Social network identity association method and system based on big data

Abstract

The invention relates to the technical field of social network analysis, in particular to a social network identity association method based on big data, which comprises the following steps: constructing an active time period model, a browsing preference model, a friend relationship model and an attention relationship model, calculating an active feature homogeneity degree and a browsing feature homogeneity degree, judging whether the active feature homogeneity degree and the browsing feature homogeneity degree are both greater than a preset homogeneity degree threshold value or not, and if yes, judging whether the homogeneity degree is greater than the preset homogeneity degree threshold value. Extracting second iteration user features and recalculating active feature homogeneity and browsing feature homogeneity if not all the features are greater than the homogeneity threshold value, calculating friend feature homogeneity and attention feature homogeneity if all the features are greater than the homogeneity threshold value, calculating mixed homogeneity based on feature weights, constructing a feature function, and calculating the content of the active feature homogeneity and the browsing feature homogeneity; and inputting the mixed homogeneity into the characteristic function to obtain a function output value, judging whether the function output value is 1 or not, and if the function output value is 1, confirming social network identity association. According to the invention, the association accuracy of the social network identity can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of social network analysis, and in particular to a social network identity association method, system, electronic device and computer-readable storage medium based on big data. Background Art

[0002] Social network identity association refers to confirming whether two accounts on different online platforms belong to the same natural person by analyzing relevant information of users on different online platforms.

[0003] Traditional social network identity association methods rely on simple feature matching, such as username, email address or phone number, to determine social network identity by matching a single feature.

[0004] Although the above method can realize social network identity association, it has certain limitations when facing complex data and changeable user behaviors. It is impossible to comprehensively analyze user data and cannot accurately associate identities. Therefore, the current social network association method has the problem of low association accuracy. Summary of the invention

[0005] The present invention provides a social network identity association method based on big data and a computer-readable storage medium, the main purpose of which is to improve the association accuracy of social network identities.

[0006] To achieve the above object, the present invention provides a social network identity association method based on big data, comprising:

[0007] Collect user identity big data from two network domains to build active time period models, browsing preference models, friend relationship models, and attention relationship models;

[0008] Performing data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model, and the attention relationship model to obtain a first user feature set and a second user feature set;

[0009] Extracting a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature, and a first attention feature;

[0010] Extracting second user features in sequence from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature;

[0011] Draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph respectively using the first active feature, the first browsing feature, the second active feature, and the second browsing feature;

[0012] Performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity;

[0013] Determining whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold;

[0014] If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0015] If the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculating the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature;

[0016] Calculate mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights;

[0017] Constructing a characteristic function, inputting the mixed homogeneity into the characteristic function, and obtaining a function output value;

[0018] Determine whether the function output value is 1;

[0019] If the function output value is not 1, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0020] If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

[0021] Optionally, collecting user identity big data of two network domains includes:

[0022] Obtaining API authentication mechanisms of the two network domains, and sending preset API requests to the two network domains based on the API authentication mechanisms of the two network domains and the pre-obtained API keys, obtaining request results, and determining whether the API request successfully connects the two network domains based on the request results;

[0023] If the API request fails to successfully connect the two network domains, return to the above step of obtaining the API authentication mechanisms of the two network domains;

[0024] If the API request successfully connects the two network domains, constructing a first API request using preset initial paging parameters, and acquiring first request data based on the initial paging parameters and the first API request;

[0025] Determine whether the preset total paging parameter is equal to 1;

[0026] If the total paging parameter is equal to 1, the first request data is used as user identity big data;

[0027] If the total paging parameter is not equal to 1, determining whether the initial paging parameter is equal to the total paging parameter;

[0028] If the initial paging parameter is not equal to the total paging parameter, the initial paging parameter is increased by one to obtain an updated paging parameter, the initial paging parameter is updated using the updated paging parameter, and then the update API request is constructed using the initial paging parameter, and the update request data is obtained based on the initial paging parameter and the update API request, and the step of determining whether the initial paging parameter is equal to the total paging parameter is returned to the above step;

[0029] If the initial paging parameter is equal to the total paging parameter, the update request data and the first request data are merged in a form of the first request data first and the update request data later to obtain user identity big data.

[0030] Optionally, the construction process of the active time period model is as follows:

[0031] Obtain the active state weight, work state weight, state activity, and work activity in each active time period, wherein the active time period may be: [0,6)h, [6,16)h, [16,20)h, [20,24)h;

[0032] The active time period model is constructed using the active state weight, work state weight, state activity and work activity:

[0033] T e =ε×X e +∈×Y e

[0034] Among them, T e represents the first active feature value or the second active feature value in the e-th active time period, e represents the sequence number of the active time period, ε represents the active state weight, and X e represents the state activity in the e-th active time period, ∈ represents the work state weight, and Y e Indicates the work activity in the e-th active time period.

[0035] Optionally, the construction process of the browsing preference model, the friend relationship model and the attention relationship model is as follows:

[0036] Obtain browsing time period, content type, browsing time interval, browsing frequency, browsing device, browsing location and content nature, wherein the browsing time period includes: [0,6)h, [6,16)h, [16,20)h, [20,24)h, the content type includes: news content, entertainment content, science and technology content, literary content and educational content, the browsing time interval includes: [10,20)min, [20,40)min, [40,60)min,, [60,+∞)min, the browsing frequency includes: greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 7 times / day, greater than 7 times / day and less than or equal to 9 times / day and greater than 9 times / day, the browsing device includes: mobile phone and PC, the browsing location includes: city and suburbs, the content nature includes: latest information and past content;

[0037] Arrange and combine the browsing time period, content type, browsing duration interval, browsing frequency, browsing device, browsing location and content nature to obtain a browsing dimension;

[0038] Obtaining the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension, and constructing a browsing preference model based on the browsing dimension, the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension;

[0039] The browsing preference model is as follows:

[0040]

[0041] Among them, B τ (j, k, l, v, w, f) represents the first browsing feature value or the second browsing feature value in the τth browsing time period and when the browsing dimension is (j, k, l, v, w, f), τ represents the ordinal number of the browsing time period, j represents the content type, k represents the browsing duration interval, l represents the browsing frequency, v represents the browsing device, w represents the browsing location, f represents the content nature, n represents the dimension ordinal number of the browsing dimension, α i represents the dimension weight of the i-th browsing dimension, β i Indicates the browsing preference of the i-th browsing dimension;

[0042] Obtain interaction frequency, common friend relationship, friend interaction degree of interaction frequency, interaction frequency weight, interaction weight of interaction frequency, friend relationship degree of common friend relationship, common friend weight and relationship weight of common friend relationship, wherein the interaction frequency includes: greater than 1 time / day and less than or equal to 3 times / day, greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 10 times / day and greater than 10 times / day, and the common friend relationship includes: greater than 1 common friend and less than or equal to 3 common friends, greater than 3 common friends and less than or equal to 5 common friends, greater than 5 common friends and less than or equal to 10 common friends and greater than 10 common friends;

[0043] Constructing a friend relationship model based on the interaction frequency, common friend relationship, friend interaction degree of the interaction frequency, interaction frequency weight, interaction weight of the interaction frequency, friend relationship degree of the common friend relationship, common friend weight and relationship weight of the common friend relationship;

[0044] The friendship model is as follows:

[0045]

