User data supervision system and method based on big data analysis

Abstract

The invention discloses a user data supervision system and method based on big data analysis, and relates to the technical field of big data. Historical service data of a user is called based on an application service database; classifying user historical service data, and constructing a corresponding type of data set; respectively carrying out data integration analysis, determining set center data and constructing data links of the same type; performing different-type data association analysis on the same-type data links, constructing different-type data links, performing data state analysis, and performing comprehensive state analysis on user historical service data; fusing the real-time user service request data with the historical service data, generating a real-time different-type data link, performing state analysis, and performing state offset analysis on the real-time different-type data link and the historical service data; outputting a data request behavior state based on the real-time user service request data; and updating the historical service data state of the user and performing data updating on the database.

Description

Technical Field

[0001] The present invention relates to the field of big data, and in particular to a user data supervision system and method based on big data analysis. Background Art

[0002] User data supervision based on big data analysis is to monitor, analyze and process user data in real time through advanced data mining, machine learning, pattern recognition and statistical analysis technologies; it aims to identify and prevent data anomalies and ensure the security and compliance of user data; in the current environment, most application software databases lack flexibility in data management, and only implement data storage and call for data in the database, lack linkage for the correlation between data, which reduces the practical value of database data; and most current applications lack real-time dynamic analysis of abnormal data monitoring and real-time user operations, and only conduct risk analysis by setting thresholds and sensitive operation perception, and data analysis lacks sensitivity. Summary of the invention

[0003] The purpose of the present invention is to provide a user data supervision system and method based on big data analysis to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A user data supervision method based on big data analysis, the method comprising the following steps:

[0006] S100, searching user information based on the application service database, and calling the user's historical service data according to the search results; generating a temporary service data space in the cloud based on the user's service request status, and storing the user's historical service data and real-time service request data;

[0007] S200, based on the temporary service data space in the cloud, classify the corresponding user historical service data according to the data attribute type, and construct a corresponding type of data set; perform data integration analysis on each type of data set respectively, and determine the central data of each type of data set; construct a data chain of the same type based on the central data of the corresponding type of data set; perform heterogeneous data association analysis based on the contract type data chain of each type of data set, and construct a heterogeneous data chain;

[0008] S300, performing data status analysis on various data links of different types, and performing comprehensive status analysis on user historical service data based on the analysis data; generating corresponding real-time heterogeneous data links by fusing real-time user service request data with corresponding data types of user historical service data; performing status analysis on real-time heterogeneous data links, and performing status offset analysis on real-time heterogeneous data links and user historical service data;

[0009] S400: Output data request behavior status based on real-time user service request data; update user historical service data status and update the database.

[0010] The specific steps of S100 are as follows:

[0011] S101, based on the application service database, using user registration information as a data search keyword to retrieve user historical service data; the user registration information is user identity information;

[0012] S102. By determining the connection status between the application server and the user, a temporary service data space is generated in the cloud when there is a data connection between the user and the application server to store the user's historical service data and real-time request data; when there is no connection between the user and the application server, a temporary service data space is generated.

[0013] The specific steps of S200 are as follows:

[0014] S201. In the corresponding user cloud temporary service data space, the user's historical service data is classified by preset data attribute types, and a corresponding type of data set is constructed based on the classification results; based on each type of data set, the central data of each type of data set is determined through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of the corresponding data in each type of data set, and analyzing the centralization degree of each data, and the calculation formula is:

[0015]

[0016] Among them, S(n,m) is the degree of centralization of the data with label m in the data set corresponding to type number n; q(n,m) is the word frequency of the data with label m in the data set corresponding to type number n; γ(n,m) is the value influence parameter of the data with label m in the data set corresponding to type number n; q(n) is the total word frequency of the data in the data set corresponding to type number n;

[0017] Based on the analysis of the degree of centralization of each data in each type of data set, the data with the maximum degree of centralization in each type of data set is selected as the central data, and the central data is used as the head of the data chain, and the data is sorted in descending order from large to small to construct the same type of data chain of the corresponding type of data set;

[0018] S202, combining each type of data set with the corresponding data chain of the same type to perform association analysis between different data types, the calculation formula is:

[0019]

[0020] Where R[(n,m):(n o,m)] is the data with label m in the data set corresponding to type number n, and the data with label m in the data set corresponding to type number n o The correlation index between the data labeled m in the corresponding data set; is the feature vector of the data numbered m in the data set corresponding to type number n; is the corresponding type number n o The feature vector corresponding to the data numbered m in the data set; where n o is the number of any data set except the data set numbered n; the characteristic vector of the data is constructed by the centralization degree and value influence parameter of the corresponding data; the above data correlation analysis is performed by performing similarity analysis on each data in any data set and each data in the other data sets;

[0021] Based on the association analysis data, set the data association judgment parameter R 0 , judge the correlation index between each data, if R[(n,m):(n o ,m)]≥R 0 , then it is judged that there is a correlation between the two data; if R[(n,m):(n o ,m)]<R 0 , it is determined that there is no association between the two data; the data with association between different types are connected in series, and the heterogeneous data chain between the different types of data is obtained based on the connection result.

