User data processing method and system, storage medium and electronic equipment

Abstract

The invention relates to a user data processing method and system, a storage medium and electronic equipment, and relates to the technical field of big data process.The method comprises the steps that data cleaning is conducted on original user attribute data of to-be-classified users to obtain target user attribute data, the to-be-classified users are subjected to bucket dividing according to the target user attribute data, and the buckets are obtained; obtaining a user bucket dividing result; determining an attribute association relationship between user entities corresponding to the to-be-classified users included in the same user bucket dividing result, and constructing an entity triple according to the attribute association relationship and the user entities; determining collaborative association edges among the user entities according to the entity triples, and constructing a user association feature map based on the collaborative association edges, the user entities and the target user attribute data; and inputting the user association feature map into a preset user classification model to obtain a user classification result, and determining the user category of the to-be-classified user according to the user classification result. The accuracy of the user classification result is improved.

Description

Technical Field

[0001] This disclosure relates to the field of big data processing technology, and more specifically, to a user data processing method, a user data processing apparatus, a computer-readable storage medium, and an electronic device. Background Technology

[0002] Existing user classification methods can only construct relationships and classify users based on their original attributes; however, the relationships obtained by this method have the problem of a single relationship dimension, which cannot reveal implicit collaborative relationships based on user behavior, thus resulting in low accuracy of user classification results.

[0003] It should be noted that the information in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this disclosure is to provide a method for processing user data, a device for processing user data, a computer-readable storage medium, and an electronic device, thereby overcoming, at least to some extent, the problem of low accuracy of user classification results due to limitations and defects in related technologies.

[0005] According to one aspect of this disclosure, a method for processing user data is provided, comprising: cleaning the original user attribute data of users to be classified to obtain target user attribute data, and binning the users to be classified according to the target user attribute data to obtain user binning results; determining the attribute association relationships between user entities corresponding to the users to be classified included in the same user binning results, and constructing entity triples according to the attribute association relationships and user entities; determining collaborative association edges between user entities according to the entity triples, and constructing a user association feature map based on the collaborative association edges, user entities, and target user attribute data; inputting the user association feature map into a preset user classification model to obtain user classification results, and determining the user category of the users to be classified according to the user classification results.

[0006] In one exemplary embodiment of this disclosure, the user to be classified is bucketed according to target user attribute data to obtain user bucketing results, including: determining a primary attribute dimension from the target attribute dimension corresponding to the target user attribute data, and obtaining the data recording time of the primary attribute data corresponding to the primary attribute dimension; sorting the users to be classified according to the data recording time to obtain user sorting results, and determining the time difference between the data recording times of two adjacent users to be classified in the user sorting results; determining the data storage bucket to which the user to be classified belongs according to the time difference to obtain the user bucketing results; wherein, each user to be classified corresponds to one data storage bucket.

[0007] In one exemplary embodiment of this disclosure, determining the data storage bucket to which the user to be classified belongs based on the time difference to obtain the user bucketing result includes: in response to the time difference being less than or equal to a preset time threshold, determining that the user to be classified belongs to the same data storage bucket, and writing the target user attribute data of the user to be classified into the same data storage bucket; in response to the time difference being greater than the preset time threshold, determining that the user to be classified belongs to different data storage buckets, and writing the target user attribute data of the user to be classified into different data storage buckets; and generating the user bucketing result based on the user to be classified written in the data storage bucket and the target user attribute data associated with the user to be classified.

[0008] In one exemplary embodiment of this disclosure, determining the attribute association relationships between user entities corresponding to the user to be classified, included in the same user bucketing result, includes: traversing the user entities corresponding to the user to be classified, included in the same user bucketing result, determining the user attribute values ​​of the target user attribute data corresponding to the user entity, and performing embedding mapping processing on the user attribute values ​​to obtain the attribute feature vector of the user entity corresponding to the user to be classified in the target attribute dimension; determining the attribute association relationships between user entities based on the attribute feature vector; wherein, the attribute association relationships include at least one of user name association relationships, user unique identifier association relationships, user IP address association relationships, user device fingerprint association relationships, user address information association relationships, and user-associated personnel information association relationships.

[0009] In an exemplary embodiment of this disclosure, constructing entity triples based on the attribute association and user entities includes: determining a group of user entities with attribute association; wherein the user entity group includes a first user entity and a second user entity; determining a target attribute dimension corresponding to the attribute association; wherein the target user dimension includes at least one of a user name dimension, a user unique identifier dimension, a user device fingerprint dimension, a user address information dimension, and a user associated personnel information dimension; constructing the entity triples based on the first user entity, the second user entity, and the target attribute dimension; wherein each entity triple corresponds to one target attribute dimension.

[0010] In an exemplary embodiment of this disclosure, determining the collaborative association edge between user entities based on the entity triplet includes: determining the number of entity groups of user entity groups between the first user entity and the second user entity, and determining whether the number of entity groups is greater than a preset number threshold; in response to the number of entity groups being greater than the preset number threshold, determining that there is a collaborative association edge between the first user entity and the second user entity.

[0011] In one exemplary embodiment of this disclosure, constructing a user association feature graph based on the collaborative association edges, user entities, and target user attribute data includes: using the user entity as a first vertex, using the collaborative association edges between user entities as first connecting edges, and determining the weight value of the collaborative association edges according to the number of entity groups of user entity groups between the user entities; using the target user attribute data as a second vertex, and using the correspondence between the user entity and the target user attribute data as a second connecting edge; and constructing the user association feature graph based on the first vertex, the first connecting edge and the weight value, the second vertex and the second connecting edge.

