Computer network-based privacy data protection method

Abstract

The present application relates to the technical field of data processing, in particular to a privacy data protection method based on a computer network, which comprises the following steps: obtaining communication data of a smart wearable device; dividing reference data in the communication data according to time sequence information to obtain a plurality of period data; determining a traffic anomaly representation degree corresponding to the plurality of period data; determining a total monitoring anomaly performance value of the plurality of period data based on monitoring values of the plurality of period data; determining a risk value of current data in the plurality of period data based on the traffic anomaly representation degree corresponding to the plurality of period data and the total monitoring anomaly performance value; and updating a key of the current data in the plurality of period data based on the risk value. The method can more accurately calculate the risk value of data, replace the key in time according to the risk value, and thus improve the security of the privacy data of the smart wearable device.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically to a method for protecting privacy data based on computer networks within the field of data processing technology. Background Technology

[0002] With the rapid development of IoT technology, smartwatches and wearable devices have become an important part of the smart terminal ecosystem. By integrating sensors and advanced algorithms, wearable devices can monitor users' physiological indicators and provide corresponding personalized services. However, the widespread adoption of wearable devices also brings challenges to privacy data protection. The data collected by wearable devices is highly sensitive personal information. This data not only relates to users' personal privacy but may also involve health information.

[0003] Privacy protection in related technologies involves encrypting the data and ensuring key security through periodic key changes. However, periodic key changes are insufficient to withstand complex network attacks. When the key change interval is long, key security is compromised, while frequent key changes impose a significant burden. Therefore, neither approach adequately guarantees the security of privacy data. Summary of the Invention

[0004] To address the technical problem that existing technologies cannot adequately guarantee the security of privacy data, the present invention aims to provide a privacy data protection method based on computer networks. The specific technical solution adopted is as follows:

[0005] In a first aspect, embodiments of the present invention provide a privacy data protection method based on a computer network, the method comprising:

[0006] Acquire communication data from smart wearable devices;

[0007] According to the time sequence information, the reference data in the communication data is divided into multiple time period data;

[0008] Determine the traffic anomaly characterization degree corresponding to the multiple time period data;

[0009] Based on the monitoring values ​​of the data from the multiple time periods, the total monitoring anomaly value of the data from the multiple time periods is determined;

[0010] Based on the traffic anomaly characterization degree and the total monitored anomaly performance value corresponding to the multiple time period data, the risk value of the current data in the multiple time period data is determined;

[0011] Based on the risk value, update the key of the current data in the multiple time period data.

[0012] Secondly, a privacy data protection device for a smart wearable device is provided, the device comprising:

[0013] The acquisition module is used to acquire communication data from smart wearable devices;

[0014] The segmentation module is used to segment the reference data in the communication data according to the time sequence information to obtain multiple time period data;

[0015] The first determining module is used to determine the traffic anomaly characterization degree corresponding to the multiple time period data;

[0016] The second determining module is used to determine the total monitoring anomaly value of the multiple time period data based on the monitoring values ​​of the multiple time period data;

[0017] The third determining module is used to determine the risk value of the current data in the multiple time period data based on the traffic anomaly characterization degree corresponding to the multiple time period data and the total monitored anomaly performance value;

[0018] An update module is used to update the key of the current data in the multiple time period data based on the risk value.

[0019] Thirdly, a computer program product is provided, comprising: computer program code, which, when run on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0020] Fourthly, a computer-readable storage medium is provided that stores computer program code, which, when executed on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0021] This invention offers the following advantages: After acquiring communication data from a smart wearable device, the reference data within the communication data is divided according to time sequence information to obtain multiple time-segment data. The traffic anomaly characterization degree corresponding to each of these multiple time-segment data is determined, allowing the identification of whether historical communication traffic has exhibited anomalies. Subsequently, the total monitoring anomaly performance value of the multiple time-segment data is determined based on the monitoring values. Analyzing this total monitoring anomaly performance value accurately reflects whether the monitoring values ​​experienced by the smart wearable device have become abnormal. Finally, based on the traffic anomaly characterization degree and the total monitoring anomaly performance value corresponding to the multiple time-segment data, a risk value for each time-segment data is determined; and based on this risk value, the key for the current data in the communication data is updated. Thus, by combining the monitoring values ​​in the communication data of the smart wearable device to calculate the total monitoring anomaly performance value for each time-segment data, the risk value of the current data can be calculated more accurately. Timely key replacement based on the risk value improves the security of the smart wearable device's privacy data. Attached Figure Description

[0022] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram illustrating the implementation process of the privacy data protection method based on computer networks provided in an embodiment of the present invention;

[0024] Figure 2 This is a schematic diagram illustrating another implementation of the privacy data protection method based on computer networks provided in this embodiment of the invention.

