Multi-type data collection system and method based on Internet platform
By introducing technologies such as screening, encryption, privacy protection, gene expression profile analysis and artificial intelligence monitoring in the medical data collection system, the problems of data leakage and lack of analysis in the existing technology have been solved, and efficient and secure collection of medical data have been achieved. manage.
Patent Information
- Application Number
- CN202411346069.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-26
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-09-26
AI Technical Summary
The lack of encryption or security authentication measures in the collection and transmission of medical data in the prior art leads to a high risk of data leakage, no data processing or analysis, and lack of real-time monitoring and emergency response mechanisms.
Using a multi-class data acquisition system based on the Internet platform, through setting screening conditions and phased screening methods, combining real-time data stream processing and differential privacy protection, using improved homomorphic encryption algorithms and multi-factor authentication to ensure data security, and introducing gene expression Spectral analysis and multi-dimensional classification algorithms perform data prediction and storage, and use artificial intelligence to real-time monitoring and threat intelligence to achieve anomaly detection and automatic emergency response.
It improves the accuracy and security of medical data collection, reduces the risk of data leakage, ensures the security and management efficiency of data transmission, and promptly detects potential security threats, and automatically triggers emergency response measures.
Smart Images

Figure CN118861789B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of data processing, and more specifically to a multi-type data collection system and method based on an Internet platform. Background Art
[0002] With the rapid development of Internet technology and the growing demand for medical and health care, the traditional medical service model is gradually shifting to digitalization and intelligence. The emergence of hospital consultation platforms and health data management systems has not only improved the efficiency and quality of medical services, but also promoted the optimal allocation of medical resources.
[0003] In digital medicine, medical data resources mainly come from hospital information systems. How to effectively collect medical data information obtained by hospital front-end machines and realize rapid processing of hospital internal network information is the key to improving hospital efficiency and service quality. Medical and health big data refers to all data related to the health of the target group generated by medical institutions in medical activities, including but not limited to electronic medical record data, medical imaging data, wearable device data, and social media data.
[0004] With the advancement of technology and the rise of Internet medicine, data collection and analysis capabilities have been continuously improved, providing stronger support for clinical treatment and disease prediction. At the same time, we also need to pay attention to data security and privacy protection challenges to ensure the legal and compliant use of medical data.
[0005] For example, the patent with authorization announcement number CN204759420U discloses a mobile medical data collection and transmission device based on the Internet of Things, including: a GPS module for geographic information collection, an ARM host for data storage and integration, a 3G wireless communication module for data transmission, and a prenatal health monitoring device for medical data collection. The data collected by the GPS module and the prenatal health monitoring device are processed by the ARM host and transmitted to the hospital terminal system through the 3G wireless communication module to provide data for doctors' diagnosis and treatment.
[0006] The above existing technologies all have the following problems: 1) There is no mention of any encryption or security authentication measures, resulting in the risk of data leakage during data transmission; 2) The collected data is only transmitted to the hospital terminal system without any further data processing or analysis; 3) There is no mechanism for real-time monitoring or emergency response. Summary of the invention
[0007] In view of the shortcomings of the prior art, the present invention proposes a multi-category data collection system and method based on the Internet platform. By setting screening conditions, combining a phased screening method and real-time data stream processing technology, multi-category data in the Internet platform are screened and preprocessed, and differential privacy protection is implemented to ensure the data security of the target group; an improved homomorphic encryption algorithm is used to encrypt and process multi-category data, and a multi-factor authentication mechanism is combined to ensure data transmission security; the data center decrypts the encrypted data after receiving it, introduces a gene expression spectrum analysis strategy to predict the disease risk and drug status response of the target group, and classifies and stores the data through a multi-dimensional classification algorithm; the system uses artificial intelligence real-time monitoring and threat intelligence integration to achieve anomaly detection and automatic emergency response, thereby improving the accuracy and security of multi-category data collection.