[0012] In one exemplary embodiment of this disclosure, inputting the user association feature map into a preset user classification model to obtain a user classification result includes: inputting the user association feature map into a preset community detection algorithm model to perform similarity fusion and community optimization on the user association feature map based on the preset community detection algorithm model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map; and / or, inputting the user association feature map into a preset graph convolutional neural network model to perform node feature updates on the user association feature map based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map.

[0013] In an exemplary embodiment of this disclosure, the preset graph convolutional neural network model includes a first graph convolutional layer, a second graph convolutional layer, ..., an Nth graph convolutional layer, and a classification layer. The process of updating node features of the user association feature map based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map, includes: updating node features of the user association feature map based on the first graph convolutional layer to obtain a first graph convolutional processing result; updating node features of the first graph convolutional processing result based on the second graph convolutional layer to obtain a second graph convolutional processing result; repeating the determination process of the second graph convolutional processing result sequentially to obtain a third graph convolutional processing result, a fourth graph convolutional processing result, ..., an Nth graph convolutional processing result; using the Nth graph convolutional processing result as the optimized user association feature map, and performing classification processing on the optimized user association feature map based on the classification layer to obtain the user classification result.

[0014] According to one aspect of this disclosure, a user data processing system is provided, comprising: a data acquisition system, a data preprocessing system, and a Spark computing engine, wherein the data preprocessing system is communicatively connected to the data acquisition system and the Spark computing engine respectively; wherein: the data acquisition system is used to acquire raw user attribute data of users to be classified, and send the raw user attribute data to the data preprocessing system; the data preprocessing system is used to perform data cleaning on the raw user attribute data to obtain target user attribute data, and to bin the users to be classified according to the target user attribute data to obtain user binning results; the Spark computing engine is used to determine the attribute association relationships between user entities corresponding to the users to be classified included in the same user binning results, and to construct entity triples according to the attribute association relationships and user entities; and to determine collaborative association edges between user entities according to the entity triples, and to construct a user association feature map based on the collaborative association edges, user entities, and target user attribute data; and to input the user association feature map into a preset user classification model to obtain user classification results, and to determine the user category of the users to be classified according to the user classification results.

[0015] According to one aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the user data processing method described in any of the preceding claims.

[0016] According to one aspect of this disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform a user data processing method as described in any one of the preceding claims by executing the executable instructions.

[0017] This disclosure provides a user data processing method. On one hand, it involves cleaning the original user attribute data of users to be classified to obtain target user attribute data, and then binning the users to be classified based on the target user attribute data to obtain user binning results. Next, it determines the attribute association relationships between user entities corresponding to the users to be classified within the same user binning result, and constructs entity triples based on the attribute association relationships and user entities. Then, it determines collaborative association edges between user entities based on the entity triples, and constructs a user association feature map based on the collaborative association edges, user entities, and target user attribute data. Finally, it inputs the user association feature map into a preset user classification model to obtain user classification results. This method achieves association of user entities from multiple attribute dimensions, thus solving the problem in existing technologies where association can only be performed from a single dimension, leading to the inability to reveal implicit collaborative relationships based on user behavior and resulting in low accuracy of user classification results. This improves the accuracy of the obtained user classification results. On the other hand, since the user category of the user to be classified can be determined based on the user classification results, and then it can be determined whether the user is a risky user based on the obtained user category, the problem of user fraud can be avoided.

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0020] Figure 1 The diagram illustrates a flowchart of a user data processing method according to an example embodiment of the present disclosure.

[0021] Figure 2 The illustration shows a scenario example diagram of a user data processing system according to an exemplary embodiment of the present disclosure.

[0022] Figure 3 The diagram schematically illustrates an example structure of a preset graph convolutional neural network model according to an exemplary embodiment of the present disclosure.

[0023] Figure 4 The diagram illustrates a scenario example of user bucketing results obtained according to an exemplary embodiment of this disclosure.

[0024] Figure 5 An example diagram illustrating a triplet obtained according to an exemplary embodiment of the present disclosure is shown.

[0025] Figure 6 An example diagram illustrating a collaborative associated edge obtained according to an exemplary embodiment of the present disclosure is shown.

[0026] Figure 7 The illustration schematically shows a user association feature map obtained according to an example embodiment of the present disclosure.

[0027] Figure 8 The illustration shows a scenario example diagram illustrating a specific process for determining a user classification result according to an example embodiment of the present disclosure.

[0028] Figure 9 This schematically illustrates a block diagram of a user data processing apparatus according to an exemplary embodiment of the present disclosure.

[0029] Figure 10 The diagram schematically illustrates an example structure of an electronic device for implementing a user data processing method according to an example embodiment of the present disclosure. Detailed Implementation

[0030] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0031] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0032] In some related solutions, anti-fraud gang collaborative relationships can be constructed based on original user attributes. Specifically, in practical applications, the original attributes of graph nodes can be used to construct relationships, forming a triple of <node 1, node 2, attribute>. However, this method has the following drawbacks: First, the relationships constructed based on original attributes have a single dimension, thus failing to reveal implicit collaborative relationships based on behavior. Second, relationships constructed based on original attributes can only discover relationships with directly shared attributes, thus failing to identify entity relationships that do not have directly shared attributes but are closely linked through business behavior logic. Furthermore, it also fails to identify business behavior relationships within directly shared attributes. However, in practical applications, these two types of relationships represent collaborative relationships with strong logical connections manifested through business behavior. Therefore, if entity relationships and business behavior relationships cannot be further identified, it is difficult to deal with organized and highly concealed gang activities.