[0022] The specific steps of S300 are as follows:

[0023] S301. Coordinate and build a corresponding heterogeneous data link set for each heterogeneous data link, and perform status evaluation and analysis on each heterogeneous data link in the set. The calculation formula is:

[0024]

[0025] Wherein, Ad(i) is the security risk index of the heterogeneous data link numbered i in the corresponding heterogeneous data link set; Y(i,u) is the value loss index of the data numbered u in the corresponding heterogeneous data link numbered i; L(i,u) is the type impact weight of the data numbered u in the corresponding heterogeneous data link numbered i; γ(i,u) is the value impact parameter of the data numbered u in the corresponding heterogeneous data link numbered i; S(i,u) is the degree of centralization of the data numbered u in the corresponding heterogeneous data link numbered i; wherein the data value loss index is a reference index of the interval in which the degree of centralization of the data is located in the corresponding data value loss form; the data value loss form is a reference form that corresponds the degree of centralization of the data to the corresponding data loss index;

[0026] Based on the status analysis data of each type of data link, the status is judged; the historical security risk average index Ad(v) is quoted. If Ad(i)≥Ad(v), the corresponding heterogeneous data link data is judged to be normal; if Ad(i)<Ad(v), the corresponding heterogeneous data link data is judged to be abnormal; if there is no abnormal heterogeneous data link, a comprehensive status assessment is performed on the heterogeneous data link set, and the calculation formula is:

[0027]

[0028] Where Ad(al) is the comprehensive security risk index of the corresponding heterogeneous data link set;

[0029] If there are abnormal heterogeneous data links, the corresponding data of the current heterogeneous data links will be marked and removed from the user's historical service data; after the removal operation is completed, a comprehensive status assessment will be conducted on the heterogeneous data link set;

[0030] S302, retrieve the corresponding data types of real-time user service request data and user historical service data for fusion processing; divide and classify the real-time user service request data according to the data attribute type, and complete the serial connection of the same type of data chain to the corresponding type of the same type of data chain after data centralization analysis; associate and complete the serial construction of the newly generated heterogeneous data chain through heterogeneous data; analyze the security risk index of the newly generated heterogeneous data chain and make a status judgment; if the judgment is abnormal, give an abnormal prompt for the current user service request data; if the judgment is normal, include the newly generated heterogeneous data chain into the heterogeneous data chain set and conduct a comprehensive status evaluation of the new heterogeneous data chain set to obtain the comprehensive security risk index of the new heterogeneous data chain set; perform a status offset analysis on the user historical service data based on the comprehensive security risk index of the new heterogeneous data chain set, and the calculation formula is:

[0031]

[0032] Among them, DF(e) is the state deviation index of the user historical service data of the newly generated heterogeneous data link numbered e; Ad(el) is the comprehensive security risk index of the new heterogeneous data link set; Ad(e) corresponds to the security risk index of the newly generated heterogeneous data link numbered e; Among them, if there are multiple newly generated heterogeneous data links, the state deviation analysis is performed separately;

[0033] Based on the comprehensive security risk index of the new heterogeneous data link set, the state deviation analysis data of the user historical service data is performed, and the restriction parameter DF(th) is introduced to make an abnormal judgment on the state deviation of the newly generated heterogeneous data link to the user historical service data; if DF(e)>DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is abnormal, and it is prompted that there is an abnormal risk in the current user service request data operation; if DF(e)≤DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is normal.

[0034] The specific steps of S400 are as follows:

[0035] S401, outputting the user's historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, storing the data in the cloud temporary service data space into the application service database, completing the database update and closing the cloud temporary service data space;

[0036] S402: Issue warnings for abnormal historical service data, abnormal real-time user service request data, and abnormal risk real-time user service request operations.

[0037] A user data supervision system based on big data analysis, the system comprising a data retrieval module, a multi-dimensional data processing module, a real-time data status analysis module and a data feedback module;

[0038] The data retrieval module searches for user information based on the application service database, and calls the user's historical service data according to the search results; based on the user service request status, a temporary service data space is generated on the cloud, and the user's historical service data and real-time service request data are stored; the multi-dimensional data processing module classifies the corresponding user's historical service data according to the data attribute type in the temporary service data space on the cloud, and constructs a corresponding type of data set; performs data integration analysis on each type of data set, and determines the central data of each type of data set; constructs a data chain of the same type based on the central data of the corresponding type of data set; and constructs a data chain of the same type based on the contract type of each type of data set. Perform heterogeneous data association analysis and construct heterogeneous data chains; the real-time data status analysis module performs data status analysis on each heterogeneous data chain, and performs comprehensive status analysis on user historical service data based on the analysis data; generates corresponding real-time heterogeneous data chains by fusing real-time user service request data with corresponding data types of user historical service data; performs status analysis on real-time heterogeneous data chains, and performs status offset analysis on real-time heterogeneous data chains and user historical service data; the data feedback module outputs data request behavior status based on real-time user service request data; updates the status of user historical service data and updates the database.

[0039] The data retrieval module includes a historical data retrieval unit and a historical data retrieval unit;

[0040] The historical data retrieval unit retrieves the user's historical service data based on the application service database, using the user registration information as a data retrieval keyword;

[0041] The historical data retrieval unit determines the connection status between the application server and the user, and generates a temporary cloud service data space when there is a data connection between the user and the application server to store the user's historical service data and real-time request data.