[0025] Figure 3 This is a schematic diagram of another implementation process of the privacy data protection method based on computer networks provided in this embodiment of the invention;

[0026] Figure 4 This is a schematic diagram of another implementation process of the privacy data protection method based on computer networks provided in the embodiments of the present invention. Detailed Implementation

[0027] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the privacy data protection method based on computer networks proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0028] In the description of the embodiments of the present invention, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present invention, "multiple" means two or more.

[0029] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0031] The specific solution of the privacy data protection method based on computer networks provided by the present invention will be described in detail below with reference to the accompanying drawings. Please refer to... Figure 1 The diagram illustrates a flowchart of a privacy data protection method based on a computer network according to an embodiment of the present invention, the method comprising:

[0032] 101, Obtain communication data from smart wearable devices.

[0033] Here, communication records are obtained from the communication history of the user end and the server end of the smart wearable device. This invention uses the historical communication records of the previous 24 hours as reference data. The communication between the user end and the server end of the smart wearable device is mainly divided into two communication modes: a question-and-answer mode and a monitoring mode.

[0034] Among them, the question-and-answer mode is mainly the command interaction between the user end and the server end of the smart wearable device. That is, the server end sends a query or control command, and the user end returns the corresponding acknowledgment data or adjusts the response parameters after receiving the query command, so as to bring a more comfortable experience to the user.

[0035] The monitoring mode involves continuously transmitting various data to the server at fixed frequencies and intervals. For example, continuously transmitting monitoring data from a smart wearable device allows the server to assess user comfort and adjust parameters such as the device's shape accordingly. Within the reference data set, all commands sent from the server to the user are retrieved as query data, and all control instructions are selected based on the communication protocol content to obtain a set A of all control instructions. Similarly, all commands sent from the user to the server are retrieved from the reference data set as transmission data, resulting in a dataset B of all transmission time periods. Since the transmission data contains multiple data types, it is also necessary to divide it according to the communication address to obtain datasets of multiple transmission time periods. The datasets for each time period consist of data collected by various sensors of the smart wearable device, primarily monitoring data (which is typically distributed across multiple sensors to obtain more accurate data, thus often comprising datasets from multiple time periods). Additionally, the most recent historical communication data of length L is read as the current data; in this embodiment, L is set to 5.

[0036] 102. According to the time sequence information, the reference data in the communication data is divided to obtain multiple time period data.

[0037] Here, the timing information includes the duration of the communication data and the time points along that duration. In some possible implementations, the duration corresponding to the reference data is divided according to the timing information to obtain multiple time periods; and the data of the reference data within each of the multiple time periods is determined to obtain the multiple time period data. Thus, by dividing the duration corresponding to the reference data in the communication data into multiple time periods, the reference data can be divided into multiple time period data, facilitating subsequent anomaly analysis of the data in each time period.

[0038] In some possible implementations, due to the large number of smart wearable devices in certain scenarios, and the large amount of diverse data involved in the interaction between the smart wearable devices and the user, security vulnerabilities inevitably exist, leading to key leaks. When the key is at risk of being leaked, attackers can mimic normal interaction data to interact with the server, thereby stealing more private data.

[0039] While the generated data can often be well disguised in a single data sensor, the hidden data connections between multiple sensors are difficult to imitate. Therefore, the greater the change in the correlation between the various transmitted data of a smart wearable device, the more abnormal the historical communication data is considered, the higher the risk of key leakage, and the more necessary it is to replace the key.

[0040] In the aforementioned reference data, the duration of the reference data is divided into several time periods, each in 1-minute increments, based on the timestamp. ,in, This represents the m-th time period.

[0041] 103. Determine the traffic anomaly characterization degree corresponding to the multiple time period data.

[0042] Here, in the historical communication records of smart wearable devices, abnormal signals are often reflected in traffic data. When communication traffic data is abnormal, the degree of signal abnormality is considered higher. In some possible implementations, step 103 above can be achieved through... Figure 2 The steps shown are to be implemented as follows:

[0043] 201. Determine the statistical flow values ​​of the multiple time period data to obtain multiple statistical flow values.