[0033] Based on this, this exemplary embodiment first provides a method for processing user data, which can run on a server, server cluster, or cloud server, etc. Of course, those skilled in the art can also run the method disclosed herein on other platforms as needed, and this exemplary embodiment does not impose any special limitations on this. Specifically, refer to... Figure 1 As shown, the method for processing this user data may include the following steps:

[0034] Step S110. Perform data cleaning on the original user attribute data of the users to be classified to obtain target user attribute data, and divide the users to be classified into buckets according to the target user attribute data to obtain user bucketing results;

[0035] Step S120. Determine the attribute association relationship between the user entities corresponding to the user to be classified included in the same user bucketing result, and construct entity triples based on the attribute association relationship and the user entities;

[0036] Step S130. Determine the collaborative association edges between user entities based on the entity triples, and construct a user association feature graph based on the collaborative association edges, user entities, and target user attribute data;

[0037] Step S140. Input the user association feature map into a preset user classification model to obtain the user classification result, and determine the user category of the user to be classified based on the user classification result.

[0038] In the user data processing method described above, on the one hand, target user attribute data is obtained by cleaning the original user attribute data of the users to be classified, and the users to be classified are binned according to the target user attribute data to obtain user binning results; then, the attribute association relationships between the user entities corresponding to the users to be classified included in the same user binning results are determined, and entity triples are constructed according to the attribute association relationships and user entities; then, collaborative association edges between user entities are determined according to the entity triples, and user association feature maps are constructed based on collaborative association edges, user entities, and target user attribute data; finally, the user association feature maps are input into a preset user classification model to obtain user classification results. This realizes the association of user entities from multiple attribute dimensions, thereby solving the problem that existing technologies can only associate from a single dimension, which leads to the inability to reveal implicit collaborative relationships based on user behavior, resulting in low accuracy of user classification results, and improving the accuracy of the obtained user classification results. On the other hand, since the user category of the user to be classified can be determined according to the user classification results, and then it can be determined whether the user is a risky user based on the obtained user category, the problem of avoiding user fraud can be avoided.

[0039] The following will provide a detailed explanation and description of the user data processing method described in the exemplary embodiments of this disclosure, in conjunction with the accompanying drawings.

[0040] First, the technical implementation principles of the exemplary embodiments of this disclosure will be explained and described. Specifically, the user data processing method described in the exemplary embodiments of this disclosure addresses the core deficiency of existing technologies in effectively identifying implicit collaborative relationships based on dynamic business behaviors. By introducing time windows and statistical indicators on top of static attributes, a collaborative relationship that reflects the behavioral logic between entities is constructed. On this basis, the time efficiency of relationship construction under large-scale data is improved through algorithm optimization such as bucketing coding and aggregation calculation. Furthermore, by inputting the constructed collaborative relationship edges into a community detection algorithm, and then performing community scoring and individual scoring on the divided communities, the accuracy and optimization of the results of risk individual identification in risky communities are improved.

[0041] Specifically, in practical applications, the input data is first cleaned, removing missing and outlier values ​​from entity and attribute fields. Then, data records are bucketed based on time attributes, and collaborative relationships are built between entities within each bucket. A graph model is established based on these collaborative relationships, and a community partitioning algorithm is used to identify high-probability initial fraud groups, significantly improving the accuracy of fraud group identification. Based on this, an efficient and scalable method for constructing collaborative relationships can be implemented for data volumes ranging from tens of millions to hundreds of millions. Furthermore, through bucketing strategies and parallel computing optimization, the computational efficiency problem of generating relationships from massive data downloads can be solved, ensuring the feasibility and timeliness of the method in real-world business scenarios.

[0042] The following will explain and describe the user data processing system involved in the exemplary embodiments of this disclosure. Specifically, refer to... Figure 2 As shown, the user data processing system may include a data acquisition system 210, a data preprocessing system 220, and a Spark computing engine 230. Specifically, the data preprocessing system is communicatively connected to both the data acquisition system and the Spark computing engine. In practical applications, the data acquisition system can collect raw user attribute data of users to be classified and send the raw user attribute data to the data preprocessing system. The data preprocessing system can clean the raw user attribute data to obtain target user attribute data and bin the users to be classified based on the target user attribute data to obtain user binning results. The Spark computing engine can determine the attribute association relationships between user entities corresponding to the users to be classified in the same user binning results and construct entity triples based on the attribute association relationships and user entities. It can determine collaborative association edges between user entities based on the entity triples and construct a user association feature map based on the collaborative association edges, user entities, and target user attribute data. The user association feature map is input into a preset user classification model to obtain user classification results, and the user category of the users to be classified is determined based on the user classification results.

[0043] It should be noted that since the preprocessing of user attribute data and the data bucketing process are both implemented in the data processing system, the computational burden on the Spark computing engine can be greatly reduced, thereby improving data processing efficiency.

[0044] The following will explain and describe the preset graph convolutional neural network model involved in the exemplary embodiments of this disclosure. Specifically, refer to... Figure 3As shown, the preset graph convolutional neural network model may include a first graph convolutional layer, a second graph convolutional layer, ..., an Nth graph convolutional layer 301 and a classification layer 302; in the actual user classification process, the role of each model layer will be detailed later, and will not be elaborated further here.