[0042] The multidimensional data processing module includes a data centralization analysis unit and a data association analysis unit;

[0043] The data centralization analysis unit classifies the user's historical service data in the corresponding user's cloud temporary service data space by preset data attribute types, and constructs a corresponding type of data set based on the classification result; based on each type of data set, determines the central data of each type of data set through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of each data corresponding to each type of data set, and analyzing the centralization degree of each data; based on the centralization degree analysis of each data in each type of data set, selects the corresponding data with the maximum centralization degree in each type of data set as the central data, and takes the central data as the head of the data chain, sorts the data in descending order from large to small, and constructs a data chain of the same type of the corresponding type of data set;

[0044] The data association analysis unit performs association analysis between different types of data in combination with the same type of data chains corresponding to each type of data set; sets data association judgment parameters based on the association analysis data to judge the association index between each type of data; and concatenates the data between different types respectively, and obtains the heterogeneous data chains between different types of data based on the concatenation results.

[0045] The real-time data status analysis module includes a data status evaluation unit and a real-time data offset analysis unit;

[0046] The data status evaluation unit comprehensively constructs a corresponding heterogeneous data link set for each heterogeneous data link, and performs status evaluation and analysis on each heterogeneous data link in the set; based on the status analysis data of each heterogeneous data link, the status is judged; if there is no abnormal heterogeneous data link, a comprehensive status evaluation is performed on the heterogeneous data link set; if there is an abnormal heterogeneous data link, the corresponding data of the current heterogeneous data link is marked and removed in the user's historical service data; after the removal operation is completed, a comprehensive status evaluation is performed on the heterogeneous data link set;

[0047] The real-time data offset analysis unit retrieves the corresponding data types of the real-time user service request data and the user historical service data for fusion processing; by dividing and classifying the real-time user service request data according to the data attribute type, and after analyzing the degree of centralization of the data, the same type of data chain is connected in series to the corresponding type of the same type of data chain; by associating heterogeneous data and completing the serial construction of the newly generated heterogeneous data chain; the security risk index of the newly generated heterogeneous data chain is analyzed, and the state judgment is made; if the judgment is abnormal, the current user service request data is prompted for abnormality; if the judgment is normal, the newly generated heterogeneous data chain is included in the heterogeneous data chain set and the comprehensive state evaluation of the new heterogeneous data chain set is performed to obtain the comprehensive security risk index of the new heterogeneous data chain set; based on the comprehensive security risk index of the new heterogeneous data chain set, the state offset analysis of the user historical service data is performed; based on the comprehensive security risk index of the new heterogeneous data chain set, the state offset analysis data of the user historical service data is performed, and the restriction parameters are introduced to make an abnormal judgment on the state offset of the newly generated heterogeneous data chain to the user historical service data.

[0048] The data feedback module includes a data feedback update unit and an abnormal warning unit;

[0049] The data feedback update unit outputs the user historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, stores the data in the cloud temporary service data space into the application service database, completes the database update and closes the cloud temporary service data space;

[0050] The abnormal warning unit warns of abnormal historical service data, abnormal real-time user service request data and abnormal risk real-time user service request operations.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention establishes a data connection between the user and the application server, builds a temporary cloud space to store the user's historical data and real-time data and conducts a series of analyses; by comprehensively dividing the historical data and conducting central data analysis and determination, and conducting correlation analysis of various types of data, a historical data corresponding correlation heterogeneous data chain is constructed; by analyzing the heterogeneous data chain status and the comprehensive historical data status, anomaly detection of historical data is achieved, and the real-time data is analyzed in combination with the fusion analysis of real-time data and historical data to perform status analysis on the real-time data; deviation impact analysis is performed based on real-time data and historical data to determine the user's real-time risk operations; the present invention establishes correlations between data through correlation analysis of historical data, mines the impact of data status, and improves the accuracy of data detection; by conducting historical data impact analysis on real-time user operation data, data anomalies and risk operations are determined, thereby improving the singleness of traditional abnormal data and risk operation detection and improving the sensitivity and diversity of data detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a structural schematic diagram of a user data supervision system based on big data analysis in the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] Example: Figure 1 As shown, the present invention provides a technical solution:

[0056] A user data supervision method based on big data analysis, the method comprising the following steps:

[0057] S100, searching user information based on the application service database, and calling the user's historical service data according to the search results; generating a temporary service data space in the cloud based on the user's service request status, and storing the user's historical service data and real-time service request data;

[0058] S200, based on the temporary service data space in the cloud, classify the corresponding user historical service data according to the data attribute type, and construct a corresponding type of data set; perform data integration analysis on each type of data set respectively, and determine the central data of each type of data set; construct a data chain of the same type based on the central data of the corresponding type of data set; perform heterogeneous data association analysis based on the contract type data chain of each type of data set, and construct a heterogeneous data chain;

[0059] S300, performing data status analysis on various data links of different types, and performing comprehensive status analysis on user historical service data based on the analysis data; generating corresponding real-time heterogeneous data links by fusing real-time user service request data with corresponding data types of user historical service data; performing status analysis on real-time heterogeneous data links, and performing status offset analysis on real-time heterogeneous data links and user historical service data;

[0060] S400: Output data request behavior status based on real-time user service request data; update user historical service data status and update the database.