[0044] Here, after obtaining data from various time periods... Next, obtain the statistical flow values ​​for each time period. Since the data packets sent contain data size information, the statistical traffic value can be obtained by summing the traffic from multiple communications per minute.

[0045] 202. Based on the multiple statistical flow values, determine the flow anomaly characterization degree corresponding to the multiple time period data.

[0046] Here, multiple traffic clusters are obtained by clustering the multiple statistical traffic values. For example, K-means clustering can be performed on the multiple statistical traffic values ​​to obtain multiple traffic clusters. Then, based on the multiple traffic clusters, the traffic anomaly characterization degree corresponding to the multiple time periods is determined. Here, the difference between the statistical traffic value of any time period and the traffic value of the cluster center is calculated through multiple traffic clusters. This difference is combined with the traffic variance to accurately calculate the traffic anomaly characterization degree of each time period.

[0047] In some possible implementations, for any given time period, firstly, the cluster center flow value and flow variance of the cluster to which the data belongs are determined; then, the difference between the statistical flow value of the data and the cluster center flow value is determined; finally, based on the difference and the flow variance, the flow anomaly characterization degree corresponding to the data in that time period is determined. In this way, by obtaining the statistical flow values ​​of data for each time period, the flow anomaly characterization degree of each time period can be accurately calculated, facilitating the analysis of whether flow anomalies occur in the data for each time period.

[0048] Here, since the communication data of smart wearable devices typically falls into two modes—working mode and idle mode—K-means clustering is performed on the statistical traffic for each time period with K=2, resulting in two traffic clusters, W1 and W2. Then, the mode of each cluster is used as the cluster center, and the traffic at each cluster center is... Calculate the variance of flow values ​​within a cluster. ;

[0049] The traffic anomaly characterization degree of each time period data is calculated according to the performance of its respective cluster, as shown in formula (1):

[0050] (1);

[0051] in: Indicates time period The flow rate value; For time period The cluster center flow value of the cluster to which it belongs; Indicates time period The flow variance of the cluster to which the cluster belongs; Indicates time period The degree of characterization of traffic anomalies.

[0052] 104. Based on the monitoring values ​​of the multiple time period data, determine the total monitoring anomaly value of the multiple time period data.

[0053] Here, multiple monitoring sensors are used to collect initial monitoring values ​​for each time period, resulting in multiple initial monitoring values ​​for each time period. The average of these multiple initial monitoring values ​​is then used as the monitoring value for each time period. In this way, by collecting initial monitoring values ​​for each time period from multiple angles of monitoring sensors and using the average of these initial monitoring values ​​as the monitoring value for each time period, the monitoring values ​​for each time period can be made more accurate. In some possible implementations, multiple initial monitoring values ​​Bn exist for multiple monitoring sensors within each time period, and the average of the initial monitoring values ​​from each monitoring sensor is calculated. As the monitoring value of the time period to which the time period data belongs.

[0054] In some possible implementations, when communicating between the server and user of a smart wearable device, a large amount of sensor data is transmitted, including monitoring data collected by sensors in different locations. By analyzing this monitoring data, the user's usage habits and behavioral patterns can be obtained.

[0055] When attackers mimic normal interaction data to interact with the server, they often generate random, simulated data. This simulated data can often be well disguised within a single data sensor, but the hidden data connections between multiple sensors are difficult to imitate. Therefore, the degree of anomaly can be determined by examining the relationships between multiple monitoring data points. When a user is using a smart wearable device normally, the overall monitoring data is relatively stable. Although there may be some variations in individual monitoring data, the overall monitoring data from multiple sensors is usually quite stable. Furthermore, since users may have different usage patterns, the degree of anomaly can be determined by assessing the overall performance of the monitoring data. Based on this, step 104 above can be... Figure 3 The steps shown are to be implemented as follows:

[0056] 301. Cluster the multiple monitoring values ​​corresponding to the data in each time period to obtain the monitoring clusters of the data in each time period.