[0045] The following, combined with Figure 2 as well as Figure 3 right Figure 1 The user data processing method shown will be further explained and illustrated. Specifically:

[0046] In step S110, the original user attribute data of the users to be classified is cleaned to obtain target user attribute data, and the users to be classified are bucketed according to the target user attribute data to obtain user bucketing results.

[0047] In this embodiment of the disclosure, the original user attribute data may include, but is not limited to, user name, user unique identifier, user IP address, user device fingerprint, user address information, and user-associated personnel information, etc.; wherein, the user name may include, but is not limited to, the username and password recorded in the webpage in the form of cookie information, etc.; the user unique identifier may be, for example, a mobile phone number, bank card number, ID card number, or email address, etc.; the user IP address may be the Internet Protocol address of the mobile or fixed terminal used by the user; the user device fingerprint may be the unique device identifier of the mobile or fixed terminal used by the user, such as the device serial number, or it may be generated based on other attribute information of the device (such as device model, device code, etc.), and this example does not impose any special restrictions on this; the user address information recorded here may include, for example, home landline number, home address, work phone number, work name and work address, etc.; the user-associated personnel information recorded here may include, but is not limited to, the phone number of immediate family members, the name of immediate family members, the phone number of contact person and the name of contact person, etc.

[0048] In this embodiment of the disclosure, the original user attribute data of users to be classified is cleaned to obtain target user attribute data, which can be achieved in the following way: First, the collected user data is converted into a table of entity fields, time fields, and attribute fields; then, data with empty or incorrect entity and attribute values ​​are removed to obtain target user attribute data.

[0049] Secondly, the users to be classified are bucketed according to the target user attribute data to obtain the user bucketing result. Specifically, this can be achieved as follows: determine the primary attribute dimension from the target attribute dimension corresponding to the target user attribute data, and obtain the data recording time of the primary attribute data corresponding to the primary attribute dimension; sort the users to be classified according to the data recording time to obtain the user sorting result, and determine the time difference between the data recording times of two adjacent users to be classified in the user sorting result; determine the data storage bucket to which the user to be classified belongs according to the time difference to obtain the user bucketing result; wherein, each user to be classified corresponds to one data storage bucket.

[0050] In one example embodiment, determining the data storage bucket to which the user to be classified belongs based on the time difference to obtain the user bucketing result can be achieved as follows: In response to the time difference being less than or equal to a preset time threshold, determining that the users to be classified belong to the same data storage bucket, and writing the target user attribute data of the users to be classified into the same data storage bucket; In response to the time difference being greater than the preset time threshold, determining that the users to be classified belong to different data storage buckets, and writing the target user attribute data of the users to be classified into different data storage buckets; Based on the users to be classified written in the data storage buckets and the target user attribute data associated with the users to be classified, generating the user bucketing result.

[0051] The following will further explain and illustrate the specific bucketing process. Specifically, in practical applications, the primary attribute dimension must first be determined. This dimension can be set according to actual needs. For example, if classification is based on bank card transaction history, the bank card number from the user's unique identifier can be used as the primary attribute dimension. Similarly, if classification is based on IP address, the user's IP address can be used as the primary attribute dimension. In practice, this can be determined according to actual needs; this example does not impose any special restrictions. Furthermore, given the existence of the primary attribute dimension, the data recording time of the corresponding attribute value can be determined. Then, bucketing is performed based on this data recording time. During bucketing, if the time interval between consecutive records is less than or equal to a preset time threshold X min, the batch of records is placed into a single data storage bucket. The preset time threshold mentioned here can be determined according to actual needs; this example does not impose any special restrictions.

[0052] For example, assuming the primary attribute dimension is IP address, and multiple accounts correspond to the same IP address, the records for that IP address are first sorted in descending order of time. The time interval between the current record and the previous record is then calculated to be less than or equal to a set threshold (X min). If so, the records are grouped into the same bucket. Based on this, the final buckets constructed based on time intervals are obtained, with no overlap between records in each bucket. The resulting user bucketing results can be referenced... Figure 4 As shown.

[0053] In step S120, the attribute association relationships between the user entities corresponding to the user to be classified included in the same user bucketing results are determined, and entity triples are constructed based on the attribute association relationships and the user entities.

[0054] In the example embodiments of this disclosure, firstly, the attribute association relationships between user entities corresponding to the user to be classified, included in the same user bucketing results, are determined. Specifically, this can be achieved as follows: traversing the user entities corresponding to the user to be classified, included in the same user bucketing results, determining the user attribute values ​​of the target user attribute data corresponding to the user entity, and performing embedding mapping processing on the user attribute values ​​to obtain the attribute feature vector of the user entity corresponding to the user to be classified in the target attribute dimension; determining the attribute association relationships between user entities based on the attribute feature vector; wherein, the attribute association relationships described herein may include, but are not limited to, user name association relationships, user unique identifier association relationships, user IP address association relationships, user device fingerprint association relationships, user address information association relationships, and user-associated personnel information association relationships, etc. In other words, in practical applications, for user entities included in the same user bucketing result, if any attribute dimension has the same user attribute value, then it is determined that there is an attribute association relationship between the two user entities; for example, if the IP attribute feature vectors of the first user entity (e.g., user A) and the second user entity (e.g., user B) are the same, then there is an attribute association relationship; if the unit address attribute feature vectors are the same, then there is another attribute association relationship.