[0061] The specific steps of S100 are as follows:

[0062] S101, based on the application service database, using the user registration information as the data search keyword, retrieve the user's historical service data;

[0063] S102, by judging the connection status between the application server and the user, when there is a data connection between the user and the application server, a temporary service data space is generated in the cloud to store the user's historical service data and real-time request data;.

[0064] The specific steps of S200 are as follows:

[0065] S201. In the corresponding user cloud temporary service data space, the user's historical service data is classified by preset data attribute types, and a corresponding type of data set is constructed based on the classification results; based on each type of data set, the central data of each type of data set is determined through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of the corresponding data in each type of data set, and analyzing the centralization degree of each data, and the calculation formula is:

[0066]

[0067] Among them, S(n,m) is the degree of centralization of the data with label m in the data set corresponding to type number n; q(n,m) is the word frequency of the data with label m in the data set corresponding to type number n; γ(n,m) is the value influence parameter of the data with label m in the data set corresponding to type number n; q(n) is the total word frequency of the data in the data set corresponding to type number n;

[0068] Based on the analysis of the degree of centralization of each data in each type of data set, the data with the maximum degree of centralization in each type of data set is selected as the central data, and the central data is used as the head of the data chain, and the data is sorted in descending order from large to small to construct the same type of data chain of the corresponding type of data set;

[0069] S202, combining each type of data set with the corresponding data chain of the same type to perform association analysis between different data types, the calculation formula is:

[0070]

[0071] Where R[(n,m):(n o ,m)] is the data with label m in the data set corresponding to type number n, and the data with label m in the data set corresponding to type number n o The correlation index between the data labeled m in the corresponding data set; is the feature vector of the data numbered m in the data set corresponding to type number n; is the corresponding type number n o The feature vector corresponding to the data numbered m in the data set;

[0072] Based on the association analysis data, set the data association judgment parameter R 0 , judge the correlation index between each data, if R[(n,m):(n o ,m)]≥R 0 , then it is judged that there is a correlation between the two data; if R[(n,m):(n o ,m)]<R 0 , it is determined that there is no association between the two data; the data with association between different types are connected in series, and the heterogeneous data chain between the different types of data is obtained based on the connection result.

[0073] The specific steps of S300 are as follows:

[0074] S301. Coordinate and build a corresponding heterogeneous data link set for each heterogeneous data link, and perform status evaluation and analysis on each heterogeneous data link in the set. The calculation formula is:

[0075]

[0076] Wherein, Ad(i) is the security risk index of the heterogeneous data link numbered i in the corresponding heterogeneous data link set; Y(i,u) is the value loss index of the data numbered u in the corresponding heterogeneous data link numbered i; L(i,u) is the type influence weight of the data numbered u in the corresponding heterogeneous data link numbered i; γ(i,u) is the value influence parameter of the data numbered u in the corresponding heterogeneous data link numbered i; S(i,u) is the degree of centralization of the data numbered u in the corresponding heterogeneous data link numbered i;

[0077] Based on the status analysis data of each type of data link, the status is judged; the historical security risk average index Ad(v) is quoted. If Ad(i)≥Ad(v), the corresponding heterogeneous data link data is judged to be normal; if Ad(i)<Ad(v), the corresponding heterogeneous data link data is judged to be abnormal; if there is no abnormal heterogeneous data link, a comprehensive status assessment is performed on the heterogeneous data link set, and the calculation formula is:

[0078]

[0079] Where Ad(al) is the comprehensive security risk index of the corresponding heterogeneous data link set;

[0080] If there are abnormal heterogeneous data links, the corresponding data of the current heterogeneous data links will be marked and removed from the user's historical service data; after the removal operation is completed, a comprehensive status assessment will be conducted on the heterogeneous data link set;

[0081] S302, retrieve the corresponding data types of real-time user service request data and user historical service data for fusion processing; divide and classify the real-time user service request data according to the data attribute type, and complete the serial connection of the same type of data chain to the corresponding type of the same type of data chain after data centralization analysis; associate and complete the serial construction of the newly generated heterogeneous data chain through heterogeneous data; analyze the security risk index of the newly generated heterogeneous data chain and make a status judgment; if the judgment is abnormal, give an abnormal prompt for the current user service request data; if the judgment is normal, include the newly generated heterogeneous data chain into the heterogeneous data chain set and conduct a comprehensive status evaluation of the new heterogeneous data chain set to obtain the comprehensive security risk index of the new heterogeneous data chain set; perform a status offset analysis on the user historical service data based on the comprehensive security risk index of the new heterogeneous data chain set, and the calculation formula is:

[0082]

[0083] Among them, DF(e) is the state deviation index of the historical service data of the user of the newly generated heterogeneous data link corresponding to the number e; Ad(el) is the comprehensive security risk index of the new heterogeneous data link set; Ad(e) is the security risk index of the newly generated heterogeneous data link corresponding to the number e;

[0084] Based on the comprehensive security risk index of the new heterogeneous data link set, the state deviation analysis data of the user historical service data is performed, and the restriction parameter DF(th) is introduced to make an abnormal judgment on the state deviation of the newly generated heterogeneous data link to the user historical service data; if DF(e)>DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is abnormal, and it is prompted that there is an abnormal risk in the current user service request data operation; if DF(e)≤DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is normal.