[0057] Here, the Bn monitoring values ​​of multiple monitoring sensors are calculated for each time period. The sum of these values ​​serves as the total monitoring value for each time period. N represents the number of sensors. This represents the total monitoring value for the m-th time period. (The remaining text appears to be incomplete and requires further context.) Total monitoring value Clustering was performed using mean drift as a clustering parameter, with a mean drift radius of r = 1 kPa, resulting in v monitored clusters. Each cluster represents a user's usage pattern.

[0058] 302. Based on the monitoring clusters corresponding to the data in each time period, determine the degree of fluctuation of the monitoring values ​​of the data in each time period.

[0059] Here, in each usage pattern cluster Calculate data for each time period Total monitoring value The variance is used as the degree of fluctuation of the monitored value. .

[0060] 303. Based on the fluctuation of the monitored values ​​of the data in each time period, determine the degree of belonging to the cluster to which the data in each time period belongs.

[0061] Here, for each time period in the current data, we obtain its corresponding cluster. Calculate its degree of belonging to its respective cluster, as shown in formula (2):

[0062] (2);

[0063] in: Indicates the time period in the current data Overall monitoring; Represents each time period in the current data To which category Average total monitoring; Representing a cluster The monitoring fluctuation variance; Indicates the time period in the current data Its affiliation with its respective cluster.

[0064] 304. Based on the degree of attribution, determine the abnormal performance of the total monitoring value of the data for each time period.

[0065] Here, the number of attribute changes (cow) for each time period in the current data is obtained; the total monitoring anomaly value of the current data in each time period is calculated, as shown in formula (3):

[0066] (3);

[0067] Where: cow represents the number of times the monitored data of the smart wearable device changes. The larger the value, the more disordered the data changes and the greater the degree of abnormality. There are multiple different clusters of data in the time period after clustering. The number of times the data of different clusters alternate in the time sequence of the current data segment is cow. Indicates the time period in the current data The data is categorized by its cluster affiliation; L represents the length of the current data; and E represents the total abnormal monitoring value of the current data. Thus, by calculating the fluctuation of monitoring values ​​in each time period and analyzing the cluster affiliation of data in each time period, the abnormal monitoring value of the total monitoring value in each time period can be accurately analyzed.

[0068] 105. Based on the traffic anomaly characterization degree corresponding to the multiple time period data and the total monitoring anomaly performance value, determine the risk value of the multiple time period data.

[0069] Here, when users use smart wearable devices, the monitoring data will fluctuate and change. These changes are often caused by the user's usage habits. Due to these habits, there are certain correlations between the monitoring data read by different sensors on the smart wearable device. That is, changes in the monitoring data of sensors in different locations will lead to corresponding changes in the monitoring data of other locations. This correlation is personalized and difficult to imitate. Therefore, the more dissimilar and unreasonable the monitoring performance of multiple sensors is to historical correlations, the greater the degree of anomaly in the communication signal is considered.

[0070] In some possible implementations, the total monitoring data collected by the user during normal use is relatively fixed, and the correlation between monitoring data from different locations is also strong. The more abnormal the total monitoring data in the current data, and the more abnormal the correlation between monitoring data from different time periods, the greater the degree of signal anomaly in the current data, and the higher the risk of data transmission. Based on this, step 105 above can be... Figure 4 The steps shown are to be implemented as follows:

[0071] 401. Based on the monitoring values ​​of the multiple time periods, determine the correlation of the overall monitoring data of the multiple time periods.

[0072] Here, the monitoring correlation between data from different time periods and the overall data are analyzed by using the monitoring values ​​of data from different time periods, and the risk value of the current data is calculated by using the monitoring correlation.

[0073] In some possible implementations, step 401 above can be achieved through the following steps 411 to 413 (not shown in the figures):

[0074] 411. Based on the monitoring values ​​of the multiple time period data, determine the correlation of data changes among the multiple time period data and the differences in the amount of monitoring data changes.

[0075] 412. Based on the correlation of the data changes and the differences in the amount of changes in the monitoring data, determine the monitoring correlation between the data in the multiple time periods to obtain multiple monitoring correlations.

[0076] 413. Based on the multiple monitoring correlations, determine the correlation of the overall monitoring data.