[0055] In one possible example embodiment, the specific process of determining the attribute feature vector can be based on a corresponding vector embedding model; wherein, the vector embedding model can be BERT or Embedding, or other neural network models, and this example does not impose any special restrictions on it.

[0056] Secondly, entity triples are constructed based on the attribute associations and user entities. Specifically, this can be achieved as follows: First, a group of user entities with attribute associations is determined, including a first user entity and a second user entity. Second, the target attribute dimension corresponding to the attribute association is determined, where the target user dimension may include, but is not limited to, user name dimension, user unique identifier dimension, user device fingerprint dimension, user address information dimension, and user associated personnel information dimension, etc. Third, entity triples are constructed based on the first user entity, the second user entity, and the target attribute dimension, where each entity triple corresponds to one target attribute dimension. In practical applications, nodes in the same bucket can be associated by attribute to form triples <node 1, node 2, attribute>. In practice, triples between different entities under each associated attribute can be generated sequentially. For example, for the IP attribute, if multiple different accounts have the same IP, then these accounts are associated with each other in the IP attribute, thus establishing a triple. Using a similar method, triples for all accounts in all the aforementioned associated attributes can be established. The resulting triples can be referenced... Figure 5 As shown.

[0057] In step S130, collaborative association edges between user entities are determined based on the entity triples, and a user association feature graph is constructed based on the collaborative association edges, user entities, and target user attribute data.

[0058] In this example embodiment, firstly, collaborative association edges between user entities are determined based on the entity triples. Specifically, this can be achieved as follows: determining the number of entity groups between the first user entity and the second user entity; determining whether the number of entity groups is greater than a preset threshold; and determining that a collaborative association edge exists between the first user entity and the second user entity in response to the number of entity groups being greater than the preset threshold. That is, in practical applications, a "collaboration strength" index can be calculated for node combinations with attribute relationships. This index value is the number of all different attribute triples connecting the two nodes. Then, a preset threshold (e.g., 2) is set. If the collaboration strength of two nodes reaches or exceeds this threshold, a collaborative association edge is established between them. It should be noted that the preset threshold used in this example can be determined according to actual needs, and this example does not impose any special restrictions on it. The obtained collaborative association edges can be referenced... Figure 6 As shown.

[0059] Secondly, a user association feature graph is constructed based on the collaborative association edges, user entities, and target user attribute data. Specifically, this can be achieved as follows: The user entity is used as the first vertex, the collaborative association edges between user entities are used as the first connecting edges, and the weight value of the collaborative association edges is determined according to the number of entity groups between the user entities; the target user attribute data is used as the second vertex, and the correspondence between the user entity and the target user attribute data is used as the second connecting edge; the user association feature graph is constructed based on the first vertex, the first connecting edge and its weight value, the second vertex and its second connecting edge. Specifically, in practical applications, the larger the number of entity groups between two user entities, the larger the weight value of the first connecting edge between the two user entities. The resulting user association feature graph can be referenced... Figure 7 As shown.

[0060] In step S140, the user association feature map is input into a preset user classification model to obtain the user classification result, and the user category of the user to be classified is determined based on the user classification result.

[0061] In this exemplary embodiment, firstly, the user association feature map is input into a preset user classification model to obtain a user classification result. Specifically, this can be achieved as follows: the user association feature map is input into a preset community detection algorithm model to perform similarity fusion and community optimization on the user association feature map based on the preset community detection algorithm model, resulting in an optimized user association feature map, and the user classification result is determined based on the optimized user association feature map; and / or, the user association feature map is input into a preset graph convolutional neural network model to perform node feature updates on the user association feature map based on the preset graph convolutional neural network model, resulting in an optimized user association feature map, and the user classification result is determined based on the optimized user association feature map.

[0062] In one possible example embodiment, taking the preset community detection algorithm model as Attributed Louvain as an example, the process of performing similarity fusion and community optimization on the user association feature map based on the preset community detection algorithm model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map, can be implemented in the following ways: First, attribute similarity calculation; specifically, for each first vertex, calculate the similarity between its attribute vector (i.e., the second vertex) and the attribute vectors of other first vertices. Common methods include cosine similarity or Euclidean distance to quantify the similarity at the attribute level; second, topological similarity calculation; specifically, the similarity of the topological structure can be evaluated based on the adjacency relationship of the network, through the sharing pattern of the node neighbors (such as Jaccard similarity) or random walk; further, similarity fusion; specifically, attribute similarity and topological similarity can be weighted and fused to generate a comprehensive similarity matrix between the first vertices, and the weights can be dynamically adjusted based on data characteristics or optimization objectives; finally, modularity optimization; specifically, in Attributed Louvain... Under the Louvain framework, the modularity is iteratively optimized through a greedy strategy, which includes two stages: the first stage assigns nodes to the local optimal community to improve the modularity; the second stage treats the community as a super node and repeats the first stage to form a multi-level community structure; finally, the community division is output; that is, the community division (i.e., user classification result) based on the comprehensive similarity and modularity optimization results can be output.