[0046] Among them, A represents the first friend's characteristic value or the second friend's characteristic value, θ represents the interaction frequency weight, and B b represents the friend interaction degree of the bth interaction frequency, B b1 represents the interaction weight of the bth interaction frequency, μ represents the common friend weight, O m represents the friendship degree of the mth common friendship, O m1 represents the relationship weight of the mth common friend relationship;

[0047] Acquire a common attention relationship, a content similarity interval, a common attention weight, a common attention degree of the common attention relationship, an attention weight of the common attention relationship, a content similarity interval weight, a content similarity of the content similarity interval, and a similarity interval weight of the content similarity interval, wherein the common attention relationship includes: more than 1 common attention and less than or equal to 3 common attentions, more than 3 common attentions and less than or equal to 5 common attentions, more than 5 common attentions and less than or equal to 10 common attentions, and more than 10 common attentions, and the content similarity interval includes: [0, 10%), [10%, 30%), [30%, 60%), and [60%, 100%);

[0048] Constructing the attention relationship model using the common attention relationship, content similarity interval, common attention weight, common attention degree of the common attention relationship, attention weight of the common attention relationship, content similarity interval weight, content similarity of the content similarity interval and similarity interval weight of the content similarity interval;

[0049] The following relationship model is as follows:

[0050]

[0051] Where Y represents the first attention feature value or the second attention feature value, ρ represents the joint attention weight, and M m represents the common attention degree of the mth common attention relationship, M m1 represents the attention weight of the mth common attention relationship, represents the weight of content similarity interval, U u represents the content similarity of the u-th content similarity interval, U u1 Represents the similarity interval weight of the u-th content similarity interval.

[0052] Optionally, the using the first active feature, the first browsing feature, the second active feature and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph and a second browsing feature graph respectively includes:

[0053] Constructing a first active time data matrix based on the active time period model and the first active feature value;

[0054] Draw a first active feature graph according to the first active time data matrix, wherein the x-axis of the first active feature graph represents the active time period, and the y-axis represents the first active feature value;

[0055] Constructing a first browsing preference data matrix based on the browsing preference model and the first browsing feature value;

[0056] Draw a first browsing characteristic graph according to the first browsing preference data matrix, wherein the x-axis of the first browsing characteristic graph represents the browsing time period, and the y-axis represents the first browsing characteristic value;

[0057] constructing a second active time data matrix based on the active time period model and the second active feature value;

[0058] Draw a second active feature graph according to the second active time data matrix, wherein the x-axis of the second active feature graph represents the active time period, and the y-axis represents the second active feature value;

[0059] Constructing a second browsing preference data matrix based on the browsing preference model and the second browsing characteristic value;

[0060] A second browsing characteristic graph is drawn according to the second browsing preference data matrix, wherein the x-axis of the second browsing characteristic graph represents the browsing time period, and the y-axis represents the second browsing characteristic value.

[0061] Optionally, constructing a first active time data matrix based on the active time period model and the first active feature value includes:

[0062] Acquire the activity state and the working state, wherein the activity state includes: high activity state, medium activity state, low activity state and inactive state, and the working state includes: working and resting;

[0063] constructing a first active time data matrix using the active state and the working state;

[0064] The first active time data matrix is ​​as follows:

[0065]

[0066] Among them, R refers to the first active time data matrix, T 11 It refers to the first active feature value when the active state is high active state and the working state is working in the e-th active time period, T 12 Refers to the first active feature value when the active state is high active state and the working state is resting in the e-th active time period, T 41 It refers to the first active feature value when the active state is inactive and the working state is working in the e-th active time period, T 42 It refers to the first active feature value when the active state is inactive and the working state is resting in the e-th active time period.

[0067] Optionally, performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity includes:

[0068] Obtain the time weight of the active time period, the second active feature value of the active time period, and the maximum possible distance of the feature value, and calculate the homogeneity of the active feature based on the time weight of the active time period, the second active feature value, and the maximum possible distance of the feature value:

[0069]

[0070] Among them, H1 represents the homogeneity of active features, Q e represents the time weight of the e-th active time period, T e represents the first active feature value of the e-th active time period, T e ' represents the second active feature value of the e-th active time period, D max Represents the maximum possible distance of eigenvalues.

[0071] Optionally, the calculating of mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights includes:

[0072] The mixed homogeneity is calculated using the feature weights according to the following formula, wherein the feature weights include: active feature weight, browsing feature weight, friend feature weight, and attention feature weight:

[0073] H m =ω d ×H1+ω z ×H2+ω f ×H3+ω c ×H4

[0074] Among them, H m represents the mixing homogeneity, ω d represents the active feature weight, ω z represents the browsing feature weight, ω f represents the friend feature weight, ω c represents the weight of the attention feature, H2 represents the homogeneity of the browsing feature, H3 represents the homogeneity of the friend feature, and H4 represents the homogeneity of the attention feature.

[0075] Optionally, constructing a characteristic function includes:

[0076] Construct the characteristic function based on the preset mixed homogeneity threshold:

[0077]

[0078] Among them, f(H m ) represents the characteristic function, H y Represents the mixture homogeneity threshold.

[0079] To achieve the above objectives, the present invention further provides a social network identity association system based on big data, comprising:

[0080] A first user feature collection module is used to collect user identity big data of two network domains, construct an active time period model, a browsing preference model, a friend relationship model and a follow-up relationship model, perform data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model and the follow-up relationship model to obtain a first user feature set and a second user feature set, and extract a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature and a first follow-up feature;

[0081] a homogeneity calculation module, configured to sequentially extract second user features from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature; using the first active feature, the first browsing feature, the second active feature, and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph, respectively; performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain an active feature homogeneity and a browsing feature homogeneity;

[0082] a homogeneity threshold judgment module, used to judge whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold; if the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, then return to the above step of sequentially extracting the second user features from the second user feature set; if the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculate the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature;

[0083] The social network identity association module is used to calculate the mixed homogeneity based on the active feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights, construct a feature function, input the mixed homogeneity into the feature function, obtain a function output value, and determine whether the function output value is 1. If the function output value is not 1, return to the above-mentioned step of sequentially extracting the second user features from the second user feature set. If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

[0084] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0085] A memory storing at least one instruction;

[0086] The processor executes the instructions stored in the memory to implement the above-mentioned social network identity association method based on big data.

[0087] In order to solve the above problem, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned social network identity association method based on big data.

[0088] The present invention solves the problem described in the background technology. By combining the user identity big data collected from two network domains, constructing multiple user feature models and comprehensively processing the user identity big data, the accurate association of social network identities is achieved. First, the user identity big data of the two network domains is collected, and an active time period model, a browsing preference model, a friend relationship model and a follow-up relationship model are constructed to achieve comprehensive capture of user behaviors and characteristics, accurately describe the user's behavior patterns in different network domains, and improve the comprehensiveness and representativeness of the data. Then, a first active feature graph, a first browsing feature graph, a second active feature graph and a second browsing feature graph are drawn, and homogeneity processing is performed, so that the similarity of the behaviors of two users in different network domains can be clearly displayed. By calculating the active The feature homogeneity and browsing feature homogeneity are calculated, and it is judged whether they meet the preset homogeneity threshold, which can effectively improve the accuracy of identity recognition; afterwards, if the active feature homogeneity and browsing feature homogeneity do not all meet the preset threshold, the second user feature set is iteratively updated. This iterative update can continuously optimize user features and improve the efficiency of social network identity association; further, by calculating the friend feature homogeneity and the attention feature homogeneity, and calculating the mixed homogeneity based on the preset feature weight, multiple feature dimensions of the user are comprehensively considered, providing more comprehensive and accurate data for the final identity confirmation; finally, by constructing a feature function and inputting the mixed homogeneity into the feature function, the matching of the user identity can be automatically judged, so that the social network identity association has a higher accuracy. Therefore, the present invention can improve the association accuracy of social network identities. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 A schematic diagram of a process of a social network identity association method based on big data provided by an embodiment of the present invention;

[0090] Figure 2 A functional module diagram of a social network identity association system based on big data provided by an embodiment of the present invention;

[0091] Figure 3 A schematic diagram of the structure of an electronic device for implementing the big data-based social network identity association method provided by an embodiment of the present invention.

