CN111260372B

CN111260372B - Resource transfer user group determination method, device, computer equipment and storage medium

Info

Publication number: CN111260372B
Application number: CN202010070871.6A
Authority: CN
Inventors: 林海雄
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2020-01-21
Filing date: 2020-01-21
Publication date: 2023-04-21
Anticipated expiration: 2040-01-21
Also published as: CN111260372A

Abstract

The application relates to a resource transfer user group determining method, a device, computer equipment and a storage medium, wherein the method comprises the following steps: acquiring target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources; obtaining the target resource transfer relation degree between the target users according to the target resource transfer data between the target users; taking target users in the target user set as target nodes, and obtaining the connection degree corresponding to edges between the nodes according to the target resource transfer relation degree between the target users to obtain a target resource transfer relation diagram; and carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group. The method can improve the dividing efficiency of the target resource transfer user group.

Description

Resource transfer user group determination method, device, computer equipment and storage medium

Technical Field

The present invention relates to the field of data processing, and in particular, to a method, an apparatus, a computer device, and a storage medium for determining a resource transfer user group.

Background

With the rapid development of science and technology, more and more users are carrying out resource transfer, and great convenience is brought to the life of people. For example, people can conduct various transactions through paymate in the internet without having to carry cash with them. However, there are some illegal molecules that use this resource transfer convenience to make fraudulent or like illegal events.

The related authorities need to determine the user group to which the lawbreaker belongs, so as to precisely hit the group, thereby reducing the possibility that the lawbreaker uses the user account to cheat resources. Currently, the population to which lawless persons belong is mainly identified manually, so that the efficiency of determining the user population is low.

Disclosure of Invention

Based on this, it is necessary to provide a resource transfer user group determining method, apparatus, computer device and storage medium, in order to solve the above problem of low efficiency in determining user groups.

A method of resource transfer user group determination, the method comprising: acquiring target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources; obtaining the target resource transfer relation degree between the target users according to the target resource transfer data between the target users; taking target users in the target user set as target nodes, and obtaining the connection degree corresponding to edges between the nodes according to the target resource transfer relation degree between the target users to obtain a target resource transfer relation diagram; and carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group.

A resource transfer user group determination apparatus, the apparatus comprising: the target resource transfer data acquisition module is used for acquiring target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources; the target resource transfer relation obtaining module is used for obtaining the target resource transfer relation between the target users according to the target resource transfer data between the target users; the target resource transfer relation diagram obtaining module is used for taking target users in the target user set as target nodes, obtaining the connection degree corresponding to edges between the nodes according to the target resource transfer relation degree between the target users, and obtaining a target resource transfer relation diagram; and the target resource transfer user group dividing module is used for carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group.

In some embodiments, the target user set deriving unit is configured to: acquiring resource transfer characteristics corresponding to each resource transfer user in the resource transfer user sequence; and screening the resource transfer user sequence according to the resource transfer characteristics corresponding to the resource transfer users to obtain target users to form a target user set.

In some embodiments, the target user set deriving unit is configured to: for the resource transfer users in the resource transfer user sequence, acquiring resource association users corresponding to the resource transfer users; and obtaining a target user set according to the resource associated user and the resource transfer user sequence.

In some embodiments, the adjustment unit is configured to: the first current relation weight and the second current relation weight are adjusted towards the direction of increasing the user similarity corresponding to the current training resource transfer user group, and updated first current relation weight and second current relation weight are obtained; the apparatus further comprises: and the return module is used for returning to the step of obtaining the third current relation degree according to the third resource transfer data and the corresponding first current relation weight and obtaining the fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight until the similarity of the corresponding users of the current training resource transfer user group is larger than the preset similarity.

In some embodiments, the target resource transfer user group partitioning module is configured to: taking the target resource transfer relation diagram as a current resource transfer relation diagram; obtaining a current node to be divided in a current resource transfer relation diagram, and calculating a corresponding module degree increment when the current node and a corresponding adjacent node form a current node group; when the modularity increment meets a preset increment condition, merging the nodes in the current node group into new nodes to obtain an updated current resource transfer relation diagram; and returning to obtain the current node to be divided in the current resource transfer relation diagram based on the updated current resource transfer relation diagram, and calculating the corresponding module degree increment when the current node and the corresponding adjacent node form the current node group until the update of the current resource transfer relation diagram is terminated, thereby obtaining the termination resource transfer relation diagram. And dividing the users corresponding to the same node in the termination resource transfer relation diagram into the same target resource transfer user group.

In some embodiments, the target resource transfer user group partitioning module is configured to: calculating the current modularity corresponding to the current node group; calculating according to the node connectivity corresponding to the current node and the node connectivity corresponding to the adjacent node to obtain a reference modularity; and obtaining corresponding module degree increment when the current node and the corresponding adjacent node form a current node group according to the difference between the current module degree and the reference module degree.

A computer device comprising a memory and a processor, the memory having stored therein a computer program which, when executed by the processor, causes the processor to perform the steps of the resource transfer user group determination method described above.

A computer readable storage medium having stored thereon a computer program which, when executed by a processor, causes the processor to perform the steps of the resource transfer user group determination method described above.

According to the resource transfer user group determining method, the device, the computer equipment and the storage medium, the target users in the target user set are the resource transfer related users corresponding to the target risk resources, the target resource transfer relation degree among the target users can be obtained through the target resource transfer data among the target users, group division is carried out according to the target resource transfer relation diagram, wherein the target resource transfer relation diagram takes the target users in the target user set as nodes, and the target resource transfer relation degree among the target users is taken as the connection degree corresponding to the edges among the nodes. Because the target user is a resource transfer related user corresponding to the target risk resource, and the transfer relation degree determined by the resource transfer data can accurately reflect the transfer relation among the users, the target user set can be accurately subjected to group division according to the target resource transfer relation diagram, and the division accuracy and efficiency of the user group for transferring the risk resource are improved.