[0077] In steps 411 to 413 above, since changes in monitoring data often occur in the distribution of the monitored data, while the sum of the monitored values ​​changes relatively little, a first-order difference sequence is obtained from the set of monitored values ​​PN of the datasets for each time period. Then, the average monitoring change within the dataset for each time period is calculated. Where M represents the total number of elements in the monitoring value set PN. Calculate the variance in the difference sequence. Then calculate the correlation of data changes between datasets in each time period, as shown in formula (4):

[0078] (4);

[0079] in: for Each element in the dataset represents the monitored changes of the dataset Bn during the time period m. This represents the average monitored change in dataset Bn, which contains data for a given time period. The dataset Bu represents the monitored changes of data over a specific time period, specifically during time period m. This represents the average monitored change in dataset Bu, which contains data for a given time period. and The variances of the monitoring differences between datasets Bn and Bu, representing time-period data respectively; This indicates the correlation of compression data changes between datasets Bn and Bu, representing the data over different time periods. The closer the value is to 1, the stronger the correlation between the two datasets; the closer it is to (-1), the stronger the inverse correlation; and the closer it is to 0, the weaker the correlation. The correlation between the datasets mentioned above only reflects the synergistic relationship between data changes and does not represent the quantitative relationship of data changes. Therefore, it is also necessary to calculate the relationship between the quantitative changes of each dataset, that is, to monitor the differences in the quantitative changes of data, as shown in formula (5):

[0080] (5);

[0081] in: This represents the average monitored change in dataset Bn, which contains data for a given time period. This represents the average monitored change in dataset Bu, which contains data for a given time period. This represents the difference in monitoring data changes between datasets Bn and Bu for different time periods. The larger this value, the less the monitoring changes between the two datasets influence each other. The monitoring correlation between data from different time periods is shown in formula (6):

[0082] (6);

[0083] in: This indicates the correlation between changes in the monitoring data of datasets Bn and Bu; This indicates the difference in the amount of change in the monitored data between datasets Bn and Bu; The monitoring correlation between datasets Bn and Bu when monitoring changes in time period data is represented.

[0084] The above analysis shows that the total monitoring data collected during normal user operation is relatively stable, and the correlation between monitoring data from different locations is also strong. The more abnormal the total monitoring data is, and the more abnormal the correlation between monitoring data from different datasets, the greater the degree of signal anomaly in the current data, and the higher the risk of data transmission. The analysis also involves obtaining the monitoring correlation between the aforementioned datasets. Take the dataset Bu that has the highest correlation with dataset Bn, and denote dataset Bu as the related dataset of dataset Bn; the overall correlation of the monitoring data is shown in formula (7):

[0085] (7);

[0086] in: Let S represent the monitoring correlation between dataset Bn of the time period data and its related dataset Bu; S represents the correlation of the overall monitoring data. Thus, by calculating the monitoring correlation of the data supports for each time period, the correlation of the overall monitoring data can be accurately obtained.

[0087] 402. The total monitoring anomaly values ​​and the flow anomaly characterization are fused to obtain the fusion result.

[0088] Here, the fusion result is obtained by multiplying the total monitored abnormal performance value and the traffic abnormality characterization degree and taking the average.

[0089] 403. Based on the correlation and fusion results of the overall monitoring data, determine the risk value of the current data.

[0090] Here, the risk value of the current data is obtained by dividing the correlation of the overall monitoring data by the fusion result. In some possible implementations, the monitoring correlation SR of the current data is calculated based on the monitoring correlation between the datasets of the time period data to obtain the risk value of the current data, as shown in formula (8):

[0091] (8);

[0092] Where: sigmoid() represents the activation function, used to map variables to the range (0, 1). Indicates time period The degree of characterization of traffic anomalies; The current data serves as the overall traffic anomaly characterization level; E represents the total abnormal performance value of the current monitoring data; S represents the correlation of the overall monitoring data; SR represents the correlation between the current monitoring data and the overall monitoring data, for example, by calculating the covariance between the current monitoring data and the overall monitoring data. c is a constant, set to 1 to prevent the denominator from being 0; DC represents the risk value of the current data, i.e., the degree of risk in the current signal transmission. Thus, by combining the correlation of the overall monitoring data for each time period with the total abnormal performance value and the traffic anomaly characterization level, the risk value of the current data can be accurately analyzed.

[0093] 106. Based on the risk value, update the key of the current data in the multiple time period data.