[0092] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0093] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0094] The embodiment of the present application provides a social network identity association method based on big data. The execution subject of the social network identity association method based on big data includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the social network identity association method based on big data can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0095] Reference Figure 1 FIG. 1 is a flow chart of a method for associating social network identities based on big data provided by an embodiment of the present invention. In this embodiment, the method for associating social network identities based on big data includes:

[0096] S1. Collect user identity big data from two network domains and build active time period models, browsing preference models, friend relationship models and attention relationship models.

[0097] Explainable, network domain refers to online social platforms, such as Weibo, Twitter and Baidu Tieba, etc., user identity big data refers to the publicly available data when users register on online social platforms and the data generated when users use online social platforms, such as browsed content, browsing time, browsing frequency, etc., active time period model refers to the model used to analyze the active feature values ​​of users in different time periods of the day, and the active feature values ​​refer to the degree of user activity in different time periods of the day. The browsing preference model refers to the model used to analyze the browsing preferences of users in different time periods of the day, and browsing preference refers to the content that users prefer to browse. The friend relationship model refers to the model used to analyze the friend relationship of users, and the friend relationship refers to the relationship between friends added online by users on the online social platform. The attention relationship model refers to the model used to analyze the user's attention relationship, and the attention relationship refers to the relationship between users paying attention to other users on the online social platform.

[0098] In detail, the collection of user identity big data of two network domains includes:

[0099] Obtaining API authentication mechanisms of the two network domains, and sending preset API requests to the two network domains based on the API authentication mechanisms of the two network domains and the pre-obtained API keys, obtaining request results, and determining whether the API request successfully connects the two network domains based on the request results;

[0100] If the API request fails to successfully connect the two network domains, return to the above step of obtaining the API authentication mechanisms of the two network domains;

[0101] If the API request successfully connects the two network domains, constructing a first API request using preset initial paging parameters, and acquiring first request data based on the initial paging parameters and the first API request;

[0102] Determine whether the preset total paging parameter is equal to 1;

[0103] If the total paging parameter is equal to 1, the first request data is used as user identity big data;

[0104] If the total paging parameter is not equal to 1, determining whether the initial paging parameter is equal to the total paging parameter;

[0105] If the initial paging parameter is not equal to the total paging parameter, the initial paging parameter is increased by one to obtain an updated paging parameter, the initial paging parameter is updated using the updated paging parameter, and then the update API request is constructed using the initial paging parameter, and the update request data is obtained based on the initial paging parameter and the update API request, and the step of determining whether the initial paging parameter is equal to the total paging parameter is returned to the above step;

[0106] If the initial paging parameter is equal to the total paging parameter, the update request data and the first request data are merged in a form of the first request data first and the update request data later to obtain user identity big data.

[0107] It can be explained that the API authentication mechanism refers to the mechanism for verifying identity in the application program interface (API), which is used to ensure that the API request comes from a legitimate user or system, and is usually used to protect sensitive data and prevent unauthorized access. The API key refers to a unique string or code generated by the API provider, which is used to verify the identity of the API caller and authorize its access. The API request refers to a request sent to the network domain through the application program interface to access the data in the network domain. The request result refers to the content returned by the network domain according to the API request. The initial paging parameter refers to the initial page number used to control the amount of identity data obtained when obtaining user identity data. The initial paging parameter starts from 1 and is used to process a large amount of data to avoid excessive pressure on the server or network caused by obtaining all data at one time. The first API request refers to the request actually used to obtain data. By constructing this request, the acquisition of user identity data is started. The first request data refers to the user identity data of the first page obtained according to the initial paging parameter. The total paging parameter refers to the maximum number of pages specified. For example, if the user identity data is obtained from the first page, and the maximum number of pages specified is 100 pages, the acquisition operation is terminated after the user identity data of the 100th page is obtained. 100 is the total paging parameter.

[0108] It can be understood that user identity big data refers to data obtained based on the first request data or data obtained based on the first request data, update request data and merging rules. The merging rule refers to the rule that the first request data comes first and the update request data comes later. The update paging parameter refers to the paging parameter obtained by adding one to the initial paging parameter when the number of pages is less than or equal to the maximum number of pages specified by the total paging parameter. For example, the maximum paging parameter is 10 pages, the initial paging parameter is 1 page, and the update paging parameter is 2 pages. The update API request refers to an API request reconstructed according to the update paging parameter, and the update API request uses the updated page number to obtain the user identity data of a new page. The update request data refers to the user identity data of a new page obtained using the updated page number.

[0109] Exemplarily, if the initial paging parameter is 1 and the total paging parameter is 2, the first request data is obtained based on the initial paging parameter and the first API request, the initial paging parameter is increased by 1 to obtain an updated paging parameter of 2, the initial paging parameter is updated using the updated paging parameter, and then the initial paging parameter and the update API request are used to obtain the updated request data. The initial paging parameter is equal to the total paging parameter, then the first request data is placed in front of the updated request data to obtain user identity big data.

[0110] S2. Perform data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model and the attention relationship model to obtain a first user feature set and a second user feature set.

[0111] It can be understood that performing data processing on the user identity big data refers to the action of inputting the user identity big data into the active time period model, browsing preference model, friend relationship model and attention relationship model, the first user feature set refers to the set formed by the data obtained after the user identity big data collected from one network social platform is processed by the active time period model, browsing preference model, friend relationship model and attention relationship model, and the second user feature set refers to the set formed by the data obtained after the user identity big data collected from another network social platform is processed by the active time period model, browsing preference model, friend relationship model and attention relationship model.

[0112] In detail, the active time period model building process is as follows:

[0113] Obtain the active state weight, work state weight, state activity, and work activity in each active time period, wherein the active time period may be: [0,6)h, [6,16)h, [16,20)h, [20,24)h;

[0114] The active time period model is constructed using the active state weight, work state weight, state activity and work activity:

[0115] T e =ε×X e +∈×Y e

[0116] Among them, T e represents the first active feature value or the second active feature value in the e-th active time period, e represents the sequence number of the active time period, ε represents the active state weight, and X e represents the state activity in the e-th active time period, ∈ represents the work state weight, and Y e Indicates the work activity in the e-th active time period.