[0008] To achieve the above object, the present invention provides the following technical solutions:
[0009] Various data collection methods based on the Internet platform include:
[0010] Step S1: Set screening conditions, use a phased screening method to screen multiple types of data from the Internet platform according to the screening conditions, introduce a real-time data stream processing method, pre-process the screened multiple types of data, and perform differential privacy protection on the pre-processed multiple types of data;
[0011] Step S2: Use an improved homomorphic encryption algorithm to encrypt the multi-category data protected by differential privacy, and combine multi-factor authentication to authenticate the identities of the entities involved in the transmission of the multi-category data;
[0012] Step S3: The data center decrypts the received encrypted multi-category data to obtain the decrypted multi-category data, introduces a gene expression spectrum analysis strategy, predicts the state type of the target group based on the decrypted multi-category data, and uses a multidimensional classification algorithm to classify and store the decrypted multi-category data;
[0013] Step S4: Use artificial intelligence methods to conduct real-time monitoring and integrate external threat intelligence sources. If there is abnormal process data, it is judged as a hacker attack or a leak, and the emergency response mechanism is automatically triggered;
[0014] The formula for differential privacy protection in step S1 is:
[0015] y = P [ f ( D , t , i ) + L a p ( max D , D ′ | f ( D , t , i ) − f ( D ′ , t , i ) | e ( α ) ) ] ≤ e x p ( e ) × P [ f ( D ′ , t , i ) + L a p ( max D , D ′ | f ( D , t , i ) − f ( D ′ , t , i ) | e ( α ) ) ] ;
[0016] Among them, y represents the multi-class data output after differential privacy protection, D and Represents any two adjacent multi-class data, represents the query function of multi-class data D, Representing multi-class data The query function, Represents multi-class data D and The global sensitivity between represents random noise drawn from a Laplace distribution with scale parameter *, represents the probability function, t represents the time parameter, Indicates the query parameters. represents the privacy budget function, represents the privacy loss parameter, represents the exponential function, Indicates loss of privacy;
[0017] The formula of the improved homomorphic encryption algorithm in step S2 is:
[0018] ;
[0019] in, Represents multiple types of data after homomorphic encryption, represents a parameterized function, represents encryption operation, represents the nth multi-class data after differential privacy protection, n represents the number of multi-class data after differential privacy protection, , , Respectively express The key, algorithm, and context parameters used when performing encryption;
[0020] The specific steps of step S3 include:
[0021] S301: The data center receives the encrypted multi-category data and decrypts the encrypted multi-category data to obtain decrypted multi-category data. The decryption formula is:
[0022] ;
[0023] in, represents the decryption operation, b, c, and d represent the key, algorithm, and context parameters used in decryption, respectively;
[0024] S302: preprocessing the decrypted multi-class data, and constructing a gene expression profile for each decrypted multi-class data according to the preprocessed decrypted multi-class data;
[0025] The specific steps of step S3 also include:
[0026] S303: extracting characteristic genes of the state type from the gene expression spectrum using the PCA feature extraction algorithm, building a random forest prediction model based on the characteristic genes and the existing medical knowledge base, and inputting the decrypted multi-class data into the trained random forest prediction model to predict the state type of the target group;
[0027] S304: Classify the multi-category data into m categories according to different health indicators, disease types and treatment strategies, classify the multi-category data using a multi-dimensional classification algorithm, and store the classified multi-category data in a database;
[0028] The specific steps of step S4 include:
[0029] S401: using AI methods to collect process data in real time, and preprocessing the collected process data;
[0030] S402: Integrate the preprocessed process data with the external threat intelligence data, and use the PCA algorithm to extract features;
[0031] S403: Obtain historical process data, combine known network security events and hacker attack cases, train AI anomaly detection models, and deploy the trained AI anomaly detection models to monitor the Internet platform in real time. At the same time, continuously integrate the latest external threat intelligence data;
[0032] S404: The AI anomaly detection model identifies abnormal process data based on preset thresholds and rules, and combines external threat intelligence data to determine whether the anomaly is caused by a hacker attack or leak. If a hacker attack or leak is identified, the emergency response mechanism is automatically triggered.
[0033] A variety of data collection systems based on the Internet platform, including: data collection module, encryption transmission module, data analysis module, emergency response module, and interaction module;
[0034] The data collection module is used to collect multiple types of data of the target group from the Internet platform;
[0035] The encryption transmission module is used to encrypt multiple types of data during the data transmission process;
[0036] The data analysis module is used to analyze and mine the encrypted multiple types of data and extract feature information;
[0037] The emergency response module is used to formulate response measures when an emergency occurs based on the feature information;
[0038] The interaction module is used to provide a user interface, and the user interface includes countermeasures.