[0094] Here, by comparing a risk value with a preset risk threshold, if the risk value exceeds the preset risk threshold, the key of the current data is replaced to obtain an updated key; and the updated key is verified; if the updated key verification fails, the current data is marked and a prompt message is output. The prompt message can be in the form of voice prompts or text prompts. The preset risk threshold can be custom-set, for example, a preset risk threshold of 0.5. Thus, by setting a risk threshold Y=0.5, when the risk level exceeds the preset risk threshold, the communication risk is considered high, insecure, and at high risk of key leakage, requiring key replacement. Key replacement is performed according to a preset key table or using other methods, and then the replaced key is verified. If verification passes, the new key is considered to guarantee communication security. If verification fails, the current data is marked as abnormal, and a prompt message is output to the system for further processing by staff. In this way, emergency key replacement when the communication risk is high can improve communication security, thereby improving the security of privacy data in smart wearable devices.

[0095] In this embodiment of the invention, after acquiring the communication data of the smart wearable device, the reference data in the communication data is divided according to time sequence information to obtain multiple time period data. The traffic anomaly characterization degree corresponding to each of the multiple time period data is determined, so as to reflect whether the communication traffic of the communication data is abnormal. Then, the total monitoring anomaly performance value of the multiple time period data is determined based on the monitoring values ​​of the multiple time period data. Thus, by analyzing the total monitoring anomaly performance value, it is possible to accurately reflect whether the monitoring values ​​received by the smart wearable device are abnormal. Finally, based on the traffic anomaly characterization degree corresponding to the multiple time period data and the total monitoring anomaly performance value, the risk value of the multiple time period data is determined; and based on the risk value, the key of the current data is updated. In this way, by combining the monitoring values ​​in the historical communication data of the smart wearable device to calculate the total monitoring anomaly performance value of the time period data, the risk value of the data can be calculated more accurately. Timely key replacement based on the risk value can improve the security of the privacy data of the smart wearable device.

[0096] In some possible implementations, the segmentation module is further configured to divide the duration corresponding to the reference data according to the time sequence information to obtain multiple time periods; and to determine the data of the reference data within each of the multiple time periods to obtain the multiple time period data.

[0097] In some possible implementations, the first determining module is further configured to determine the statistical flow values ​​of the multiple time period data to obtain multiple statistical flow values; and based on the multiple statistical flow values, determine the flow anomaly characterization degree corresponding to the multiple time period data.

[0098] In some possible implementations, the first determining module is further configured to cluster the plurality of statistical traffic values ​​to obtain a plurality of traffic clusters; and based on the plurality of traffic clusters, determine the traffic anomaly characterization degree corresponding to the plurality of time period data.

[0099] In some possible implementations, the first determining module is further configured to, for any given time period data, determine the cluster center flow value and flow variance of the cluster to which the data belongs; determine the difference between the statistical flow value of the data and the cluster center flow value; and, based on the difference and the flow variance, determine the flow anomaly characterization degree corresponding to the data in any given time period.

[0100] In some possible implementations, the second determining module is further configured to use multiple monitoring sensors to collect initial monitoring values ​​of data for each time period, thereby obtaining multiple initial monitoring values ​​of data for each time period; and to determine the average of the multiple initial monitoring values ​​as the monitoring value of data for each time period.

[0101] In some possible implementations, the second determining module is further configured to cluster multiple monitoring values ​​corresponding to the data in each time period to obtain monitoring clusters of the data in each time period; determine the degree of fluctuation of the monitoring values ​​of the data in each time period based on the monitoring clusters corresponding to the data in each time period; determine the degree of affiliation of the data in each time period to the cluster based on the degree of affiliation; and determine the abnormal performance of the total monitoring values ​​of the data in each time period based on the degree of affiliation.

[0102] In some possible implementations, the third determining module is further configured to determine the correlation of the overall monitoring data of the multiple time periods based on the monitoring values ​​of the multiple time periods; fuse the total monitoring anomaly performance value and the traffic anomaly characterization degree to obtain a fusion result; and determine the risk value of the current data based on the correlation of the overall monitoring data and the fusion result.

[0103] In some possible implementations, the third determining module is further configured to determine the correlation of data changes and the difference in the amount of changes in the monitoring data among the multiple time periods based on the monitoring values ​​of the multiple time period data; determine the monitoring correlation among the multiple time period data based on the correlation of data changes and the difference in the amount of changes in the monitoring data, thereby obtaining multiple monitoring correlations; and determine the correlation of the overall monitoring data based on the multiple monitoring correlations.