[0085] The specific steps of S400 are as follows:

[0086] S401, outputting the user's historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, storing the data in the cloud temporary service data space into the application service database, completing the database update and closing the cloud temporary service data space;

[0087] S402: Issue warnings for abnormal historical service data, abnormal real-time user service request data, and abnormal risk real-time user service request operations.

[0088] A user data supervision system based on big data analysis, the system comprising a data retrieval module, a multi-dimensional data processing module, a real-time data status analysis module and a data feedback module;

[0089] The data retrieval module searches for user information based on the application service database, and calls the user's historical service data according to the search results; based on the user service request status, a temporary service data space is generated on the cloud, and the user's historical service data and real-time service request data are stored; the multi-dimensional data processing module classifies the corresponding user's historical service data according to the data attribute type in the temporary service data space on the cloud, and constructs a corresponding type of data set; performs data integration analysis on each type of data set, and determines the central data of each type of data set; constructs a data chain of the same type based on the central data of the corresponding type of data set; and constructs a data chain of the same type based on the contract type of each type of data set. Perform heterogeneous data association analysis and construct heterogeneous data chains; the real-time data status analysis module performs data status analysis on each heterogeneous data chain, and performs comprehensive status analysis on user historical service data based on the analysis data; generates corresponding real-time heterogeneous data chains by fusing real-time user service request data with corresponding data types of user historical service data; performs status analysis on real-time heterogeneous data chains, and performs status offset analysis on real-time heterogeneous data chains and user historical service data; the data feedback module outputs data request behavior status based on real-time user service request data; updates the status of user historical service data and updates the database.

[0090] The data retrieval module includes a historical data retrieval unit and a historical data retrieval unit;

[0091] The historical data retrieval unit retrieves the user's historical service data based on the application service database, using the user registration information as a data retrieval keyword;

[0092] The historical data retrieval unit determines the connection status between the application server and the user, and generates a temporary cloud service data space when there is a data connection between the user and the application server to store the user's historical service data and real-time request data.

[0093] The multidimensional data processing module includes a data centralization analysis unit and a data association analysis unit;

[0094] The data centralization analysis unit classifies the user's historical service data in the corresponding user's cloud temporary service data space by preset data attribute types, and constructs a corresponding type of data set based on the classification result; based on each type of data set, determines the central data of each type of data set through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of each data corresponding to each type of data set, and analyzing the centralization degree of each data; based on the centralization degree analysis of each data in each type of data set, selects the corresponding data with the maximum centralization degree in each type of data set as the central data, and takes the central data as the head of the data chain, sorts the data in descending order from large to small, and constructs a data chain of the same type of the corresponding type of data set;

[0095] The data association analysis unit performs association analysis between different types of data in combination with the same type of data chains corresponding to each type of data set; sets data association judgment parameters based on the association analysis data to judge the association index between each type of data; and concatenates the data between different types respectively, and obtains the heterogeneous data chains between different types of data based on the concatenation results.

[0096] The real-time data status analysis module includes a data status evaluation unit and a real-time data offset analysis unit;

[0097] The data status evaluation unit comprehensively constructs a corresponding heterogeneous data link set for each heterogeneous data link, and performs status evaluation and analysis on each heterogeneous data link in the set; based on the status analysis data of each heterogeneous data link, the status is judged; if there is no abnormal heterogeneous data link, a comprehensive status evaluation is performed on the heterogeneous data link set; if there is an abnormal heterogeneous data link, the corresponding data of the current heterogeneous data link is marked and removed in the user's historical service data; after the removal operation is completed, a comprehensive status evaluation is performed on the heterogeneous data link set;

[0098] The real-time data offset analysis unit retrieves the corresponding data types of the real-time user service request data and the user historical service data for fusion processing; by dividing and classifying the real-time user service request data according to the data attribute type, and after analyzing the degree of centralization of the data, the same type of data chain is connected in series to the corresponding type of the same type of data chain; by associating heterogeneous data and completing the serial construction of the newly generated heterogeneous data chain; the security risk index of the newly generated heterogeneous data chain is analyzed, and the state judgment is made; if the judgment is abnormal, the current user service request data is prompted for abnormality; if the judgment is normal, the newly generated heterogeneous data chain is included in the heterogeneous data chain set and the comprehensive state evaluation of the new heterogeneous data chain set is performed to obtain the comprehensive security risk index of the new heterogeneous data chain set; based on the comprehensive security risk index of the new heterogeneous data chain set, the state offset analysis of the user historical service data is performed; based on the comprehensive security risk index of the new heterogeneous data chain set, the state offset analysis data of the user historical service data is performed, and the restriction parameters are introduced to make an abnormal judgment on the state offset of the newly generated heterogeneous data chain to the user historical service data.