[0063] In an exemplary embodiment, updating the node features of the user association feature map based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map, can be implemented as follows: updating the node features of the user association feature map based on the first graph convolutional layer to obtain a first graph convolutional processing result, and updating the node features of the first graph convolutional processing result based on the second graph convolutional layer to obtain a second graph convolutional processing result; repeating the determination process of the second graph convolutional processing result sequentially to obtain a third graph convolutional processing result, a fourth graph convolutional processing result, ..., an Nth graph convolutional processing result; using the Nth graph convolutional processing result as the optimized user association feature map, and performing classification processing on the optimized user association feature map based on the classification layer to obtain the user classification result. The specific determination process can be found in [reference needed]. Figure 8 As shown.

[0064] Thus, the user data processing method described in the exemplary embodiments of this disclosure has been fully implemented. Based on the foregoing description, the user data processing method described in the exemplary embodiments of this disclosure has at least the following advantages: Firstly, the exemplary embodiments of this disclosure establish collaborative relationships between samples based on business data. Compared to general static attribute direct association, this relationship can capture dynamic relationships in the business and has stronger business significance. Secondly, the data utilization rate implemented in this exemplary embodiment is high, which can cope with organized and highly concealed gang activities that are difficult to handle by relying on static attribute association schemes, and improve the identification and discovery of risky gangs. Thirdly, the bucketing technology adopted in the exemplary embodiments of this disclosure greatly reduces the amount of queries required when constructing relationships and improves data processing efficiency.

[0065] The following are embodiments of the apparatus disclosed herein, which can be used to execute embodiments of the method disclosed herein. For details not disclosed in the apparatus embodiments of this disclosure, please refer to the embodiments of the method disclosed herein.

[0066] This disclosure also provides an example embodiment of a user data processing apparatus. Specifically, refer to... Figure 9 As shown, the user data processing device may include a user bucketing result determination module 910, an entity triplet construction module 920, a user association feature map construction module 930, and a user category determination module 940.

[0067] The user binning result determination module 910 can be used to clean the original user attribute data of the users to be classified to obtain target user attribute data, and to bin the users to be classified according to the target user attribute data to obtain user binning results; the entity triple construction module 920 can be used to determine the attribute association relationship between the user entities corresponding to the users to be classified included in the same user binning result, and to construct entity triples according to the attribute association relationship and user entities; the user association feature graph construction module 930 can be used to determine the collaborative association edge between user entities according to the entity triple, and to construct a user association feature graph based on the collaborative association edge, user entities and target user attribute data; the user category determination module 940 can be used to input the user association feature graph into a preset user classification model to obtain user classification results, and to determine the user category of the users to be classified according to the user classification results.

[0068] In one exemplary embodiment of this disclosure, the user to be classified is bucketed according to target user attribute data to obtain user bucketing results, including: determining a primary attribute dimension from the target attribute dimension corresponding to the target user attribute data, and obtaining the data recording time of the primary attribute data corresponding to the primary attribute dimension; sorting the users to be classified according to the data recording time to obtain user sorting results, and determining the time difference between the data recording times of two adjacent users to be classified in the user sorting results; determining the data storage bucket to which the user to be classified belongs according to the time difference to obtain the user bucketing results; wherein, each user to be classified corresponds to one data storage bucket.

[0069] In one exemplary embodiment of this disclosure, determining the data storage bucket to which the user to be classified belongs based on the time difference to obtain the user bucketing result includes: in response to the time difference being less than or equal to a preset time threshold, determining that the user to be classified belongs to the same data storage bucket, and writing the target user attribute data of the user to be classified into the same data storage bucket; in response to the time difference being greater than the preset time threshold, determining that the user to be classified belongs to different data storage buckets, and writing the target user attribute data of the user to be classified into different data storage buckets; and generating the user bucketing result based on the user to be classified written in the data storage bucket and the target user attribute data associated with the user to be classified.

[0070] In one exemplary embodiment of this disclosure, determining the attribute association relationships between user entities corresponding to the user to be classified, included in the same user bucketing result, includes: traversing the user entities corresponding to the user to be classified, included in the same user bucketing result, determining the user attribute values ​​of the target user attribute data corresponding to the user entity, and performing embedding mapping processing on the user attribute values ​​to obtain the attribute feature vector of the user entity corresponding to the user to be classified in the target attribute dimension; determining the attribute association relationships between user entities based on the attribute feature vector; wherein, the attribute association relationships include at least one of user name association relationships, user unique identifier association relationships, user IP address association relationships, user device fingerprint association relationships, user address information association relationships, and user-associated personnel information association relationships.

[0071] In an exemplary embodiment of this disclosure, constructing entity triples based on the attribute association and user entities includes: determining a group of user entities with attribute association; wherein the user entity group includes a first user entity and a second user entity; determining a target attribute dimension corresponding to the attribute association; wherein the target user dimension includes at least one of a user name dimension, a user unique identifier dimension, a user device fingerprint dimension, a user address information dimension, and a user associated personnel information dimension; constructing the entity triples based on the first user entity, the second user entity, and the target attribute dimension; wherein each entity triple corresponds to one target attribute dimension.

[0072] In an exemplary embodiment of this disclosure, determining the collaborative association edge between user entities based on the entity triplet includes: determining the number of entity groups of user entity groups between the first user entity and the second user entity, and determining whether the number of entity groups is greater than a preset number threshold; in response to the number of entity groups being greater than the preset number threshold, determining that there is a collaborative association edge between the first user entity and the second user entity.