[0039] Specifically, the data acquisition module includes: a data screening unit, a preprocessing unit, and a differential privacy protection unit;
[0040] The data screening unit is used to screen multiple types of data from the Internet platform using a phased screening method;
[0041] The preprocessing unit is used to preprocess the screened multiple types of data;
[0042] The differential privacy protection unit is used to protect the privacy of users during data screening and preprocessing.
[0043] Specifically, the data analysis module includes: a data decryption unit, a classification storage unit, and a data analysis unit;
[0044] The data decryption unit is used to decrypt and restore the received encrypted multiple types of data;
[0045] The classification storage unit is used to store the decrypted multiple types of data in a specific format and structure;
[0046] The data analysis unit is used to use a gene expression spectrum analysis method to predict the state type of the target group according to the decrypted multi-category data, and use a multi-dimensional classification method to classify and store the restored multi-category data.
[0047] Specifically, an electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a multi-category data collection method based on an Internet platform when executing the computer program.
[0048] Compared with the prior art, the present invention has the following beneficial effects:
[0049] 1. The present invention proposes a multi-type data collection system based on the Internet platform, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.
[0050] 2. The present invention proposes a multi-category data collection method based on the Internet platform. On the one hand, the accuracy and timeliness of multi-category data are ensured by using phased screening and real-time data stream processing methods. The differential privacy protection technology can allow statistics and analysis while ensuring data privacy, thereby reducing the risk of data leakage of the target group. On the other hand, the improved homomorphic encryption algorithm provides a strong security guarantee for the encrypted multi-category data, making the data not easily cracked during transmission. Multi-factor authentication ensures the identity authenticity of each entity involved in data transmission, further enhancing the security of data transmission.
[0051] 3. The present invention proposes a multi-category data collection method based on the Internet platform. After the data center restores the original multi-category data through decryption processing, a gene expression spectrum analysis strategy is introduced to more accurately predict the status type of the target group; the multi-dimensional classification algorithm enables multi-category data to be classified and stored according to different dimensions, thereby improving the efficiency and convenience of data management; the Internet platform is monitored in real time using artificial intelligence methods to promptly detect potential security threats; the integration of external threat intelligence sources enables the system to have a more comprehensive understanding of the current network security situation; once abnormal process data is found, the system can automatically trigger the emergency response mechanism, take countermeasures quickly, and reduce security risks. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] Figure 1 It is a schematic diagram of a multi-category data collection method based on an Internet platform of the present invention;
[0053] Figure 2 This is a flow chart of the working principle of the multi-category data collection method based on the Internet platform of the present invention;
[0054] Figure 3 This is a flow chart of encryption and decryption of multiple types of data in the multiple types of data collection method based on the Internet platform of the present invention;
[0055] Figure 4 This is an architecture diagram of the multi-type data collection system based on the Internet platform of the present invention. DETAILED DESCRIPTION
[0056] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.
[0057] Example 1
[0058] See also Figure 4 , an embodiment provided by the present invention: a multi-type data collection system based on an Internet platform, whose application fields include but are not limited to a hospital consultation platform, and the multi-type data includes health data collected through the hospital consultation platform, including: a data collection module, an encryption transmission module, a data analysis module, an emergency response module, and an interaction module;
[0059] The data collection module is used to collect multiple types of data of the target group from the Internet platform, including physiological parameter data collected in real time by medical sensor devices, such as blood pressure, blood sugar, and heart rate; image data of the target group obtained from medical imaging equipment; and electronic medical record data extracted from the hospital information system;
[0060] The encryption transmission module is used to encrypt multiple types of data during the data transmission process to ensure the security and confidentiality of the data. When data is transmitted within the hospital or between doctors and target groups, the use of encryption methods can effectively prevent the data from being illegally obtained or tampered with;
[0061] The data analysis module is used to analyze and mine encrypted multi-category data and extract characteristic information of state reactions to discover early symptoms of diseases, predict disease risks, and evaluate drug responses;
[0062] The emergency response module is used to formulate response measures based on feature information when an emergency occurs, such as an emergency or a sudden serious illness in the target group, including collecting and analyzing relevant information to provide decision makers with timely and accurate intelligence support to formulate response measures;
[0063] The interactive module is used to provide a user interface, which includes countermeasures, including displaying multiple types of data of the target group, analyzing results, providing online consultation services, and managing user permissions.