[0099] The data feedback module includes a data feedback update unit and an abnormal warning unit;

[0100] The data feedback update unit outputs the user historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, stores the data in the cloud temporary service data space into the application service database, completes the database update and closes the cloud temporary service data space;

[0101] The abnormal warning unit warns of abnormal historical service data, abnormal real-time user service request data, and abnormal risk real-time user service request operations;

[0102] In the examples:

[0103] A software company currently uses the user data supervision system based on big data analysis of the present invention to supervise the user data of a certain application; it uses the user registration information as the data search keyword based on the application service database to retrieve the user's historical service data; by judging the connection status between the application server and the user, when the user and the application server have a data connection, a temporary service data space in the cloud is generated to store the user's historical service data and real-time request data;

[0104] In the corresponding user cloud temporary service data space, the user's historical service data is classified by preset data attribute types, and the corresponding type of data set is constructed based on the classification results; based on each type of data set, the central data of each type of data set is determined through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of the corresponding data in each type of data set and analyzing the centralization degree of each data. The calculation formula is:

[0105]

[0106] Based on the analysis of the degree of centralization of each data in each type of data set, the data with the maximum degree of centralization in each type of data set is selected as the central data, and the central data is used as the head of the data chain, and the data is sorted in descending order from large to small to construct the same type of data chain of the corresponding type of data set;

[0107] Combine each type of data set with the corresponding data chain of the same type to perform association analysis between different data types. The calculation formula is:

[0108]

[0109] Based on the association analysis data, set the data association judgment parameter R 0 , judge the correlation index between each data, if R[(n,m):(n o ,m)]≥R 0 , then it is judged that there is a correlation between the two data; if R[(n,m):(n o ,m)]<R 0 , it is determined that there is no association between the two data; the data of different types that are associated are connected in series, and the heterogeneous data chain between the different types of data is obtained based on the connection result;

[0110] The corresponding heterogeneous data links are constructed in a coordinated manner for each type of data link, and the status of each type of data link in the set is evaluated and analyzed. The calculation formula is:

[0111]

[0112] Based on the status analysis data of each type of data link, the status is judged; the historical security risk average index Ad(v) is quoted. If Ad(i)≥Ad(v), the corresponding heterogeneous data link data is judged to be normal; if Ad(i)<Ad(v), the corresponding heterogeneous data link data is judged to be abnormal; if there is no abnormal heterogeneous data link, a comprehensive status assessment is performed on the heterogeneous data link set, and the calculation formula is:

[0113]

[0114] If there are abnormal heterogeneous data links, the corresponding data of the current heterogeneous data links will be marked and removed from the user's historical service data; after the removal operation is completed, a comprehensive status assessment will be conducted on the heterogeneous data link set;

[0115] Retrieve the corresponding data types of real-time user service request data and user historical service data for fusion processing; divide and classify the real-time user service request data according to the data attribute type, and complete the connection of the same type of data chain to the corresponding type of the same type of data chain after data concentration analysis; associate and complete the construction of the newly generated heterogeneous data chain series through heterogeneous data; analyze the security risk index of the newly generated heterogeneous data chain and make a status judgment; if the judgment is abnormal, give an abnormal prompt for the current user service request data; if the judgment is normal, include the newly generated heterogeneous data chain into the heterogeneous data chain set and conduct a comprehensive status assessment of the new heterogeneous data chain set to obtain the comprehensive security risk index of the new heterogeneous data chain set; perform a status offset analysis on the user historical service data based on the comprehensive security risk index of the new heterogeneous data chain set, and the calculation formula is:

[0116]

[0117] Based on the comprehensive security risk index of the new heterogeneous data link set, the state deviation analysis data of the user's historical service data is performed, and the restriction parameter DF(th) is introduced to judge the abnormality of the state deviation of the newly generated heterogeneous data link to the user's historical service data; if DF(e)>DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user's historical service data is abnormal, and it is prompted that the current user service request data operation has abnormal risks; if DF(e)≤DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user's historical service data is normal;

[0118] Output the user's historical service data status and real-time user service request data status evaluation data, and when the user service request connection is disconnected, store the data in the cloud temporary service data space into the application service database, complete the database update and close the cloud temporary service data space; issue warnings for abnormal historical service data, abnormal real-time user service request data and abnormal risk real-time user service request operations.

[0119] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A user data supervision method based on big data analysis, characterized in that: The method comprises the following steps: S100, searching user information based on the application service database, and calling the user's historical service data according to the search results; generating a temporary service data space in the cloud based on the user's service request status, and storing the user's historical service data and real-time service request data; S200, based on the temporary service data space in the cloud, classify the corresponding user historical service data according to the data attribute type, and construct a corresponding type of data set; perform data integration analysis on each type of data set respectively, and determine the central data of each type of data set; construct a data chain of the same type based on the central data of the corresponding type of data set; perform heterogeneous data association analysis based on the contract type data chain of each type of data set, and construct a heterogeneous data chain; S300, performing data status analysis on various data links of different types, and performing comprehensive status analysis on user historical service data based on the analysis data; generating corresponding real-time heterogeneous data links by fusing real-time user service request data with corresponding data types of user historical service data; performing status analysis on real-time heterogeneous data links, and performing status offset analysis on real-time heterogeneous data links and user historical service data; S400: Output data request behavior status based on real-time user service request data; update user historical service data status and update the database.

2. The user data supervision method based on big data analysis according to claim 1 is characterized by: The specific steps of S100 are as follows: S101, based on the application service database, using the user registration information as the data search keyword, retrieve the user's historical service data; S102, by judging the connection status between the application server and the user, when there is a data connection between the user and the application server, a temporary service data space is generated in the cloud to store the user's historical service data and real-time request data.