[0073] In one exemplary embodiment of this disclosure, constructing a user association feature graph based on the collaborative association edges, user entities, and target user attribute data includes: using the user entity as a first vertex, using the collaborative association edges between user entities as first connecting edges, and determining the weight value of the collaborative association edges according to the number of entity groups of user entity groups between the user entities; using the target user attribute data as a second vertex, and using the correspondence between the user entity and the target user attribute data as a second connecting edge; and constructing the user association feature graph based on the first vertex, the first connecting edge and the weight value, the second vertex and the second connecting edge.

[0074] In one exemplary embodiment of this disclosure, inputting the user association feature map into a preset user classification model to obtain a user classification result includes: inputting the user association feature map into a preset community detection algorithm model to perform similarity fusion and community optimization on the user association feature map based on the preset community detection algorithm model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map; and / or, inputting the user association feature map into a preset graph convolutional neural network model to perform node feature updates on the user association feature map based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map.

[0075] In an exemplary embodiment of this disclosure, the preset graph convolutional neural network model includes a first graph convolutional layer, a second graph convolutional layer, ..., an Nth graph convolutional layer, and a classification layer. The process of updating node features of the user association feature map based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and determining the user classification result based on the optimized user association feature map, includes: updating node features of the user association feature map based on the first graph convolutional layer to obtain a first graph convolutional processing result; updating node features of the first graph convolutional processing result based on the second graph convolutional layer to obtain a second graph convolutional processing result; repeating the determination process of the second graph convolutional processing result sequentially to obtain a third graph convolutional processing result, a fourth graph convolutional processing result, ..., an Nth graph convolutional processing result; using the Nth graph convolutional processing result as the optimized user association feature map, and performing classification processing on the optimized user association feature map based on the classification layer to obtain the user classification result.

[0076] The specific details of each module in the aforementioned user data processing device have been described in detail in the corresponding user data processing methods, so they will not be repeated here.

[0077] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0078] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0079] In an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided.

[0080] Those skilled in the art will understand that various aspects of this disclosure can be implemented as a system, method, or program product. Therefore, various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, collectively referred to herein as a "circuit," "module," or "system."

[0081] The following reference Figure 10 To describe an electronic device 1000 according to such an embodiment of the present disclosure. Figure 10 The electronic device 1000 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0082] like Figure 10 As shown, the electronic device 1000 is manifested in the form of a general-purpose computing device. The components of the electronic device 1000 may include, but are not limited to: at least one processing unit 1010, at least one storage unit 1020, a bus 1030 connecting different system components (including storage unit 1020 and processing unit 1010), and a display unit 1040.

[0083] The storage unit stores program code that can be executed by the processing unit 1010, causing the processing unit 1010 to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure. For example, the processing unit 1010 can perform actions such as... Figure 1 Step S110: Clean the original user attribute data of the users to be classified to obtain target user attribute data, and bin the users to be classified according to the target user attribute data to obtain user binning results; Step S120: Determine the attribute association relationship between the user entities corresponding to the users to be classified included in the same user binning result, and construct entity triples according to the attribute association relationship and user entities; Step S130: Determine the collaborative association edges between user entities according to the entity triples, and construct a user association feature map based on the collaborative association edges, user entities, and target user attribute data; Step S140: Input the user association feature map into a preset user classification model to obtain user classification results, and determine the user category of the users to be classified according to the user classification results.

[0084] Storage unit 1020 may include readable media in the form of volatile storage units, such as random access memory (RAM) 10201 and / or cache memory 10202, and may further include read-only memory (ROM) 10203.

[0085] Storage unit 1020 may also include a program / utility 10204 having a set (at least one) program module 10205, such program module 10205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0086] Bus 1030 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0087] Electronic device 1000 can also communicate with one or more external devices 1100 (e.g., keyboard, pointing device, Bluetooth device, etc.), one or more devices that enable a user to interact with electronic device 1000, and / or any device that enables electronic device 1000 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 1050. Furthermore, electronic device 1000 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 1060. As shown, network adapter 1060 communicates with other modules of electronic device 1000 via bus 1030. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0088] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0089] In exemplary embodiments of this disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible implementations, various aspects of this disclosure may also be implemented as a program product including program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of this disclosure described in the "Exemplary Methods" section above.

[0090] The program product for implementing the above-described method according to embodiments of the present disclosure may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0091] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0092] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0093] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0094] Program code for performing the operations of this disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0095] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this disclosure and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0096] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention described herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not invented by this disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

Claims

1. A method for processing user data, characterized in that, include: Data cleaning is performed on the original user attribute data of the users to be classified to obtain target user attribute data, and the users to be classified are binned according to the target user attribute data to obtain user binning results; Determine the attribute associations between the user entities corresponding to the user to be classified in the same user bucketing results, and construct entity triples based on the attribute associations and user entities; Based on the entity triples, collaborative association edges between user entities are determined, and a user association feature graph is constructed based on the collaborative association edges, user entities, and target user attribute data. The user association feature map is input into a preset user classification model to obtain the user classification result, and the user category of the user to be classified is determined based on the user classification result.

2. The user data processing method according to claim 1, characterized in that, The users to be classified are bucketed based on the target user attribute data to obtain the user bucketing results, including: Determine the primary attribute dimension from the target attribute dimensions corresponding to the target user attribute data, and obtain the data recording time of the primary attribute data corresponding to the primary attribute dimension; The users to be classified are sorted according to the data recording time to obtain the user sorting result, and the time difference between the data recording times of two adjacent users to be classified in the user sorting result is determined. The data storage bucket to which the user to be classified belongs is determined based on the time difference, so as to obtain the user bucketing result; wherein, each user to be classified corresponds to one data storage bucket.