[0064] The data collection module includes: data screening unit, preprocessing unit, and differential privacy protection unit;
[0065] A data screening unit for screening multiple categories of data from Internet platforms using a phased screening approach;
[0066] The preprocessing unit is used to preprocess the selected data, including data cleaning, standardization and privacy protection;
[0067] Differential privacy protection unit, used to protect user privacy during data screening and preprocessing.
[0068] The encryption transmission module includes: a data encryption unit, an identity authentication unit, and a data transmission unit;
[0069] The data encryption unit is used to encrypt the pre-processed multiple types of data using an improved homomorphic encryption algorithm.
[0070] The identity authentication unit is used to authenticate the identities of various entities involved in data transmission in combination with digital certificates.
[0071] The data transmission unit is responsible for transmitting encrypted multiple types of data to the data center through a secure communication channel.
[0072] The data analysis module includes: a data decryption unit, a classification storage unit, and a data analysis unit;
[0073] A data decryption unit, used to decrypt and restore the received encrypted multiple types of data;
[0074] A classification storage unit is used to store the decrypted multiple types of data in a specific format and structure;
[0075] The data analysis unit is used to use the gene expression spectrum analysis method to predict the state type of the target group according to the decrypted multi-category data, and use the multi-dimensional classification method to classify and store the restored multi-category data.
[0076] The emergency response module includes: security monitoring unit and emergency response unit;
[0077] Safety monitoring unit, used to monitor the system's operating status in real time and detect abnormal process data;
[0078] The emergency response unit is used to automatically trigger the emergency response mechanism when a security threat is detected, including isolating the infected system, notifying the security team, and initiating backup and recovery procedures.
[0079] Example 2
[0080] See also Figure 1-Figure 3 Another embodiment provided by the present invention is a multi-type data collection method based on an Internet platform, comprising:
[0081] Step S1: Set screening conditions, use a phased screening method to screen multiple types of data from the Internet platform according to the screening conditions, introduce a real-time data stream processing method, pre-process the screened multiple types of data, and perform differential privacy protection on the pre-processed multiple types of data;
[0082] In the present invention, the multi-category data refers to the health data of the target group collected. The phased screening method is to divide the screening process into multiple stages according to the disease type and time range, and apply different screening conditions to each stage. At the same time, the real-time data stream processing framework of Apache Flink is selected to configure the data stream processing logic to capture the multi-category data generated by the Internet platform in real time, and then apply the screening conditions to the real-time data stream to screen out the multi-category data that meets the conditions.
[0083] Furthermore, the screening criteria include:
[0084] (1) Data types: clearly identify the multiple types of data that need to be screened, such as medical records, test results, and prescription information;
[0085] (2) Time range: set the time range of the data, such as data from the last week, month, or a specific time period;
[0086] (3) Target group information: screening based on the target group’s age, gender, and region;
[0087] (4) Disease category: Screening for specific diseases or disease categories, such as diabetes and hypertension.
[0088] The present invention introduces a real-time data stream processing method to access the data sources of the Internet platform in real time, such as electronic medical records and smart medical equipment data, and uses the Apache Flink data stream processing framework for data cleaning, filtering, and aggregation processing to ensure the real-time and accuracy of the data.
[0089] Step S2: Use an improved homomorphic encryption algorithm to encrypt the multi-category data protected by differential privacy, and combine multi-factor authentication to authenticate the identities of the entities involved in the transmission of the multi-category data;
[0090] Furthermore, the specific steps of step S2 include:
[0091] (1) Receive multiple types of data protected by differential privacy and encrypt them using an improved homomorphic encryption algorithm;
[0092] (2) Select a multi-factor authentication process that combines ownership factors with biometric factors and deploys it in the system of the Internet platform to ensure that all entities involved in data transmission are authenticated to improve the security of data transmission.
[0093] The present invention adopts a differential privacy method to protect data privacy. By adding random noise to the original multi-category data, data perturbation is achieved, thereby preventing the leakage of sensitive information.