3. The user data supervision method based on big data analysis according to claim 2 is characterized by: The specific steps of S200 are as follows: S201. In the corresponding user cloud temporary service data space, the user's historical service data is classified by preset data attribute types, and a corresponding type of data set is constructed based on the classification results; based on each type of data set, the central data of each type of data set is determined through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of the corresponding data in each type of data set, and analyzing the centralization degree of each data, and the calculation formula is: Among them, S(n,m) is the degree of centralization of the data with label m in the data set corresponding to type number n; q(n,m) is the word frequency of the data with label m in the data set corresponding to type number n; γ(n,m) is the value influence parameter of the data with label m in the data set corresponding to type number n; q(n) is the total word frequency of the data in the data set corresponding to type number n; Based on the analysis of the degree of centralization of each data in each type of data set, the data with the maximum degree of centralization in each type of data set is selected as the central data, and the central data is used as the head of the data chain, and the data is sorted in descending order from large to small to construct the same type of data chain of the corresponding type of data set; S202, combining each type of data set with the corresponding data chain of the same type to perform association analysis between different data types, the calculation formula is: Where R[(n,m):(n o ,m)] is the data with label m in the data set corresponding to type number n, and the data with label m in the data set corresponding to type number n o The correlation index between the data labeled m in the corresponding data set; is the feature vector of the data numbered m in the data set corresponding to type number n; is the corresponding type number n o The feature vector corresponding to the data numbered m in the data set; Based on the association analysis data, set the data association judgment parameter R0 to judge the association index between each data. If R[(n,m):(n o ,m)]≥R0, then it is judged that there is a correlation between the two data; if R[(n,m):(n o ,m)]<R0, it is judged that there is no association between the two data; the data with association between different types are connected in series respectively, and the heterogeneous data chains between different types of data are obtained based on the connection results.

4. The user data supervision method based on big data analysis according to claim 3 is characterized by: The specific steps of S300 are as follows: S301. Coordinate and build a corresponding heterogeneous data link set for each heterogeneous data link, and perform status evaluation and analysis on each heterogeneous data link in the set. The calculation formula is: Wherein, Ad(i) is the security risk index of the heterogeneous data link numbered i in the corresponding heterogeneous data link set; Y(i,u) is the value loss index of the data numbered u in the corresponding heterogeneous data link numbered i; L(i,u) is the type influence weight of the data numbered u in the corresponding heterogeneous data link numbered i; γ(i,u) is the value influence parameter of the data numbered u in the corresponding heterogeneous data link numbered i; S(i,u) is the degree of centralization of the data numbered u in the corresponding heterogeneous data link numbered i; Based on the status analysis data of each type of data link, the status is judged; the historical security risk average index Ad(v) is quoted. If Ad(i)≥Ad(v), the corresponding heterogeneous data link data is judged to be normal; if Ad(i)<Ad(v), the corresponding heterogeneous data link data is judged to be abnormal; if there is no abnormal heterogeneous data link, a comprehensive status assessment is performed on the heterogeneous data link set, and the calculation formula is: Where Ad(al) is the comprehensive security risk index of the corresponding heterogeneous data link set; If there are abnormal heterogeneous data links, the corresponding data of the current heterogeneous data links will be marked and removed from the user's historical service data; after the removal operation is completed, a comprehensive status assessment will be conducted on the heterogeneous data link set; S302, retrieve the corresponding data types of real-time user service request data and user historical service data for fusion processing; divide and classify the real-time user service request data according to the data attribute type, and complete the serial connection of the same type of data chain to the corresponding type of the same type of data chain after data centralization analysis; associate and complete the serial construction of the newly generated heterogeneous data chain through heterogeneous data; analyze the security risk index of the newly generated heterogeneous data chain and make a status judgment; if the judgment is abnormal, give an abnormal prompt for the current user service request data; if the judgment is normal, include the newly generated heterogeneous data chain into the heterogeneous data chain set and conduct a comprehensive status evaluation of the new heterogeneous data chain set to obtain the comprehensive security risk index of the new heterogeneous data chain set; perform a status offset analysis on the user historical service data based on the comprehensive security risk index of the new heterogeneous data chain set, and the calculation formula is: Among them, DF(e) is the state deviation index of the historical service data of the user of the newly generated heterogeneous data link corresponding to the number e; Ad(el) is the comprehensive security risk index of the new heterogeneous data link set; Ad(e) is the security risk index of the newly generated heterogeneous data link corresponding to the number e; Based on the comprehensive security risk index of the new heterogeneous data link set, the state deviation analysis data of the user historical service data is performed, and the restriction parameter DF(th) is introduced to make an abnormal judgment on the state deviation of the newly generated heterogeneous data link to the user historical service data; if DF(e)>DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is abnormal, and it is prompted that there is an abnormal risk in the current user service request data operation; if DF(e)≤DF(th), it is judged that the state deviation of the current newly generated heterogeneous data link to the user historical service data is normal.