3. The user data processing method according to claim 2, characterized in that, The data storage bucket to which the user to be classified belongs is determined based on the time difference, in order to obtain the user bucketing result, including: In response to the time difference being less than or equal to a preset time threshold, it is determined that the users to be classified belong to the same data storage bucket, and the target user attribute data of the users to be classified is written into the same data storage bucket. In response to the time difference being greater than a preset time threshold, it is determined that the user to be classified belongs to different data storage passes, and the target user attribute data of the user to be classified is written into different data storage buckets; The user bucketing result is generated based on the user to be classified and the target user attribute data associated with the user to be classified written in the data storage bucket.

4. The user data processing method according to claim 1, characterized in that, Determine the attribute relationships between the user entities corresponding to the user to be classified, included in the same user bucketing results, including: Traverse the user entities corresponding to the user to be classified included in the same user bucketing result, determine the user attribute values ​​of the target user attribute data corresponding to the user entity, and perform embedding mapping processing on the user attribute values ​​to obtain the attribute feature vector of the user entity corresponding to the user to be classified in the target attribute dimension. The attribute associations between user entities are determined based on the attribute feature vectors; wherein, the attribute associations include at least one of the following: user name association, user unique identifier association, user IP address association, user device fingerprint association, user address information association, and user-associated personnel information association.

5. The user data processing method according to claim 1, characterized in that, Based on the attribute relationships and user entities, entity triples are constructed, including: Identify a group of user entities with attribute relationships; wherein the group of user entities includes a first user entity and a second user entity; Determine the target attribute dimension corresponding to the attribute association relationship; wherein, the target user dimension includes at least one of the following: user name dimension, user unique identifier dimension, user device fingerprint dimension, user address information dimension, and user associated personnel information dimension; The entity triples are constructed based on the first user entity, the second user entity, and the target attribute dimension; wherein each entity triple corresponds to a target attribute dimension.

6. The user data processing method according to claim 5, characterized in that, Determining collaborative association edges between user entities based on the entity triples includes: Determine the number of entity groups in the user entity group between the first user entity and the second user entity, and determine whether the number of entity groups is greater than a preset number threshold. In response to the number of entity groups exceeding a preset threshold, it is determined that there is a collaborative association edge between the first user entity and the second user entity.

7. The user data processing method according to claim 1, characterized in that, Based on the collaborative association edges, user entities, and target user attribute data, a user association feature graph is constructed, including: The user entity is taken as the first vertex, the collaborative association edge between user entities is taken as the first connecting edge, and the weight value of the collaborative association edge is determined according to the number of entity groups of user entity groups between the user entities. The target user attribute data is used as the second vertex, and the correspondence between the user entity and the target user attribute data is used as the second connecting edge; The user association feature map is constructed based on the first vertex, the first connecting edge and its weight value, the second vertex and the second connecting edge.

8. The user data processing method according to claim 1, characterized in that, The user association feature map is input into a preset user classification model to obtain user classification results, including: The user association feature map is input into a preset community detection algorithm model. Similarity fusion and community optimization are performed on the user association feature map based on the preset community detection algorithm model to obtain an optimized user association feature map. The user classification result is then determined based on the optimized user association feature map; and / or The user association feature map is input into a preset graph convolutional neural network model to update the node features of the user association feature map based on the preset graph convolutional neural network model, thereby obtaining an optimized user association feature map, and the user classification result is determined based on the optimized user association feature map.

9. The user data processing method according to claim 8, characterized in that, The preset graph convolutional neural network model includes a first graph convolutional layer, a second graph convolutional layer, ..., an Nth graph convolutional layer, and a classification layer; Specifically, the user association feature map is updated with node features based on the preset graph convolutional neural network model to obtain an optimized user association feature map, and the user classification result is determined based on the optimized user association feature map, including: Based on the first graph convolutional layer, the node features of the user-related feature graph are updated to obtain the first graph convolutional processing result. Based on the second graph convolutional layer, the node features of the first graph convolutional processing result are updated to obtain the second graph convolutional processing result. The process of determining the convolution result of the second image is repeated sequentially to obtain the convolution result of the third image, the convolution result of the fourth image, ..., the convolution result of the Nth image; The convolution result of the Nth graph is used as the optimized user association feature map, and the optimized user association feature map is classified based on the classification layer to obtain the user classification result.

10. A user data processing system, characterized in that, include: The system comprises a data acquisition system, a data preprocessing system, and a Spark computing engine, wherein the data preprocessing system is communicatively connected to both the data acquisition system and the Spark computing engine; wherein: The data acquisition system is used to collect the original user attribute data of users to be classified, and send the original user attribute data to the data preprocessing system; The data preprocessing system is used to clean the original user attribute data to obtain target user attribute data, and to classify the users to be classified into buckets based on the target user attribute data to obtain user bucketing results. The Spark computing engine is used to determine the attribute relationships between user entities corresponding to the user to be classified within the same user bucketing results, and to construct entity triples based on the attribute relationships and user entities; and Based on the entity triples, collaborative association edges between user entities are determined, and a user association feature graph is constructed based on the collaborative association edges, user entities, and target user attribute data; and The user association feature map is input into a preset user classification model to obtain the user classification result, and the user category of the user to be classified is determined based on the user classification result.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the user data processing method according to any one of claims 1-9.

12. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the user data processing method according to any one of claims 1-9 by executing the executable instructions.

Images