[0117] It can be understood that the active time period refers to the time period during which the user is active in a day. For example, [16,20)h and [20,24)h are active time periods. The active state weight refers to the influence of the active state on the first active feature value and the second active feature value. The working state weight refers to the influence of the working state on the first active feature value and the second active feature value. For example, the influence of the active state on the first active feature value and the second active feature value is 40%, and the influence of the working state on the first active feature value and the second active feature value is 60%. Then the active state weight is 0.4, and the working state weight is 0.6. The active states include: high active state, medium active state, low active state and inactive state. The high active state, medium active state, low active state and inactive state all refer to the user's activity level in the current active time period. The higher the activity level, the more active the user is in the active time period. The high active state, medium active state, low active state and inactive state all represent an interval. When the user's activity level reaches When the intervals represented by the high active state, the medium active state, the low active state and the inactive state are selected, the active state of the user in the current active time period is confirmed. For example, the interval represented by the high active state is that the user's activity level is between (80%, 100%], the working state includes: working and resting, both working and resting refer to the state of the user in the current active time period, working refers to the user working in the current active time period, resting refers to the user resting in the current active time period, state activity refers to the value of the active state of the user in the current active time period, work activity refers to the value of the working state of the user in the current active time period, the first active feature value refers to the comprehensive activity of users in one network domain under different active states and different working state conditions, the second active feature value refers to the comprehensive activity of users in another network domain under different active states and different working state conditions, the comprehensive activity refers to the activity that combines the active state and the working state, and the activity refers to the value used to determine whether the users of the two network domains are the same user.

[0118] In detail, the construction process of the browsing preference model, the friend relationship model and the attention relationship model is as follows:

[0119] Obtain browsing time period, content type, browsing time interval, browsing frequency, browsing device, browsing location and content nature, wherein the browsing time period includes: [0,6)h, [6,16)h, [16,20)h, [20,24)h, the content type includes: news content, entertainment content, science and technology content, literary content and educational content, the browsing time interval includes: [10,20)min, [20,40)min, [40,60)min,, [60,+∞)min, the browsing frequency includes: greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 7 times / day, greater than 7 times / day and less than or equal to 9 times / day and greater than 9 times / day, the browsing device includes: mobile phone and PC, the browsing location includes: city and suburbs, the content nature includes: latest information and past content;

[0120] Arrange and combine the browsing time period, content type, browsing duration interval, browsing frequency, browsing device, browsing location and content nature to obtain a browsing dimension;

[0121] Obtaining the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension, and constructing a browsing preference model based on the browsing dimension, the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension;

[0122] The browsing preference model is as follows:

[0123]

[0124] Among them, B τ (j, k, l, v, w, f) represents the first browsing feature value or the second browsing feature value in the τth browsing time period and when the browsing dimension is (j, k, l, v, w, f), τ represents the ordinal number of the browsing time period, j represents the content type, k represents the browsing duration interval, l represents the browsing frequency, v represents the browsing device, w represents the browsing location, f represents the content nature, n represents the dimension ordinal number of the browsing dimension, α i represents the dimension weight of the i-th browsing dimension, β i Indicates the browsing preference of the i-th browsing dimension;

[0125] Obtain interaction frequency, common friend relationship, friend interaction degree of interaction frequency, interaction frequency weight, interaction weight of interaction frequency, friend relationship degree of common friend relationship, common friend weight and relationship weight of common friend relationship, wherein the interaction frequency includes: greater than 1 time / day and less than or equal to 3 times / day, greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 10 times / day and greater than 10 times / day, and the common friend relationship includes: greater than 1 common friend and less than or equal to 3 common friends, greater than 3 common friends and less than or equal to 5 common friends, greater than 5 common friends and less than or equal to 10 common friends and greater than 10 common friends;

[0126] Constructing a friend relationship model based on the interaction frequency, common friend relationship, friend interaction degree of the interaction frequency, interaction frequency weight, interaction weight of the interaction frequency, friend relationship degree of the common friend relationship, common friend weight and relationship weight of the common friend relationship;

[0127] The friendship model is as follows:

[0128]

[0129] Among them, A represents the first friend's characteristic value or the second friend's characteristic value, θ represents the interaction frequency weight, and B b represents the friend interaction degree of the bth interaction frequency, B b1 represents the interaction weight of the bth interaction frequency, μ represents the common friend weight, O m represents the friendship degree of the mth common friendship, O m1 represents the relationship weight of the mth common friend relationship;

[0130] Acquire a common attention relationship, a content similarity interval, a common attention weight, a common attention degree of the common attention relationship, an attention weight of the common attention relationship, a content similarity interval weight, a content similarity of the content similarity interval, and a similarity interval weight of the content similarity interval, wherein the common attention relationship includes: more than 1 common attention and less than or equal to 3 common attentions, more than 3 common attentions and less than or equal to 5 common attentions, more than 5 common attentions and less than or equal to 10 common attentions, and more than 10 common attentions, and the content similarity interval includes: [0, 10%), [10%, 30%), [30%, 60%), and [60%, 100%);

[0131] Constructing the attention relationship model using the common attention relationship, content similarity interval, common attention weight, common attention degree of the common attention relationship, attention weight of the common attention relationship, content similarity interval weight, content similarity of the content similarity interval and similarity interval weight of the content similarity interval;

[0132] The following relationship model is as follows:

[0133]

[0134] Where Y represents the first attention feature value or the second attention feature value, ρ represents the joint attention weight, and M m represents the common attention degree of the mth common attention relationship, M m1 represents the attention weight of the mth common attention relationship, represents the weight of content similarity interval, U u represents the content similarity of the u-th content similarity interval, U u1 Represents the similarity interval weight of the u-th content similarity interval.

[0135] It can be understood that the browsing time period refers to the time interval during which the user browses the content in a day, the content type refers to the type of content browsed by the user, for example: news, sports, the browsing time interval refers to the time during which the user browses a certain content type, the browsing frequency refers to the number of times the user browses per day, the browsing device refers to the device end used for browsing the content, the browsing location refers to the geographical location when the content is browsed, the content nature refers to the newness of the content browsed by the user, and the browsing dimension refers to the combination of content type, browsing time interval, browsing frequency, browsing device, browsing location and content nature. For example, when the content type is news content, the browsing time interval is [10,20) min, the browsing frequency is greater than 3 times / day, the browsing device is a mobile phone end, the browsing location is a city, and the content nature is the latest information, it is a browsing dimension. The browsing preference of all browsing dimensions refers to the user's preference for each browsing dimension, and the preference degree refers to the degree of the user's preference. The first browsing feature value refers to the user's comprehensive preference degree under the conditions of different content types, browsing frequencies, browsing devices, browsing locations and content nature in one network domain, and the second browsing feature value refers to the user's comprehensive preference degree under the conditions of different content types, browsing frequencies, browsing devices, browsing locations and content nature in another network domain.

[0136] Exemplarily, assume that there are three browsing dimensions, τ=1, and the corresponding browsing time period is [0,6)h. The dimension weight of the first browsing dimension α1=0.3, and the browsing preference of the first browsing dimension β1=0.5. The dimension weight of the second browsing dimension α2=0.3, and the browsing preference of the second browsing dimension β2=0.3. The dimension weight of the third browsing dimension α3=0.4, and the browsing preference of the third browsing dimension β3=0.2. B1(j,k,l,v,w,f)=0.3×0.5+0.3×0.3+0.4×0.2=0.44, and the first browsing feature value or the second browsing feature value in the first browsing time period is 0.44.