[0104] In some possible implementations, the update module is further configured to: replace the key of the current data to obtain an updated key if the risk value is greater than a preset risk threshold; perform key verification on the updated key; and if the updated key verification fails, mark the current data and output a prompt message.

[0105] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous. The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. The above content is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for protecting privacy data based on computer networks, characterized in that, The method includes: Acquire communication data from smart wearable devices; According to the time sequence information, the reference data in the communication data is divided into multiple time period data; Determining the traffic anomaly characterization degree corresponding to the multiple time period data includes: determining the statistical traffic values ​​of the multiple time period data to obtain multiple statistical traffic values; clustering the multiple statistical traffic values ​​to obtain multiple traffic clusters; for any time period data, determining the cluster center traffic value and traffic variance of the traffic cluster to which the traffic data of any time period belongs; determining the difference between the statistical traffic value and the cluster center traffic value of the data of any time period; and determining the traffic anomaly characterization degree corresponding to the data of any time period based on the difference and the traffic variance. Based on the monitoring values ​​of the multiple time periods, the total abnormal monitoring value of the multiple time periods is determined, including: clustering the multiple monitoring values ​​corresponding to the data of each time period to obtain monitoring clusters of the data of each time period, each monitoring cluster representing a user's usage pattern; determining the fluctuation degree of the monitoring values ​​of the data of each time period based on the monitoring clusters corresponding to the data of each time period; determining the degree of belonging to the monitoring cluster to which the data of each time period belongs based on the fluctuation degree of the monitoring values ​​of the data of each time period; determining the total abnormal monitoring value of the data of each time period based on the degree of belonging; the monitoring value is the average of multiple initial monitoring values ​​of the data of each time period collected by multiple monitoring sensors; the total abnormal monitoring value is expressed as: Where: cow represents the number of times the monitored data attribution of the smart wearable device has changed; Indicates the time period in the current data The degree of affiliation with its monitoring cluster; L is the length of the current data, specifically the number of time periods included in the current data; E represents the total monitoring anomaly value of the current data; Norm represents the normalization function; Based on the traffic anomaly characterization degree and the total monitored anomaly performance value corresponding to the multiple time period data, the risk value of the current data in the multiple time period data is determined; Based on the risk value, update the key of the current data in the multiple time period data.

2. The privacy data protection method based on computer networks according to claim 1, characterized in that, The reference data in the communication data is divided according to the time sequence information to obtain multiple time period data, including: The reference data is divided into multiple time periods according to the time sequence information. The reference data is determined within each of the multiple time periods to obtain the multiple time period data.

3. The privacy data protection method based on computer networks according to claim 1, characterized in that, Before determining the total abnormal monitoring value of the multiple time period data based on the monitoring values ​​of the multiple time period data, the method further includes: Multiple monitoring sensors are used to collect initial monitoring values ​​of data for each time period, resulting in multiple initial monitoring values ​​for the data in each time period. The average of the multiple initial monitoring values ​​is determined as the monitoring value for each time period.

4. The privacy data protection method based on computer networks according to claim 1, characterized in that, The step of determining the risk value of the current data in the multiple time period data based on the traffic anomaly characterization degree corresponding to the multiple time period data and the total monitored anomaly performance value includes: Based on the monitoring values ​​of the data from the multiple time periods, determine the correlation of the overall monitoring data from the multiple time periods; The total monitored abnormal performance value and the flow abnormality characterization degree are fused to obtain the fusion result; Based on the correlation and fusion results of the overall monitoring data, the risk value of the current data is determined.

5. The privacy data protection method based on computer networks according to claim 4, characterized in that, Determining the correlation of the overall monitoring data based on the monitoring values ​​from the multiple time periods includes: Based on the monitoring values ​​of the data from the multiple time periods, determine the correlation of data changes among the multiple time periods and the differences in the amount of monitoring data changes; Based on the correlation of the data changes and the differences in the amount of changes in the monitored data, the monitoring correlation between the data in the multiple time periods is determined, and multiple monitoring correlations are obtained; Based on the multiple monitoring correlations, the correlation of the overall monitoring data is determined.

6. The privacy data protection method based on computer networks according to claim 1, characterized in that, The step of updating the key for the current data in the multiple time-period data based on the risk value includes: If the risk value is greater than a preset risk threshold, the key of the current data is replaced to obtain an updated key; Perform key verification on the updated key; If the updated key verification fails, the current data is marked and a prompt message is output.

Images