5. The user data supervision method based on big data analysis according to claim 4 is characterized in that: The specific steps of S400 are as follows: S401, outputting the user's historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, storing the data in the cloud temporary service data space into the application service database, completing the database update and closing the cloud temporary service data space; S402: Issue warnings for abnormal historical service data, abnormal real-time user service request data, and abnormal risk real-time user service request operations.

6. A user data supervision system based on big data analysis, characterized in that: The system includes a data retrieval module, a multi-dimensional data processing module, a real-time data status analysis module and a data feedback module; The data retrieval module searches for user information based on the application service database, and calls the user's historical service data according to the search results; based on the user's service request status, a temporary service data space is generated in the cloud to store the user's historical service data and real-time service request data; The multi-dimensional data processing module classifies the corresponding user historical service data according to the data attribute type in the cloud temporary service data space, and constructs a corresponding type of data set; performs data integration analysis on each type of data set respectively, and determines the central data of each type of data set; constructs a same type of data chain based on the central data of the corresponding type of data set; performs heterogeneous data association analysis based on the contract type data chain of each type of data set, and constructs a heterogeneous data chain; the real-time data status analysis module performs data status analysis on each heterogeneous data chain, and performs a comprehensive status analysis on the user historical service data based on the analysis data; generates a corresponding real-time heterogeneous data chain by fusing the real-time user service request data with the corresponding data type of the user historical service data; Perform status analysis on real-time heterogeneous data links, and perform status offset analysis on real-time heterogeneous data links and user historical service data; the data feedback module outputs data request behavior status based on real-time user service request data; updates the user historical service data status and updates the database.

7. A user data supervision system based on big data analysis according to claim 6, characterized in that: The data retrieval module includes a historical data retrieval unit and a historical data retrieval unit; The historical data retrieval unit retrieves the user's historical service data based on the application service database, using the user registration information as a data retrieval keyword; The historical data retrieval unit determines the connection status between the application server and the user, and generates a temporary cloud service data space when there is a data connection between the user and the application server to store the user's historical service data and real-time request data.

8. The user data supervision system based on big data analysis according to claim 7 is characterized by: The multidimensional data processing module includes a data centralization analysis unit and a data association analysis unit; The data centralization analysis unit classifies the user's historical service data in the corresponding user's cloud temporary service data space by preset data attribute types, and constructs a corresponding type of data set based on the classification result; based on each type of data set, determines the central data of each type of data set through integrated analysis; the integrated analysis is performed by coordinating the word frequencies of each data corresponding to each type of data set, and analyzing the centralization degree of each data; based on the centralization degree analysis of each data in each type of data set, selects the corresponding data with the maximum centralization degree in each type of data set as the central data, and takes the central data as the head of the data chain, sorts the data in descending order from large to small, and constructs a data chain of the same type of the corresponding type of data set; The data association analysis unit performs association analysis between different types of data in combination with the same type of data chains corresponding to each type of data set; sets data association judgment parameters based on the association analysis data to judge the association index between each type of data; and concatenates the data between different types respectively, and obtains the heterogeneous data chains between different types of data based on the concatenation results.

9. The user data supervision system based on big data analysis according to claim 8 is characterized by: The real-time data status analysis module includes a data status evaluation unit and a real-time data offset analysis unit; The data status evaluation unit comprehensively constructs a corresponding heterogeneous data link set for each heterogeneous data link, and performs status evaluation and analysis on each heterogeneous data link in the set; based on the status analysis data of each heterogeneous data link, the status is judged; if there is no abnormal heterogeneous data link, a comprehensive status evaluation is performed on the heterogeneous data link set; if there is an abnormal heterogeneous data link, the corresponding data of the current heterogeneous data link is marked and removed in the user's historical service data; After the elimination operation is completed, a comprehensive status evaluation is performed on the heterogeneous data link set; The real-time data offset analysis unit retrieves the corresponding data types of the real-time user service request data and the user historical service data for fusion processing; By dividing and classifying the real-time user service request data according to the data attribute type, and analyzing the degree of centralization of the data, the same type of data link is connected in series to the corresponding type of the same type of data link; by associating heterogeneous data and completing the serial construction of the newly generated heterogeneous data link; the security risk index of the newly generated heterogeneous data link is analyzed, and the status is judged; if the judgment is abnormal, the current user service request data is prompted for abnormality; if the judgment is normal, the newly generated heterogeneous data link is included in the heterogeneous data link set and a comprehensive status evaluation of the new heterogeneous data link set is performed to obtain the comprehensive security risk index of the new heterogeneous data link set; based on the comprehensive security risk index of the new heterogeneous data link set, the user's historical service data is analyzed for state deviation; based on the comprehensive security risk index of the new heterogeneous data link set, the user's historical service data is analyzed for state deviation, and restriction parameters are introduced to make abnormal judgments on the state deviation of the newly generated heterogeneous data link to the user's historical service data.

10. A user data supervision system based on big data analysis according to claim 9, characterized in that: The data feedback module includes a data feedback update unit and an abnormal warning unit; The data feedback update unit outputs the user historical service data status and the real-time user service request data status evaluation data, and when the user service request connection is disconnected, stores the data in the cloud temporary service data space into the application service database, completes the database update and closes the cloud temporary service data space; The abnormal warning unit warns of abnormal historical service data, abnormal real-time user service request data and abnormal risk real-time user service request operations.

Images