[0137] Interpretable, interaction frequency refers to the number of times a user interacts per day, common friend relationships refer to the number of common friends of two users in one network domain and another network domain, friend interaction of interaction frequency refers to the user's activity at each interaction frequency, interaction frequency weight refers to the degree of influence of interaction frequency on the first friend's characteristic value and the second friend's characteristic value, and the corresponding interaction weight of interaction frequency refers to the proportion of each interaction frequency to all interaction frequencies. For example, the corresponding interaction weight for more than 1 time / day is 0.1, the corresponding interaction weight for more than 3 times / day is 0.2, the corresponding interaction weight for more than 5 times / day is 0.3, and the corresponding interaction weight for more than 10 times / day is 0.4. Common friend weight refers to the influence of common friend relationships on the first friend's characteristic value and The influence degree of the second friend characteristic value, the friend relationship degree of the common friend relationship refers to the user's activity in each common friend relationship, the relationship weight of the common friend relationship refers to the proportion of each common friend relationship to all common friend relationships, for example, the relationship weight of more than 1 common friend is 0.1, the relationship weight of more than 3 common friends is 0.2, the relationship weight of more than 5 common friends is 0.3, and the relationship weight of more than 10 common friends is 0.4. The first friend characteristic value refers to the comprehensive activity of the user under different interaction frequencies and different common friend relationship conditions in one network domain, and the second friend characteristic value refers to the comprehensive activity of the user under different interaction frequencies and different common friend relationship conditions in another network domain.

[0138] Explainably, the common attention relationship refers to the number of hobbies that two users in one network domain and another network domain pay common attention to, the content similarity interval refers to the interval of the similarity of the content browsed by two users in one network domain and another network domain, the common attention weight refers to the influence of the common attention relationship on the first attention feature value and the second attention feature value, the common attention degree of the common attention relationship refers to the user's activity under each common attention relationship, the attention weight of the common attention relationship refers to the proportion of each common attention relationship to all common attention relationships, the content similarity interval weight refers to the influence of the content similarity interval on the first attention feature value and the second attention feature value, the content similarity of the content similarity interval refers to the similarity of the content browsed by two users in one network domain and another network domain, and the corresponding similarity interval weight of the content similarity interval refers to the proportion of each content similarity interval to all content similarity intervals.

[0139] S3. Extracting a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature, and a first attention feature.

[0140] It can be understood that the first user feature refers to the active features, browsing features, friend features and attention features of each user in the first user feature set, and the second user feature refers to the active features, browsing features, friend features and attention features of each user in the second user feature set. The first active feature refers to the active feature of each user in the first user feature set, the first browsing feature refers to the browsing feature of each user in the first user feature set, the first friend feature refers to the friend feature of each user in the first user feature set, and the first attention feature refers to the attention feature of each user in the first user feature set.

[0141] S4. Extracting second user features in sequence from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature.

[0142] It can be understood that the second active feature refers to the active feature of each user in the second user feature set, the second browsing feature refers to the browsing feature of each user in the second user feature set, the second friend feature refers to the friend feature of each user in the second user feature set, and the second attention feature refers to the attention feature of each user in the second user feature set.

[0143] S5. Draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph respectively using the first active feature, the first browsing feature, the second active feature, and the second browsing feature.

[0144] It can be understood that the first active feature graph refers to a line graph that divides a day into four time periods and uses the four time periods as the x-axis and the first active feature value as the y-axis, the first browsing feature graph refers to a line graph that divides a day into four time periods and uses the four time periods as the x-axis and the first browsing feature value as the y-axis, the second active feature graph refers to a line graph that divides a day into four time periods and uses the four time periods as the x-axis and the second active feature value as the y-axis, the second browsing feature graph refers to a line graph that divides a day into four time periods and uses the four time periods as the x-axis and the second browsing feature value as the y-axis.

[0145] Specifically, using the first active feature, the first browsing feature, the second active feature, and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph, respectively, includes:

[0146] Constructing a first active time data matrix based on the active time period model and the first active feature value;

[0147] Draw a first active feature graph according to the first active time data matrix, wherein the x-axis of the first active feature graph represents the active time period, and the y-axis represents the first active feature value;

[0148] Constructing a first browsing preference data matrix based on the browsing preference model and the first browsing feature value;

[0149] Draw a first browsing characteristic graph according to the first browsing preference data matrix, wherein the x-axis of the first browsing characteristic graph represents the browsing time period, and the y-axis represents the first browsing characteristic value;

[0150] constructing a second active time data matrix based on the active time period model and the second active feature value;

[0151] Draw a second active feature graph according to the second active time data matrix, wherein the x-axis of the second active feature graph represents the active time period, and the y-axis represents the second active feature value;

[0152] Constructing a second browsing preference data matrix based on the browsing preference model and the second browsing characteristic value;

[0153] A second browsing characteristic graph is drawn according to the second browsing preference data matrix, wherein the x-axis of the second browsing characteristic graph represents the browsing time period, and the y-axis represents the second browsing characteristic value.

[0154] It can be understood that the first active time data matrix refers to a matrix composed of first active eigenvalues, the first browsing preference data matrix refers to a matrix composed of first browsing eigenvalues, the second active time data matrix refers to a matrix composed of second active eigenvalues, and the second browsing preference data matrix refers to a matrix composed of second browsing eigenvalues.

[0155] Specifically, constructing a first active time data matrix based on the active time period model and the first active feature value includes:

[0156] Acquire the activity state and the working state, wherein the activity state includes: high activity state, medium activity state, low activity state and inactive state, and the working state includes: working and resting;

[0157] constructing a first active time data matrix using the active state and the working state;

[0158] The first active time data matrix is ​​as follows:

[0159]

[0160] Among them, R refers to the first active time data matrix, T 11 It refers to the first active feature value when the active state is high active state and the working state is working in the e-th active time period, T 12 Refers to the first active feature value when the active state is high active state and the working state is resting in the e-th active time period, T 41It refers to the first active feature value when the active state is inactive and the working state is working in the e-th active time period, T 42 It refers to the first active feature value when the active state is inactive and the working state is resting in the e-th active time period.

[0161] S6. Perform homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity.

[0162] It can be understood that performing homogeneity processing refers to the action of calculating the homogeneity of active features and browsing features based on the first active feature graph and the second active feature graph, based on the first browsing feature graph and the second browsing feature graph. The active feature homogeneity and browsing feature homogeneity refer to the degree of similarity between the first active feature graph and the second active feature graph, and the first browsing feature graph and the second browsing feature graph. The higher the homogeneity, the more likely it is that the two user identities in the two network domains are the same natural person.

[0163] Specifically, performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity includes:

[0164] Obtain the time weight of the active time period, the second active feature value of the active time period, and the maximum possible distance of the feature value, and calculate the homogeneity of the active feature based on the time weight of the active time period, the second active feature value, and the maximum possible distance of the feature value:

[0165]

[0166] Among them, H1 represents the homogeneity of active features, Q e represents the time weight of the e-th active time period, T e represents the first active feature value of the e-th active time period, T e ' represents the second active feature value of the e-th active time period, D max Represents the maximum possible distance of eigenvalues.

[0167] It can be understood that the time weight of all active time periods refers to the weights of four active time periods, for example, the weight of the first active time period is 0.2, the weight of the first active time period is 0.4, the weight of the first active time period is 0.3, and the weight of the first active time period is 0.1. The second active feature value of all active time periods refers to the second active feature value of all active time periods under another network domain, and the maximum possible distance of the feature value refers to the maximum difference between the first active feature value and the second active feature value under all active event segments.

[0168] S7: Determine whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold.

[0169] It can be understood that the homogeneity threshold refers to a preset critical value used to measure the similarity between the active feature and the browse feature.

[0170] If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, the process returns to the step of sequentially extracting the second user features from the second user feature set.

[0171] If the activity feature homogeneity and the browsing feature homogeneity are all greater than the homogeneity threshold, S8 is executed to calculate the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature.

[0172] Explainably, the second iteration user features refer to the active features, browsing features, friend features, and attention features of the second user in the second user feature set. The friend feature homogeneity refers to the degree of similarity between the first friend feature and the second friend feature, and the attention feature homogeneity refers to the degree of similarity between the first attention feature and the second attention feature.

[0173] S9. Calculate mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights.

[0174] It can be understood that feature weight refers to the importance of friend feature homogeneity, attention feature homogeneity, browsing feature homogeneity and active feature homogeneity, and mixed homogeneity refers to an overall homogeneity that is formed by combining friend feature homogeneity, attention feature homogeneity, browsing feature homogeneity and active feature homogeneity by assigning different weights.

[0175] Specifically, the mixed homogeneity is calculated based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights, including:

[0176] The mixed homogeneity is calculated using the feature weights according to the following formula, wherein the feature weights include: active feature weight, browsing feature weight, friend feature weight, and attention feature weight:

[0177] H m =ω d ×H1+ω z ×H2+ω f ×H3+ω c ×H4

[0178] Among them, H m represents the mixing homogeneity, ω d represents the active feature weight, ωz represents the browsing feature weight, ω f represents the friend feature weight, ω c represents the weight of the attention feature, H2 represents the homogeneity of the browsing feature, H3 represents the homogeneity of the friend feature, and H4 represents the homogeneity of the attention feature.

[0179] It can be understood that the active feature weight refers to the influence of the active feature homogeneity on the mixed homogeneity, the browsing feature weight refers to the influence of the browsing feature homogeneity on the mixed homogeneity, the friend feature weight refers to the influence of the friend feature homogeneity on the mixed homogeneity, and the attention feature weight refers to the influence of the attention feature homogeneity on the mixed homogeneity.

[0180] S10, constructing a characteristic function, inputting the mixed homogeneity into the characteristic function, and obtaining a function output value.

[0181] Explainable, characteristic function refers to a function for judging whether two user identities in two network domains are the same natural person based on mixed homogeneity and mixed homogeneity threshold, mixed homogeneity threshold refers to a preset critical value used to measure the degree of similarity, and function output value refers to the output used to judge whether two user identities in two network domains are the same natural person.

[0182] In detail, the construction characteristic function includes:

[0183] Construct the characteristic function based on the preset mixed homogeneity threshold:

[0184]

[0185] Among them, f(H m ) represents the characteristic function, H y Represents the mixture homogeneity threshold.

[0186] Exemplarily, assuming that the mixing homogeneity threshold is 0.9 and the mixing homogeneity is 0.92, the characteristic function outputs 1.

[0187] S11. Determine whether the function output value is 1.

[0188] If the function output value is not 1, return to the above step of sequentially extracting the second user features from the second user feature set.

[0189] If the function output value is 1, S12 is executed to confirm that the first user feature and the second user feature complete the social network identity association.

[0190] The present invention solves the problem described in the background technology. By combining the user identity big data collected from two network domains, constructing multiple user feature models and comprehensively processing the user identity big data, the accurate association of social network identities is achieved. First, the user identity big data of the two network domains is collected, and an active time period model, a browsing preference model, a friend relationship model and a follow-up relationship model are constructed to achieve comprehensive capture of user behaviors and characteristics, accurately describe the user's behavior patterns in different network domains, and improve the comprehensiveness and representativeness of the data. Then, a first active feature graph, a first browsing feature graph, a second active feature graph and a second browsing feature graph are drawn, and homogeneity processing is performed, so that the similarity of the behaviors of two users in different network domains can be clearly displayed. By calculating the active The feature homogeneity and browsing feature homogeneity are calculated, and it is judged whether they meet the preset homogeneity threshold, which can effectively improve the accuracy of identity recognition; afterwards, if the active feature homogeneity and browsing feature homogeneity do not all meet the preset threshold, the second user feature set is iteratively updated. This iterative update can continuously optimize user features and improve the efficiency of social network identity association; further, by calculating the friend feature homogeneity and the attention feature homogeneity, and calculating the mixed homogeneity based on the preset feature weight, multiple feature dimensions of the user are comprehensively considered, providing more comprehensive and accurate data for the final identity confirmation; finally, by constructing a feature function and inputting the mixed homogeneity into the feature function, the matching of the user identity can be automatically judged, so that the social network identity association has a higher accuracy. Therefore, the present invention can improve the association accuracy of social network identities.

[0191] like Figure 2 , which is a functional module diagram of a social network identity association system based on big data provided by an embodiment of the present invention.

[0192] The social network identity association system 100 based on big data of the present invention can be installed in an electronic device. According to the functions implemented, the social network identity association system 100 based on big data can include a first user feature collection module 101, a homogeneity calculation module 102, a homogeneity threshold judgment module 103 and a social network identity association module 104. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0193] The first user feature collection module 101 is used to collect user identity big data of two network domains, build an active time period model, a browsing preference model, a friend relationship model and a follow relationship model, perform data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model and the follow relationship model to obtain a first user feature set and a second user feature set, and extract a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature and a first follow feature;

[0194] The homogeneity calculation module 102 is used to sequentially extract second user features from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature, and use the first active feature, the first browsing feature, the second active feature, and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph, respectively, and perform homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain an active feature homogeneity and a browsing feature homogeneity;

[0195] A homogeneity threshold judgment module 103 is used to judge whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold. If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, the module returns to the above-mentioned step of sequentially extracting the second user features from the second user feature set. If the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, the homogeneity of the friend feature and the homogeneity of the attention feature are calculated based on the first friend feature, the first attention feature, the second friend feature and the second attention feature.

[0196] The social network identity association module 104 is used to calculate the mixed homogeneity based on the active feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights, construct a feature function, input the mixed homogeneity into the feature function, obtain a function output value, and determine whether the function output value is 1. If the function output value is not 1, return to the above step of sequentially extracting the second user features from the second user feature set. If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

[0197] In detail, the modules in the social network identity association system 100 based on big data in the embodiment of the present invention are used in the same manner as described above. Figure 1 The technical means are the same as the social network identity association method based on big data described in, and can produce the same technical effects, so they will not be repeated here.

[0198] like Figure 3 , is a schematic diagram of the structure of an electronic device for implementing a social network identity association method based on big data provided by an embodiment of the present invention.

[0199] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a social network identity association method program based on big data.

[0200] The memory 11 includes at least one type of readable storage medium, including flash memory, mobile hard disk, multimedia card, card-type memory (e.g., SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card, etc. equipped on the electronic device 1. Further, the memory 11 also includes an internal storage unit of the electronic device 1 and an external storage device. The memory 11 may be used not only to store application software and various types of data installed on the electronic device 1, such as the code of the social network identity association method program based on big data, but also to temporarily store data that has been output or is to be output.

[0201] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules (such as social network identity association method programs based on big data, etc.) stored in the memory 11, and calls data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

[0202] The bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0203] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0204] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, etc. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0205] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices.

[0206] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0207] The social network identity association method program based on big data stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0208] Collect user identity big data from two network domains to build active time period models, browsing preference models, friend relationship models, and attention relationship models;

[0209] Performing data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model, and the attention relationship model to obtain a first user feature set and a second user feature set;

[0210] Extracting a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature, and a first attention feature;

[0211] Extracting second user features in sequence from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature;

[0212] Draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph respectively using the first active feature, the first browsing feature, the second active feature, and the second browsing feature;

[0213] Performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity;

[0214] Determining whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold;

[0215] If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0216] If the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculating the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature;

[0217] Calculate mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights;

[0218] Constructing a characteristic function, inputting the mixed homogeneity into the characteristic function, and obtaining a function output value;

[0219] Determine whether the function output value is 1;

[0220] If the function output value is not 1, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0221] If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

[0222] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0223] Furthermore, if the module / unit integrated in the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0224] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:

[0225] Collect user identity big data from two network domains to build active time period models, browsing preference models, friend relationship models, and attention relationship models;

[0226] Performing data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model, and the attention relationship model to obtain a first user feature set and a second user feature set;

[0227] Extracting a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature, and a first attention feature;

[0228] Extracting second user features in sequence from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature;

[0229] Draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph respectively using the first active feature, the first browsing feature, the second active feature, and the second browsing feature;

[0230] Performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity;

[0231] Determining whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold;

[0232] If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0233] If the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculating the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature;

[0234] Calculate mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights;

[0235] Constructing a characteristic function, inputting the mixed homogeneity into the characteristic function, and obtaining a function output value;

[0236] Determine whether the function output value is 1;

[0237] If the function output value is not 1, returning to the above step of sequentially extracting the second user features from the second user feature set;

[0238] If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

[0239] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative, and actual implementation may have other division methods.

[0240] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0241] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0242] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0243] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A social network identity association method based on big data, characterized in that: The method comprises: Collect user identity big data from two network domains to build active time period models, browsing preference models, friend relationship models, and attention relationship models; Performing data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model, and the attention relationship model to obtain a first user feature set and a second user feature set; Extracting a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature, and a first attention feature; Extracting second user features in sequence from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature; Draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph respectively using the first active feature, the first browsing feature, the second active feature, and the second browsing feature; Performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity; Determining whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold; If the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, returning to the above step of sequentially extracting the second user features from the second user feature set; If the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculating the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature; Calculate mixed homogeneity based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights; Constructing a characteristic function, inputting the mixed homogeneity into the characteristic function, and obtaining a function output value; Determine whether the function output value is 1; If the function output value is not 1, returning to the above step of sequentially extracting the second user features from the second user feature set; If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

2. The social network identity association method based on big data according to claim 1, characterized in that: The collecting of user identity big data of two network domains includes: Obtaining API authentication mechanisms of the two network domains, and sending preset API requests to the two network domains based on the API authentication mechanisms of the two network domains and the pre-obtained API keys, obtaining request results, and determining whether the API request successfully connects the two network domains based on the request results; If the API request fails to successfully connect the two network domains, return to the above step of obtaining the API authentication mechanisms of the two network domains; If the API request successfully connects the two network domains, constructing a first API request using preset initial paging parameters, and acquiring first request data based on the initial paging parameters and the first API request; Determine whether the preset total paging parameter is equal to 1; If the total paging parameter is equal to 1, the first request data is used as user identity big data; If the total paging parameter is not equal to 1, determining whether the initial paging parameter is equal to the total paging parameter; If the initial paging parameter is not equal to the total paging parameter, the initial paging parameter is increased by one to obtain an updated paging parameter, the initial paging parameter is updated using the updated paging parameter, and then the update API request is constructed using the initial paging parameter, and the update request data is obtained based on the initial paging parameter and the update API request, and the step of determining whether the initial paging parameter is equal to the total paging parameter is returned to the above step; If the initial paging parameter is equal to the total paging parameter, the update request data and the first request data are merged in a form of the first request data first and the update request data later to obtain user identity big data.

3. The social network identity association method based on big data as claimed in claim 2, characterized in that: The construction process of the active time period model is as follows: Obtain the active state weight, work state weight, state activity, and work activity in each active time period, wherein the active time period may be: [0,6)h, [6,16)h, [16,20)h, 20,24)h; The active time period model is constructed using the active state weight, work state weight, state activity and work activity: T e =ε×X e +∈×Y e Among them, T e represents the first active feature value or the second active feature value in the e-th active time period, e represents the sequence number of the active time period, ε represents the active state weight, and X e represents the state activity in the e-th active time period, ∈ represents the work state weight, and Y e Indicates the work activity in the e-th active time period.

4. The social network identity association method based on big data as claimed in claim 3, characterized in that: The construction process of the browsing preference model, the friend relationship model and the attention relationship model is as follows: Obtain browsing time period, content type, browsing time interval, browsing frequency, browsing device, browsing location and content nature, wherein the browsing time period includes: [0,6)h, [6,16)h, [16,20)h, [20,24)h, the content type includes: news content, entertainment content, science and technology content, literary content and educational content, the browsing time interval includes: [10,20)min, [20,40)min, [40,60)min,, [60,+∞)min, the browsing frequency includes: greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 7 times / day, greater than 7 times / day and less than or equal to 9 times / day and greater than 9 times / day, the browsing device includes: mobile phone and PC, the browsing location includes: city and suburbs, the content nature includes: latest information and past content; Arrange and combine the browsing time period, content type, browsing duration interval, browsing frequency, browsing device, browsing location and content nature to obtain a browsing dimension; Obtaining the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension, and constructing a browsing preference model based on the browsing dimension, the browsing preference of the browsing dimension, the dimension ordinal number of the browsing dimension, and the dimension weight of the browsing dimension; The browsing preference model is as follows: Among them, B τ (j, k, l, v, w, f) represents the first browsing feature value or the second browsing feature value in the τth browsing time period and when the browsing dimension is (j, k, l, v, w, f), τ represents the ordinal number of the browsing time period, j represents the content type, k represents the browsing duration interval, l represents the browsing frequency, v represents the browsing device, w represents the browsing location, f represents the content nature, n represents the dimension ordinal number of the browsing dimension, α i represents the dimension weight of the i-th browsing dimension, β i Indicates the browsing preference of the i-th browsing dimension; Obtain interaction frequency, common friend relationship, friend interaction degree of interaction frequency, interaction frequency weight, interaction weight of interaction frequency, friend relationship degree of common friend relationship, common friend weight and relationship weight of common friend relationship, wherein the interaction frequency includes: greater than 1 time / day and less than or equal to 3 times / day, greater than 3 times / day and less than or equal to 5 times / day, greater than 5 times / day and less than or equal to 10 times / day and greater than 10 times / day, and the common friend relationship includes: greater than 1 common friend and less than or equal to 3 common friends, greater than 3 common friends and less than or equal to 5 common friends, greater than 5 common friends and less than or equal to 10 common friends and greater than 10 common friends; Constructing a friend relationship model based on the interaction frequency, common friend relationship, friend interaction degree of the interaction frequency, interaction frequency weight, interaction weight of the interaction frequency, friend relationship degree of the common friend relationship, common friend weight and relationship weight of the common friend relationship; The friendship model is as follows: Among them, A represents the first friend's characteristic value or the second friend's characteristic value, θ represents the interaction frequency weight, and B b represents the friend interaction degree of the bth interaction frequency, B b1 represents the interaction weight of the bth interaction frequency, μ represents the common friend weight, O m represents the friendship degree of the mth common friendship, O m1 represents the relationship weight of the mth common friend relationship; Acquire a common attention relationship, a content similarity interval, a common attention weight, a common attention degree of the common attention relationship, an attention weight of the common attention relationship, a content similarity interval weight, a content similarity of the content similarity interval, and a similarity interval weight of the content similarity interval, wherein the common attention relationship includes: more than 1 common attention and less than or equal to 3 common attentions, more than 3 common attentions and less than or equal to 5 common attentions, more than 5 common attentions and less than or equal to 10 common attentions, and more than 10 common attentions, and the content similarity interval includes: [0, 10%), [10%, 30%), [30%, 60%), and [60%, 100%); Constructing the attention relationship model using the common attention relationship, content similarity interval, common attention weight, common attention degree of the common attention relationship, attention weight of the common attention relationship, content similarity interval weight, content similarity of the content similarity interval and similarity interval weight of the content similarity interval; The following relationship model is as follows: Where Y represents the first attention feature value or the second attention feature value, ρ represents the joint attention weight, and M m represents the common attention degree of the mth common attention relationship, M m1 represents the attention weight of the mth common attention relationship, represents the weight of content similarity interval, U u represents the content similarity of the u-th content similarity interval, U u1 Represents the similarity interval weight of the u-th content similarity interval.

5. The social network identity association method based on big data as claimed in claim 4, characterized in that: The using the first active feature, the first browsing feature, the second active feature and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph and a second browsing feature graph respectively includes: Constructing a first active time data matrix based on the active time period model and the first active feature value; Draw a first active feature graph according to the first active time data matrix, wherein the x-axis of the first active feature graph represents the active time period, and the y-axis represents the first active feature value; Constructing a first browsing preference data matrix based on the browsing preference model and the first browsing feature value; Draw a first browsing characteristic graph according to the first browsing preference data matrix, wherein the x-axis of the first browsing characteristic graph represents the browsing time period, and the y-axis represents the first browsing characteristic value; Constructing a second active time data matrix based on the active time period model and the second active feature value; Draw a second active feature graph according to the second active time data matrix, wherein the x-axis of the second active feature graph represents the active time period, and the y-axis represents the second active feature value; Constructing a second browsing preference data matrix based on the browsing preference model and the second browsing characteristic value; A second browsing characteristic graph is drawn according to the second browsing preference data matrix, wherein the x-axis of the second browsing characteristic graph represents the browsing time period, and the y-axis represents the second browsing characteristic value.

6. The social network identity association method based on big data as claimed in claim 5, characterized in that: The constructing a first active time data matrix based on the active time period model and the first active feature value includes: Acquire the activity state and the working state, wherein the activity state includes: high activity state, medium activity state, low activity state and inactive state, and the working state includes: working and resting; constructing a first active time data matrix using the active state and the working state; The first active time data matrix is ​​as follows: Among them, R refers to the first active time data matrix, T 11 Refers to the first active feature value when the active state is high active state and the working state is working in the e-th active time period, T 12 It refers to the first active feature value when the active state is high active state and the working state is resting in the e-th active time period, T 41 Refers to the first active feature value when the active state is inactive and the working state is working in the e-th active time period, T 42 It refers to the first active feature value when the active state is inactive and the working state is resting in the e-th active time period.

7. The social network identity association method based on big data according to claim 6, characterized in that: The performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain active feature homogeneity and browsing feature homogeneity includes: Obtain the time weight of the active time period, the second active feature value of the active time period, and the maximum possible distance of the feature value, and calculate the homogeneity of the active feature based on the time weight of the active time period, the second active feature value, and the maximum possible distance of the feature value: Among them, H1 represents the homogeneity of active features, Q e represents the time weight of the e-th active time period, T e represents the first active feature value of the e-th active time period, T e ' represents the second active feature value of the e-th active time period, D max Represents the maximum possible distance of eigenvalues.

8. The social network identity association method based on big data according to claim 7, characterized in that: The mixed homogeneity is calculated based on the activity feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights, including: The mixed homogeneity is calculated using the feature weights according to the following formula, wherein the feature weights include: active feature weight, browsing feature weight, friend feature weight, and attention feature weight: H m =ω d ×H1+ω z ×H2+ω f ×H3+ω c ×H4 Among them, H m represents the mixing homogeneity, ω d represents the active feature weight, ω z represents the browsing feature weight, ω f represents the friend feature weight, ω c represents the weight of the attention feature, J2 represents the homogeneity of the browsing feature, J3 represents the homogeneity of the friend feature, and H4 represents the homogeneity of the attention feature.

9. The social network identity association method based on big data as claimed in claim 1, characterized in that: The constructing characteristic function comprises: Construct the characteristic function based on the preset mixed homogeneity threshold: Among them, f(H m ) represents the characteristic function, H y Represents the mixture homogeneity threshold.

10. A social network identity association system based on big data, characterized in that: The system comprises: A first user feature collection module is used to collect user identity big data of two network domains, construct an active time period model, a browsing preference model, a friend relationship model and a follow-up relationship model, perform data processing on the user identity big data based on the active time period model, the browsing preference model, the friend relationship model and the follow-up relationship model to obtain a first user feature set and a second user feature set, and extract a first user feature based on the first user feature set, wherein the first user feature includes: a first active feature, a first browsing feature, a first friend feature and a first follow-up feature; a homogeneity calculation module, configured to sequentially extract second user features from the second user feature set, wherein the second user features include: a second active feature, a second browsing feature, a second friend feature, and a second attention feature; using the first active feature, the first browsing feature, the second active feature, and the second browsing feature to draw a first active feature graph, a first browsing feature graph, a second active feature graph, and a second browsing feature graph, respectively; performing homogeneity processing based on the first active feature graph, the first browsing feature graph, the second active feature graph, and the second browsing feature graph to obtain an active feature homogeneity and a browsing feature homogeneity; a homogeneity threshold judgment module, used to judge whether the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than a preset homogeneity threshold; if the homogeneity of the active feature and the homogeneity of the browsing feature are not all greater than the homogeneity threshold, then return to the above step of sequentially extracting the second user features from the second user feature set; if the homogeneity of the active feature and the homogeneity of the browsing feature are all greater than the homogeneity threshold, then calculate the friend feature homogeneity and the attention feature homogeneity based on the first friend feature, the first attention feature, the second friend feature and the second attention feature; The social network identity association module is used to calculate the mixed homogeneity based on the active feature homogeneity, browsing feature homogeneity, friend feature homogeneity, attention feature homogeneity and preset feature weights, construct a feature function, input the mixed homogeneity into the feature function, obtain a function output value, and determine whether the function output value is 1. If the function output value is not 1, return to the above-mentioned step of sequentially extracting the second user features from the second user feature set. If the function output value is 1, it is confirmed that the first user feature and the second user feature complete the social network identity association.

Images