[0094] Step S3: The data center decrypts the received encrypted multi-category data to obtain the decrypted multi-category data, introduces a gene expression spectrum analysis strategy, predicts the state type of the target group based on the decrypted multi-category data, and uses a multidimensional classification algorithm to classify and store the decrypted multi-category data;
[0095] Among them, the status types of the target group include: the status reaction of the target group after using the drug, and the disease risk of the target group.
[0096] Step S4: Use artificial intelligence methods to conduct real-time monitoring and integrate external threat intelligence sources. If there is abnormal process data, it is judged as a hacker attack or a leak, and the emergency response mechanism is automatically triggered. Among them, external threat intelligence sources include network security intelligence platforms and threat information sharing organizations.
[0097] Furthermore, the emergency response mechanism includes:
[0098] (1) Monitoring and early warning: Use artificial intelligence methods to monitor the network environment and system status in real time, integrate external threat intelligence sources, and promptly detect and warn of potential network security threats;
[0099] (2) Event confirmation and assessment: When abnormal process data is monitored, it is necessary to further confirm whether a security incident such as a hacker attack or data leakage has actually occurred, and to assess the nature, scope, and impact of the incident;
[0100] (3) Emergency response initiation: Once a security incident is confirmed, the emergency response mechanism will be automatically or manually triggered to initiate the corresponding emergency plan;
[0101] (4) Emergency response: According to the emergency plan, quickly organize relevant personnel to carry out emergency response, including isolating the affected systems, restoring damaged data, and tracing the source of the attack;
[0102] (5) Summary and Improvement: After the emergency response is completed, the incident is summarized and analyzed, the effectiveness and shortcomings of the emergency response mechanism are evaluated, and improvement measures are proposed to improve the emergency plan and emergency response mechanism.
[0103] The formula for differential privacy protection in step S1 is:
[0104] y = P [ f ( D , t , i ) + L a p ( max D , D ′ | f ( D , t , i ) − f ( D ′ , t , i ) | e ( α ) ) ] ≤ e x p ( e ) × P [ f ( D ′ , t , i ) + L a p ( max D , D ′ | f ( D , t , i ) − f ( D ′ , t , i ) | e ( α ) ) ] ;
[0105] Among them, y represents the multi-class data output after differential privacy protection, D and It means that there is only one record different between any two adjacent multi-class data. represents the query function of multi-class data D, Representing multi-class data The query function receives multiple types of data and query parameters and time parameter t as input, and returns the query result, Represents multi-class data D and The global sensitivity between represents random noise drawn from a Laplace distribution with scale parameter *, represents the probability function, t represents the time parameter, Indicates query parameters, including query type and query conditions. represents the exponential function, Indicates privacy loss, represents the privacy budget function, represents the privacy loss parameter, and Control the degree of privacy protection and the amount of noise, The smaller it is, the louder the noise is and the stronger the privacy protection is; The larger it is, the less noise there is and the weaker the privacy protection is.
[0106] The formula of the improved homomorphic encryption algorithm in step S2 is:
[0107] ;
[0108] in, Represents multiple types of data after homomorphic encryption, represents a parameterized function, represents encryption operation, represents the nth multi-class data after differential privacy protection, n represents the number of multi-class data after differential privacy protection, , , Respectively express The key, algorithm, and context parameters used when performing encryption.
[0109] The specific steps of step S3 include:
[0110] S301: The data center receives the encrypted multi-category data and decrypts the encrypted multi-category data to obtain decrypted multi-category data. The decryption formula is:
[0111] ;
[0112] in, represents the decryption operation, b, c, and d represent the key, algorithm, and context parameters used in decryption, respectively;
[0113] S302: preprocessing the decrypted multi-class data, and constructing a gene expression profile for each decrypted multi-class data according to the preprocessed decrypted multi-class data, wherein the gene expression profile refers to a list or matrix containing all gene expression levels.
[0114] The specific steps of step S3 also include:
[0115] S303: extracting characteristic genes of the state type from the gene expression spectrum using the PCA feature extraction algorithm, constructing a random forest prediction model based on the characteristic genes and the existing medical knowledge base, and inputting the decrypted multi-class data into the trained random forest prediction model to predict the state type of the target group, wherein the PCA feature extraction algorithm is the prior art content in this field, is not the inventive solution of the present application, and is not described in detail here;
[0116] S304: According to different health indicators, disease types and treatment strategies, the multi-category data is divided into m categories, and the multi-category data is classified using a multi-dimensional classification algorithm, and the classified multi-category data is stored in a database.
[0117] The specific steps of step S4 include:
[0118] S401: using AI methods to collect process data in real time, and preprocessing the collected process data, wherein the preprocessing includes cleaning, standardizing and formatting the collected process data, wherein the process data includes log records and process data of the Internet platform system;
[0119] S402: Integrate the preprocessed process data with the external threat intelligence data, and use the PCA algorithm to extract features;
[0120] S403: Obtain historical process data, combine known network security events and hacker attack cases, train AI anomaly detection models, and deploy the trained AI anomaly detection models to monitor the Internet platform in real time. At the same time, continuously integrate the latest external threat intelligence data;
[0121] In the present invention, the anomaly detection model selects a convolutional neural network model, extracts principal component features from historical process data, known network security events, and hacker attack case data, and uses the extracted feature data to train the convolutional neural network model to obtain a trained convolutional neural network model, i.e., an AI anomaly detection model, wherein the specific training process and principal component feature extraction formula of the convolutional neural network model are prior art contents in the art, and are not the inventive solution of the present application, and are not described in detail herein;
[0122] For example, assuming that it is necessary to perform anomaly detection on the network traffic of an Internet platform to identify potential DDoS attacks or malware propagation, key information in the network traffic data packet, such as IP address, port number, packet size, and timestamp, can be converted into a two-dimensional matrix or grayscale image, and then trained and analyzed using a convolutional neural network model. The specific steps include:
[0123] (1) Extract key information from network traffic packets and convert this information into a two-dimensional matrix or grayscale image for processing by the convolutional neural network model;
[0124] (2) Use normal traffic in historical process data and known network security event data, such as DDoS attacks and malware propagation, as training sets and train a convolutional neural network model to enable it to identify abnormal behaviors that do not match normal traffic patterns;
[0125] (3) Deploy the trained convolutional neural network model to the monitoring system of the Internet platform, collect network traffic data in real time, and convert it into a format that can be processed by the convolutional neural network model;
[0126] (4) Use convolutional neural network models to detect anomalies in real-time network traffic data. When abnormal behavior is detected, an alarm is immediately triggered and corresponding defensive measures are taken;
[0127] (5) Collect and integrate the latest external threat intelligence data, and update the training set and detection rules of the convolutional neural network model based on the new threat intelligence to improve the accuracy and adaptability of the convolutional neural network.
[0128] S404: The AI anomaly detection model identifies abnormal process data based on preset thresholds and rules, and combines external threat intelligence data to determine whether the anomaly is caused by a hacker attack or leak. If a hacker attack or leak is identified, the emergency response mechanism is automatically triggered, including notifying the security team, isolating the infected system, and initiating backup recovery.
[0129] Example 3
[0130] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of a multi-category data collection method based on an Internet platform when executing the computer program.
[0131] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.
Claims
1. Multiple data collection methods based on the Internet platform, It is characterized in that include: Step S1: Set screening conditions, use a phased screening method to screen multiple types of data from the Internet platform according to the screening conditions, introduce a real-time data stream processing method, pre-process the screened multiple types of data, and perform differential privacy protection on the pre-processed multiple types of data; Step S2: Use an improved homomorphic encryption algorithm to encrypt the multi-category data protected by differential privacy, and combine multi-factor authentication to authenticate the identities of the entities involved in the transmission of the multi-category data; Step S3: The data center decrypts the received encrypted multi-category data to obtain the decrypted multi-category data, introduces a gene expression spectrum analysis strategy, predicts the state type of the target group based on the decrypted multi-category data, and uses a multidimensional classification algorithm to classify and store the decrypted multi-category data; Step S4: Use artificial intelligence methods to conduct real-time monitoring and integrate external threat intelligence sources. If there is abnormal process data, it is judged as a hacker attack or a leak, and the emergency response mechanism is automatically triggered; The formula for differential privacy protection in step S1 is: ; Among them, y represents the multi-class data output after differential privacy protection, D and Represents any two adjacent multi-class data, represents the query function for multi-class data D, Representing multi-class data The query function, Represents multi-class data D and The global sensitivity between represents random noise drawn from a Laplace distribution with scale parameter *, represents the probability function, t represents the time parameter, Indicates the query parameters. represents the privacy budget function, represents the privacy loss parameter, represents the exponential function, Indicates loss of privacy; The formula of the improved homomorphic encryption algorithm in step S2 is: ; in, Represents multiple types of data after homomorphic encryption, represents a parameterized function, represents encryption operation, represents the nth multi-class data after differential privacy protection, n represents the number of multi-class data after differential privacy protection, , , Respectively express The key, algorithm, and context parameters used when performing encryption; The specific steps of step S3 include: S301: The data center receives the encrypted multi-category data and decrypts the encrypted multi-category data to obtain decrypted multi-category data. The decryption formula is: ; in, represents the decryption operation, b, c, and d represent the key, algorithm, and context parameters used in decryption, respectively; S302: preprocessing the decrypted multi-class data, and constructing a gene expression profile for each decrypted multi-class data according to the preprocessed decrypted multi-class data; The specific steps of step S3 also include: S303: extracting characteristic genes of the state type from the gene expression spectrum using the PCA feature extraction algorithm, building a random forest prediction model based on the characteristic genes and the existing medical knowledge base, and inputting the decrypted multi-class data into the trained random forest prediction model to predict the state type of the target group; S304: Classify the multi-category data into m categories according to different health indicators, disease types and treatment strategies, classify the multi-category data using a multi-dimensional classification algorithm, and store the classified multi-category data in a database; The specific steps of step S4 include: S401: using AI methods to collect process data in real time, and preprocessing the collected process data; S402: Integrate the preprocessed process data with the external threat intelligence data, and use the PCA algorithm to extract features; S403: Obtain historical process data, combine known network security events and hacker attack cases, train AI anomaly detection models, and deploy the trained AI anomaly detection models to monitor the Internet platform in real time. At the same time, continuously integrate the latest external threat intelligence data; S404: The AI anomaly detection model identifies abnormal process data based on preset thresholds and rules, and combines external threat intelligence data to determine whether the anomaly is caused by a hacker attack or leak. If a hacker attack or leak is identified, the emergency response mechanism is automatically triggered.
2. A multi-category data collection system based on an Internet platform, used to implement the multi-category data collection method based on an Internet platform as claimed in claim 1, It is characterized in that include: Data acquisition module, encryption transmission module, data analysis module, emergency response module, interaction module; The data collection module is used to collect multiple types of data of the target group from the Internet platform; The encryption transmission module is used to encrypt multiple types of data during the data transmission process; The data analysis module is used to analyze and mine the encrypted multiple types of data and extract feature information; The emergency response module is used to formulate response measures when an emergency occurs based on the feature information; The interaction module is used to provide a user interface, and the user interface includes countermeasures.
3. The multi-category data collection system based on the Internet platform as claimed in claim 2, It is characterized in that The data acquisition module includes: a data screening unit, a preprocessing unit, and a differential privacy protection unit; The data screening unit is used to screen multiple types of data from the Internet platform using a phased screening method; The preprocessing unit is used to preprocess the screened multiple types of data; The differential privacy protection unit is used to protect the privacy of users during data screening and preprocessing.
4. The multi-category data collection system based on the Internet platform as claimed in claim 3, It is characterized in that The data analysis module includes: a data decryption unit, a classification storage unit, and a data analysis unit; The data decryption unit is used to decrypt and restore the received encrypted multiple types of data; The classification storage unit is used to store the decrypted multiple types of data in a specific format and structure; The data analysis unit is used to use a gene expression spectrum analysis method to predict the state type of the target group according to the decrypted multi-category data, and to classify and store the restored multi-category data using a multi-dimensional classification method.
5. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, It is characterized in that When the processor executes the computer program, the steps of the multi-category data collection method based on the Internet platform described in claim 1 are implemented.
Citation Information
Patent Citations
It gathers and transmission device to remove medical data based on thing networking
CN204759420U
Crowdsourcing platform statistical data on-demand sharing method based on block chain and differential privacy
CN117216786A
Data security transmission system
CN117319030A
