing method, data processing device, computer equipment and storage medium | |
| CN119544278A (en) | Access flow control method, device, computer equipment and storage medium | |
| KR102471731B1 (en) | A method of managing network security for users | |
| CN113794699B (en) | Network analysis processing method | |
| CN118797465A (en) | Abnormal group identification method, device, equipment and storage medium | |
| CN113868593B (en) | A network security risk assessment method, device, equipment and medium | |
| EP3842966B1 (en) | Identifying a user from the sequence of windows opened on a user interface | |
| CN115713231A (en) | Wind control method, system, medium and electronic device based on user portrait |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |