Data association degree analysis method and device, computer equipment and storage medium

Abstract

The invention relates to a data association degree analysis method and device, computer equipment and a storage medium. The method comprises the steps of obtaining user data of each dimension of a target person from a database; extracting attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtaining a time factor of the attribute data, calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtaining a characteristic value set of the attribute data in the user data of all dimensions; and inputting the feature value set into a trained multi-layer perceptron model to obtain a correlation degree score output by the trained multi-layer perceptron model, the correlation degree score being used for representing the correlation degree between the attribute data and the target person. According to the method, the relevance between the low-relevance data or the current relevance failure data and the related personnel can be analyzed, so that case handling personnel can quickly and accurately extract and identify the real portrait information of the target personnel.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, computer equipment and storage medium for analyzing data correlation. Background Art

[0002] During the case analysis process, the suspect's electronic evidence data, logistics and express delivery data, vehicle registration data, social registration data, etc. are generally analyzed to identify the suspect's name, mobile phone number, address, license plate number and other identity information, providing clues for solving the case. However, it is unknown to what extent these name, mobile phone number, address, license plate number and other identity information are related to the suspect's data.

[0003] The current common technical practice is to simply classify the degree of correlation of data according to the data source. In the absence of accurate information on the degree of data correlation, some data with low correlation or data with currently invalid correlation can easily lead to misleading and erroneous judgments in case investigations. Summary of the invention

[0004] Based on this, it is necessary to provide a data correlation analysis method, device, computer equipment and storage medium to address the above technical problems, and determine the correlation between data by automatically analyzing data correlation, so as to analyze the correlation between data with low correlation or data with invalid current correlation and relevant personnel, and provide reliable judgment basis for public security personnel on the degree of data correlation of criminal suspects and whether it is applicable in the case scenario during case investigation.

[0005] A data association analysis method comprises: obtaining user data of each dimension of a target person from a database; extracting attribute data of the target person from the user data of each dimension of the target person according to data attribute type; obtaining a time factor of the attribute data, calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtaining a characteristic value set of the attribute data in the user data of all dimensions; inputting the characteristic value set into a trained multilayer perceptron model to obtain a correlation score output by the trained multilayer perceptron model, wherein the trained multilayer perceptron model is obtained by model training through the characteristic values ​​of sample attribute data and the correlation between the sample attribute data and the sample person, and the correlation score is used to characterize the correlation between the attribute data and the target person.

[0006] In one of the embodiments, the user data of each dimension of the target person is collected based on the user data of each dimension configured by the social activities of the target person; and the data attribute type is set based on the user data attribute of the target person.

[0007] In one of the embodiments, the user data of the target person in various dimensions includes social tag data, travel data, and social data; the attribute data of the target person includes user identity data and social activity data.

[0008] In one embodiment, calculating the characteristic value of the attribute data in each dimension of user data according to the time factor of the attribute data includes: Calculate the characteristic value of the attribute data in the user data of each dimension; wherein m represents the number of resource data in the user data of each dimension, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0009] In one of the embodiments, the model training of the multilayer perceptron model includes: collecting characteristic values ​​of multiple sample attribute data and the correlation between each sample attribute data and the sample person to obtain a sample data set; dividing the sample data set into a training set, a validation set and a test set; training the multilayer perceptron model through the training set, and evaluating the results output from the model training process of the multilayer perceptron model through the validation set to obtain a preliminarily trained multilayer perceptron model; testing the preliminarily trained multilayer perceptron model through the test set, and obtaining a trained multilayer perceptron model after the test passes.

[0010] In one of the embodiments, a method for analyzing data correlation also includes: obtaining multiple classified case information of the target person determined based on case investigation; associating the correlation score of the attribute data with the multiple classified case information to perform application analysis through the associated display.

[0011] In one embodiment, the multiple classified and processed case information includes any number of personnel holographic files, personnel number and item associations, number-personnel translations, mobile phone evidence data analysis reports, personnel vehicle information analysis, personnel involvement analysis, and relationship network personnel information.

[0012] A device for analyzing data association degree, comprising: an acquisition module, used for acquiring user data of each dimension of a target person from a database; an extraction module, used for extracting attribute data of a target person from the user data of each dimension of the target person according to data attribute type; a calculation module, used for acquiring a time factor of the attribute data, calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtaining a characteristic value set of the attribute data in the user data of all dimensions; an analysis module, used for inputting the characteristic value set into a trained multilayer perceptron model, and obtaining a correlation score output by the trained multilayer perceptron model, wherein the trained multilayer perceptron model is obtained by model training through the characteristic values ​​of sample attribute data and the correlation between the sample attribute data and the sample person, and the correlation score is used to characterize the correlation between the attribute data and the target person.

[0013] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of any of the above-mentioned embodiments are implemented.

[0014] A computer-readable storage medium stores a computer program, which implements the steps of any of the above-mentioned method embodiments when the computer program is executed by a processor.

[0015] The above-mentioned data association analysis method, device, computer equipment and storage medium obtain the user data of each dimension of the target person from the database; extract the attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtain the time factor of the attribute data, calculate the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtain the characteristic value set of the attribute data in the user data of all dimensions; input the characteristic value set into the trained multi-layer perceptron model to obtain the association score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the characteristic values ​​of the sample attribute data and the association between the sample attribute data and the sample person, and the association score is used to characterize the association between the attribute data and the target person.

[0016] Therefore, the attribute data of each dimension is added with the time factor to calculate the eigenvalue of each dimension, and the data correlation is analyzed and calculated through the multi-layer perceptron model to obtain the correlation between the attribute data and the target person. This can provide a reliable basis for public security personnel to judge the degree of data correlation between suspects and whether it is applicable in the case scenario during the case investigation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 An application environment diagram of a data correlation analysis method in an embodiment;

[0018] Figure 2A schematic diagram of a flow chart of a method for analyzing data relevance in an embodiment;

[0019] Figure 3 is a schematic diagram of a model structure of a multilayer perceptron model in one embodiment;

[0020] Figure 4 is a structural block diagram of a data association analysis device in one embodiment;

[0021] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0023] The present application provides a data correlation analysis method, which is applied to Figure 1 In the application environment shown. Among them, the correlation analysis platform 102 is used to execute a data correlation analysis method of the present application. Specifically, the correlation analysis platform 102 communicates with the database system 104 through the network. , obtain the user data of each dimension of the target person from the database; extract the attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtain the time factor of the attribute data, calculate the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtain the characteristic value set of the attribute data in the user data of all dimensions; input the characteristic value set into the trained multi-layer perceptron model to obtain the correlation score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through sample attribute data and the correlation between the sample attribute data and the sample personnel, and the correlation score is used to characterize the correlation between the attribute data and the target person. Among them, the correlation analysis platform 102 can be implemented with an independent server or a server cluster composed of multiple servers.

[0024] In one embodiment, Figure 2 As shown in the figure, a method for analyzing data association is provided. Figure 1 The correlation analysis platform 102 in FIG. 1 is used as an example to illustrate the method, which includes the following steps:

[0025] S202, obtaining user data of various dimensions of the target person from the database.

[0026] In this embodiment, the database system stores data information of multiple persons, and the target person may be a suspect in a criminal case. In this application, the database system is accessed and the user data of each dimension of the target person is obtained from the database for correlation analysis. In one application scenario, it is mainly used by relevant departments to analyze the data of suspects in related cases, and to provide a reliable basis for public security personnel to judge the degree of data correlation of suspects and whether it is applicable in the case scenario during the case investigation process.

[0027] In one example, user data of each dimension of the target person is obtained by collecting user data of each dimension configured according to the social activities of the target person.

[0028] S204: extracting attribute data of the target person from the user data of each dimension of the target person according to the data attribute type.

[0029] In this embodiment, for each dimension of the obtained user data, data extraction is performed on the user data of each dimension according to the data attribute type. In one example, the data attribute type is set based on the user data attribute of the target person.

[0030] In one example, the user data of the target person in various dimensions includes social tag data, travel data, and social data; the attribute data of the target person includes user identity data and social activity data.

[0031] Specifically, social tag data refers to the tags that target persons are marked with when participating in social activities. Travel data refers to the data generated when target persons travel during social activities. Social data refers to the data generated when target persons participate in social activities. User data is divided according to dimensions such as social tags, travel, and social activities. Furthermore, user data is further divided according to user identity data and social activity data to obtain corresponding attribute data. S206, obtaining a time factor of the attribute data, calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtaining a characteristic value set of the attribute data in the user data of all dimensions.

[0032] In this embodiment, the time factor of the attribute data can be the time when the attribute data is recorded. For example, the generation time of the attribute data, or the collection time of the attribute data, etc. The corresponding characteristic value is calculated for each attribute data plus its corresponding time factor, and the characteristic value set of each attribute data in all dimensions of user data is obtained. For example, take the attribute data as a mobile phone number as an example: the advertising information dimension includes a house rental advertisement 1, a personnel recruitment advertisement 2, and a used car sale advertisement 3. Each advertisement contains the mobile phone number of the suspect and the time factor of the mobile phone number. The characteristic value of the mobile phone number in the advertising information dimension is calculated based on the time factor of the mobile phone number in each advertisement. Similarly, the characteristic value of the mobile phone number in other dimensions of user data can be calculated, so the characteristic value set of the mobile phone number in all dimensions of user data can be obtained. Among them, the mobile phone number and advertising information are collected by legal and compliant means.

[0033] In one embodiment, calculating the characteristic value of the attribute data in each dimension of user data according to the time factor of the attribute data includes: Calculate the characteristic value of the attribute data in the user data of each dimension; wherein m represents the number of resource data in the user data of each dimension, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0034] Specifically, the calculation formula for calculating the corresponding eigenvalues ​​of different attribute data plus the time factor is:

[0035]

[0036] Among them, m represents the number of resource data in the user data of each dimension, the resource data includes resource information of attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0037] Take the attributes of mobile phone numbers in the user data of the advertising information dimension as an example. The user data of the advertising information dimension contains three resource data: house rental advertisement 1, personnel recruitment advertisement 2, and used car sale advertisement 3. Then the number of resource data m = 3, and the time when a person's mobile phone number was recorded in these three resource data is T1, T2, and T3; according to the formula, the characteristic value F of a person's mobile phone number in the user data of the advertising information dimension is finally extracted.

[0038] Finally, the feature value set of a person's mobile phone number in all dimensions is represented by the following vector, where the values ​​in brackets are the feature values ​​of the user in each dimension:

[0039] [2.2,0.85,3.45,3.25,1.9,3.25,1.65,2.2,2.2,0.85,3.45,3.25,1.9,3.25,1.65,1.85].

[0040] S208, inputting the feature value set into the trained multi-layer perceptron model to obtain the correlation score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the feature values ​​of the sample attribute data and the correlation between the sample attribute data and the sample personnel, and the correlation score is used to characterize the correlation between the attribute data and the target personnel.

[0041] In this embodiment, the trained multi-layer perceptron model is obtained by model training through the characteristic values ​​of the sample attribute data and the correlation between the sample attribute data and the sample personnel. Therefore, the trained multi-layer perceptron model can output the relevant correlation score when the characteristic values ​​of the attribute data are input in the application scenario.

[0042] Specifically, Figure 3 As shown in the figure, the MLP multilayer perceptron model is a feedforward neural network consisting of an input layer, several hidden layers, and an output layer. Each layer consists of multiple neurons. It is generally used for classification problems and can be trained through the back propagation algorithm. In the field of deep learning, it is widely used in image recognition, natural language processing and other fields.

[0043] The MLP multilayer perceptron model is a feedforward neural network that consists of multiple layers, including an input layer, a hidden layer, and an output layer. Each layer consists of multiple neurons (or nodes).

[0044] Input Layer: The first layer of a multilayer perceptron is the input layer, which receives the raw data or features as input.

[0045] Hidden layer: A multilayer perceptron can contain one or more hidden layers, each of which consists of multiple neurons. The neurons in the hidden layer receive the output of the previous layer, perform a weighted sum on it, and perform a nonlinear transformation through an activation function.

[0046] Output layer: The last layer of the multilayer perceptron is the output layer, which determines the number of neurons depending on the type of problem. For classification problems, neurons with softmax activation function are usually used for output; for regression problems, linear activation function or identity function is usually used for output.

[0047] Through the backpropagation algorithm, the weights and bias terms can be adjusted according to the difference between the network output and the expected output to minimize the network's loss function. In this way, the multilayer perceptron can learn and approximate complex nonlinear functional relationships.

[0048] In one embodiment, the model training of the multilayer perceptron model includes: collecting multiple sample attribute data and the correlation between each sample attribute data and the sample person to obtain a sample data set; dividing the sample data set into a training set, a validation set and a test set; training the multilayer perceptron model through the training set, and evaluating the results output by the model training process of the multilayer perceptron model through the validation set to obtain a preliminarily trained multilayer perceptron model; testing the preliminarily trained multilayer perceptron model through the test set, and obtaining a trained multilayer perceptron model after the test passes.

[0049] Specifically, the model training process of the multilayer perceptron model is as follows:

[0050] 1) Data Collection

[0051] The debugging process of the MLP multi-layer perceptron model requires a large amount of relevant data for training, verification, and testing. First, a data collection strategy is formulated, which involves determining the time, frequency, method, and scope of data collection. Specifically, user data of various dimensions is collected, and a suitable data storage solution is selected for data management and access methods, and data documents are written and metadata information of the data is recorded, including information such as data source, data format, meaning of data fields, and data quality assessment.

[0052] 2) Normalization of data eigenvalues

[0053] The eigenvalues ​​of the data, as the input parameters of the MLP multilayer perceptron model, first need to be normalized; normalization is a scaling technique in which the eigenvalues ​​of the data are moved and rescaled so that their final range is between 0 and 1.

[0054] Normalized formula:

[0055]

[0056] Among them, X represents any eigenvalue, X' represents the eigenvalue after normalization of X, and X max and X min are the maximum and minimum values ​​in the eigenvalue set, respectively.

[0057] When the value of X is the minimum value in the column, the numerator will be 0, so X' is 0; on the other hand, when the value of X is the maximum value in the column, the numerator is equal to the denominator, so the value of X' is 1; if the value of X is between the minimum and maximum values, the value of X' is between 0 and 1.

[0058] 3) Data labeling

[0059] Data labeling is the process of providing labeled training data for deep learning of the MLP multilayer perceptron model, which has a significant impact on the training and performance of the MLP multilayer perceptron model. Before labeling, the data needs to be cleaned to eliminate quality issues such as missing values, noisy data, and duplicate data. Clarify labeling requirements, classification standards, labeling methods, etc. to ensure that all labelers have a consistent understanding of the labeling rules.

[0060] 4) Divide the dataset

[0061] The data is classified by data attribute type such as name, mobile phone number, etc., and divided into training set, validation set and test set. The training set is used to establish the MLP multi-layer perceptron prediction model. At the same time, the validation set is evaluated and predictions are made based on it. The model parameters of the MLP multi-layer perceptron model can be tuned. After multiple training and verification parameter tuning, the final appropriate MLP multi-layer perceptron model parameters are obtained.

[0062] 5) Training process

[0063] 1. Import the required libraries, install and import the required program libraries.

[0064] 2. Define the model and use the Sequential model to define a multi-layer perceptron model. You can add multiple hidden layers, and each hidden layer can have a different number of nodes.

[0065] 3. Compile the model, specify the loss function, optimizer, and evaluation metrics.

[0066] 4. Train the model, fit the model using the training data, set the batch size and number of iterations.

[0067] 5. Evaluate model performance. Use test data to evaluate the performance of the model. You can view indicators such as accuracy and AUC-ROC.

[0068] 6. Grid search, use grid search GridSearchCV() to find the best parameter combination.

[0069] 7. Retrain the model using the best parameters and obtain the best parameters through multiple cycles of parameter adjustment and training.

[0070] Specifically, based on the above steps, the evaluation indicators are set, the model is fitted using training data, and the batch size and number of iterations are set. These specific settings are continuously tested, modified, and adjusted during multiple training processes, and the optimal parameter values ​​are finally obtained as shown in the following table:

[0071]

[0072]

[0073] The above-mentioned method for analyzing data association degree obtains the user data of each dimension of the target person from the database; extracts the attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtains the time factor of the attribute data, calculates the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtains the characteristic value set of the attribute data in the user data of all dimensions; inputs the characteristic value set into the trained multi-layer perceptron model to obtain the association score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the characteristic value of the sample attribute data and the association degree between the sample attribute data and the sample person, and the association score is used to characterize the association degree between the attribute data and the target person. Therefore, the characteristic value of each dimension is calculated by adding the time factor to the attribute data of each dimension, and the data association degree is analyzed and calculated by the multi-layer perceptron model to obtain the association degree between the attribute data and the target person. In an application scenario, it can provide a reliable basis for public security personnel to judge the degree of data association of the suspect during the case investigation process and whether it is applicable in the case scenario.

[0074] In one embodiment, after the above step S208, that is, after outputting the correlation score, it also includes: obtaining multiple classified case information of the target person based on case investigation; associating the correlation score of the attribute data with the multiple classified case information to perform application analysis through the associated display.

[0075] Among them, multiple classified and processed case information includes personnel holographic files, personnel number and item association, number personnel translation, mobile phone evidence data analysis report, personnel vehicle information analysis, personnel case analysis, and any number of relationship network personnel information. The goal of calculating the correlation score of the target personnel attribute data can not only be associated with the case information, but also be used for application analysis through the associated display, such as extracting the target personnel's most realistic information, drawing a portrait of the target personnel, and other applications.

[0076] Specifically, after the data's relevance score is calculated through the trained MLP multi-layer perceptron model, it is integrated with public security-related functional applications such as personnel holographic files, personnel-number-item association, number-personnel translation, mobile phone evidence collection data analysis reports, personnel-vehicle information analysis, personnel-involved-case analysis, and relationship network personnel information mining for calculation and display. The operation interface provides functions such as sorting, filtering, and aggregation statistics by data's relevance score.

[0077] A data correlation analysis method in each of the above embodiments extracts characteristic values ​​by adding a time factor to multi-dimensional user data, analyzes and calculates the data correlation through an MLP multi-layer perceptron model, and obtains a data correlation degree score between information such as name, mobile phone number, address, license plate number and a person's ID card. Accurate data correlation score calculation results are obtained through multiple model training and parameter tuning, which helps case handlers to quickly and accurately extract and identify the target person's real portrait information, and provides a reliable basis for public security personnel to judge the degree of data correlation of suspects and whether it is applicable in the case scenario during case investigation.

[0078] It should be understood that, although the steps in the flowchart are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0079] The present application also provides a data correlation analysis device. Figure 4 As shown, a data association analysis device includes: an acquisition module 402, an extraction module 404, a calculation module 406 and an analysis module 408. The acquisition module 402 is used to acquire the user data of each dimension of the target person from the database; the extraction module 404 is used to extract the attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; the calculation module 406 is used to acquire the time factor of the attribute data, calculate the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtain the characteristic value set of the attribute data in the user data of all dimensions; the analysis module 408 is used to input the characteristic value set into the trained multi-layer perceptron model to obtain the association score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the characteristic value of the sample attribute data and the association degree between the sample attribute data and the sample person, and the association score is used to characterize the association degree between the attribute data and the target person.

[0080] In one of the embodiments, the user data of each dimension of the target person is collected based on the user data of each dimension configured by the social activities of the target person; and the data attribute type is set based on the user data attribute of the target person.

[0081] In one of the embodiments, the user data of the target person in various dimensions includes social tag data, travel data, and social data; the attribute data of the target person includes user identity data and social activity data.

[0082] In one embodiment, calculating the characteristic value of the attribute data in each dimension of user data according to the time factor of the attribute data includes: Calculate the characteristic value of the attribute data in the user data of each dimension; wherein m represents the number of resource data in the user data of each dimension, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0083] In one of the embodiments, the model training of the multilayer perceptron model includes: collecting characteristic values ​​of multiple sample attribute data and the correlation between each sample attribute data and the sample person to obtain a sample data set; dividing the sample data set into a training set, a validation set and a test set; training the multilayer perceptron model through the training set, and evaluating the results output from the model training process of the multilayer perceptron model through the validation set to obtain a preliminarily trained multilayer perceptron model; testing the preliminarily trained multilayer perceptron model through the test set, and obtaining a trained multilayer perceptron model after the test passes.

[0084] In one of the embodiments, a data correlation analysis device also includes a correlation display module, which is used to obtain multiple classified case information of a target person determined based on case investigation; the correlation score of the attribute data is associated with the multiple classified case information to display, so as to perform application analysis through the associated display.

[0085] In one embodiment, the multiple classified and processed case information includes any number of personnel holographic files, personnel number and item associations, number-personnel translations, mobile phone evidence data analysis reports, personnel vehicle information analysis, personnel involvement analysis, and relationship network personnel information.

[0086] For the specific definition of a data correlation analysis device, please refer to the definition of a data correlation analysis method mentioned above, which will not be repeated here. Each module in the above-mentioned data correlation analysis device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0087] In one embodiment, a computer device is provided, which may be Figure 1The internal structure of the server in the correlation analysis platform 102 shown in FIG. Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store multiple user data. The network interface of the computer device is used to communicate with the database system through a network connection. When the computer program is executed by the processor, a method for analyzing data association is implemented.

[0088] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: obtaining user data of each dimension of a target person from a database; extracting attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtaining a time factor of the attribute data, calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtaining a characteristic value set of the attribute data in the user data of all dimensions; inputting the characteristic value set into a trained multilayer perceptron model to obtain a correlation score output by the trained multilayer perceptron model, wherein the trained multilayer perceptron model is obtained by model training using the characteristic values ​​of sample attribute data and the correlation between the sample attribute data and the sample person, and the correlation score is used to characterize the correlation between the attribute data and the target person.

[0089] In one of the embodiments, the user data of each dimension of the target person is collected based on the user data of each dimension configured by the social activities of the target person; and the data attribute type is set based on the user data attribute of the target person.

[0090] In one of the embodiments, the user data of the target person in various dimensions includes social tag data, travel data, and social data; the attribute data of the target person includes user identity data and social activity data.

[0091] In one embodiment, when the processor executes the computer program to implement the above-mentioned step of calculating the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, the following steps are specifically implemented: Calculate the characteristic value of the attribute data in the user data of each dimension; wherein m represents the number of resource data in the user data of each dimension, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0092] In one of the embodiments, when the processor executes a computer program to implement model training of a multilayer perceptron model, the following steps are specifically implemented: characteristic values ​​of multiple sample attribute data and the correlation between each sample attribute data and a sample person are collected to obtain a sample data set; the sample data set is divided into a training set, a validation set, and a test set; the multilayer perceptron model is trained using the training set, and the results output from the model training process of the multilayer perceptron model are evaluated using the validation set to obtain a preliminarily trained multilayer perceptron model; the preliminarily trained multilayer perceptron model is tested using the test set, and a trained multilayer perceptron model is obtained after the test passes.

[0093] In one of the embodiments, when the processor executes the computer program, it specifically implements the following steps: obtaining multiple classified case information of the target person determined based on case investigation; associating the correlation score of the attribute data with the multiple classified case information to perform application analysis through the associated display.

[0094] In one embodiment, the multiple classified and processed case information includes any number of personnel holographic files, personnel number and item associations, number-personnel translations, mobile phone evidence data analysis reports, personnel vehicle information analysis, personnel involvement analysis, and relationship network personnel information.

[0095] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: obtaining user data of each dimension of a target person from a database; extracting attribute data of the target person from the user data of each dimension of the target person according to the data attribute type; obtaining a time factor of the attribute data, and calculating a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data to obtain a characteristic value set of the attribute data in the user data of all dimensions; inputting the characteristic value set into a trained multilayer perceptron model to obtain a correlation score output by the trained multilayer perceptron model, wherein the trained multilayer perceptron model is obtained by model training through the characteristic values ​​of sample attribute data and the correlation between the sample attribute data and the sample person, and the correlation score is used to characterize the correlation between the attribute data and the target person.

[0096] In one of the embodiments, the user data of each dimension of the target person is collected based on the user data of each dimension configured by the social activities of the target person; and the data attribute type is set based on the user data attribute of the target person.

[0097] In one of the embodiments, the user data of the target person in various dimensions includes social tag data, travel data, and social data; the attribute data of the target person includes user identity data and social activity data.

[0098] In one embodiment, when the computer program is executed by the processor to implement the above-mentioned step of calculating the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, the following steps are specifically implemented: Calculate the characteristic value of the attribute data in the user data of each dimension; wherein m represents the number of resource data in the user data of each dimension, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

[0099] In one of the embodiments, when the computer program is executed by a processor to implement model training of a multilayer perceptron model, the following steps are specifically implemented: characteristic values ​​of multiple sample attribute data and the correlation between each sample attribute data and a sample person are collected to obtain a sample data set; the sample data set is divided into a training set, a validation set, and a test set; the multilayer perceptron model is trained using the training set, and the results output from the model training process of the multilayer perceptron model are evaluated using the validation set to obtain a preliminarily trained multilayer perceptron model; the preliminarily trained multilayer perceptron model is tested using the test set, and a trained multilayer perceptron model is obtained after the test passes.

[0100] In one of the embodiments, when the computer program is executed by a processor, the following steps are specifically implemented: obtaining multiple classified case information of the target person determined based on case investigation; associating the correlation score of the attribute data with the multiple classified case information to perform application analysis through the associated display.

[0101] In one embodiment, the multiple classified and processed case information includes any number of personnel holographic files, personnel number and item associations, number-personnel translations, mobile phone evidence data analysis reports, personnel vehicle information analysis, personnel involvement analysis, and relationship network personnel information.

[0102] It should be noted that in the embodiments of the present application, data related to user information or user data (for example, social tag data, travel data and social data, user identity data and social activity data, etc.) need to be obtained and processed after the user's authorization and consent, and all the data obtained by the user comply with the relevant policies and laws. A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0103] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for analyzing data association, characterized in that: The method comprises: Obtain user data of various dimensions of the target person from the database; Extracting attribute data of the target person from user data of each dimension of the target person according to data attribute type; Acquire a time factor of the attribute data, calculate a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtain a characteristic value set of the attribute data in the user data of all dimensions; The feature value set is input into a trained multi-layer perceptron model to obtain a correlation score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the feature values ​​of sample attribute data and the correlation between the sample attribute data and the sample personnel, and the correlation score is used to characterize the correlation between the attribute data and the target personnel.

2. The method according to claim 1, characterized in that The user data of each dimension of the target person is obtained by collecting user data of each dimension of the social activity configuration of the target person; and the data attribute type is set based on the user data attribute of the target person.

3. The method according to claim 2, characterized in that The user data of the target person in various dimensions includes social tag data, travel data and social data; The attribute data of the target person includes user identity data and social activity data.

4. The method according to claim 1, characterized in that: The calculating the characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data includes: according to Calculating characteristic values ​​of the attribute data in user data of each dimension; Among them, m represents the number of resource data in each dimension of user data, the resource data contains resource information of the attribute data, i represents the i-th resource data, and Ti represents the recording time of the attribute data in the i-th resource data.

5. The method according to claim 1, characterized in that The model training of the multi-layer perceptron model includes: Collect characteristic values ​​of multiple sample attribute data and the correlation between each sample attribute data and the sample personnel to obtain a sample data set; Divide the sample data set into training set, validation set and test set; Performing model training on the multilayer perceptron model through the training set, and evaluating the results output from the model training process of the multilayer perceptron model through the validation set, so as to obtain a preliminarily trained multilayer perceptron model; The preliminarily trained multilayer perceptron model is tested by using the test set, and the trained multilayer perceptron model is obtained after the test passes.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining multiple classified and processed case information of the target person based on case investigation; The relevance score of the attribute data is associated with the plurality of classified case information and displayed, so as to perform application analysis through the associated display.

7. The method according to claim 6, characterized in that The multiple classified and processed case information includes any number of personnel holographic files, personnel number and item associations, number-personnel translations, mobile phone evidence data analysis reports, personnel vehicle information analysis, personnel involvement analysis, and relationship network personnel information.

8. A data correlation analysis device, characterized in that: The device comprises: The acquisition module is used to obtain the user data of each dimension of the target person from the database; An extraction module, used to extract attribute data of the target person from the user data of each dimension of the target person according to data attribute type; A calculation module, used to obtain a time factor of the attribute data, calculate a characteristic value of the attribute data in the user data of each dimension according to the time factor of the attribute data, and obtain a characteristic value set of the attribute data in the user data of all dimensions; An analysis module is used to input the feature value set into a trained multi-layer perceptron model to obtain a correlation score output by the trained multi-layer perceptron model, wherein the trained multi-layer perceptron model is obtained by model training through the feature values ​​of sample attribute data and the correlation between the sample attribute data and the sample personnel, and the correlation score is used to characterize the correlation between the attribute data and the target person.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images