Drawings

FIG. 1 is an application environment diagram of a resource transfer user group determination method provided in one embodiment;

FIG. 2 is a flow diagram of a method of resource transfer user group determination, in one embodiment;

FIG. 3 is a schematic illustration of a longitudinal expansion performed in one embodiment;

FIG. 4 is a schematic illustration of a longitudinal expansion performed in one embodiment;

FIG. 5 is a flow chart of a method for obtaining a target user according to one embodiment;

FIG. 6 is a diagram of a target resource transfer relationship graph, in one embodiment;

FIG. 7 is a flow chart of performing group partitioning on a set of target users according to a target resource transfer relationship graph to obtain at least one target resource transfer user group in one embodiment;

FIG. 8 is a diagram of a current resource transfer relationship diagram in one embodiment;

FIG. 9 is a diagram of a terminating resource transfer relationship diagram, in one embodiment;

FIG. 10 is a flow chart of computing a first target relationship weight and a second target relationship weight in one embodiment;

FIG. 11 is a block diagram of an apparatus for determining a resource transfer user group in one embodiment;

FIG. 12 is a block diagram of the internal architecture of a computer device in one embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

It will be understood that the terms "first," "second," and the like, as used herein, may be used to describe various elements, but these elements are not limited by these terms unless otherwise specified. These terms are only used to distinguish one element from another element. For example, a first degree of relationship may be referred to as a second degree of relationship, and similarly, a second degree of relationship determination script may be referred to as a first degree of relationship, without departing from the scope of the present application.

Fig. 1 is an application environment diagram of a resource transfer user group determining method provided in one embodiment, as shown in fig. 1, in the application environment, including a terminal 110 and a server 120. The server 120 may acquire a plurality of user accounts for fraud, for example, bank card accounts, to form a target user set, so as to execute the method provided by the embodiment of the present application, obtain at least one target resource transfer user group corresponding to the target user set, and send the division result to the terminal 110, for example, the terminal 110 may display a bank card account corresponding to each target resource transfer user group. The server 120 may also intercept the resource transfer to the user account corresponding to the target resource transfer user group, so as to improve the security of the resource transfer. For example, when the server 120 detects that a user transfers a bank card account corresponding to a user group to a target resource, the corresponding target resource is intercepted, i.e. the resource is not transferred to the bank card account, so that the resource loss of the user is reduced.

The method provided by the implementation of the application can be used for detecting the water house partner, namely, the money washing partner which distributes the black money to a plurality of transaction accounts, such as a plurality of small bank cards, through network transfer and then takes the money out by the partner member. Money laundering is a legal act of illegally obtaining, mainly to mask or hide the source and properties of the income obtained by the illegal and generated by the illegal, and make the illegal and generated income legal in form. The illegal result may be obtained by fraud by an impersonation method, or by theft by a hacking means, for example. By acquiring the bank card account number for fraud, the fraud partner to which the bank card account number belongs is determined and provided to the public security authority, so that the public security authority can precisely strike the fraud partner.

The server 120 may be an independent physical server, or may be a server cluster formed by a plurality of physical servers, or may be a cloud server that provides basic cloud computing services such as a cloud server, a cloud database, cloud storage, and CDN. The terminal 110 may be, but is not limited to, a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, etc. The terminal 110 and the server 120 may be connected through a communication connection manner such as bluetooth, USB (Universal Serial Bus ) or a network, which is not limited herein.

As shown in fig. 2, in one embodiment, a method for determining a resource transfer user group is provided, and this embodiment is mainly exemplified by the method being applied to the server 120 in fig. 1. The method specifically comprises the following steps:

step S202, obtaining target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources.

Specifically, the resource refers to a resource that exists in an electronic account and can be circulated, for example, through an account on the internet. Accounts may include, but are not limited to: a bank card number or an internet payment account, etc. Resources include, but are not limited to, currency, virtual red packs, game pieces or virtual items, and the like. Resource transfer refers to transferring resources from one user account to another. The target user set comprises a plurality of target users, the specific number of the target users can be determined according to actual conditions, the users can be represented by accounts, and one account represents one user, such as a bank card number. Risk resources refer to resources that are at risk, such as abnormally transferred resources. Abnormal transfer means that the resource is transferred according to an unfair act, for example, the resource is stolen or is fraudulently stolen. As a practical example, when a user encounters fraud, a bank card number may be transferred to the fraud user, and the money transferred is an abnormally transferred resource. There may be multiple target risk resources. For example, multiple abnormally transferred resources may be acquired as target risk resources. The resource transfer-related users corresponding to the target risk resource may include at least one of the target resource transfer users transferred through the target risk resource or users related to the target resource transfer users. The users associated with the target resource transfer users may include users that have been resource transferred between the target resource transfer users. For example, assuming that user A is fraudulently tricked by user B, a bank card number is transferred to user B, and user B transfers the fraudulently tricked money to user C again, user B and user C are resource transfer related users. If the D user and the C user carry out resource transfer once, the D user can also be a resource transfer related user.

The target resource transfer data refers to data related to resource transfer, and may include at least one of a resource transfer amount or a number of resource transfers. The resource transfer amount refers to the amount of resources transferred between users. For example, if user a transfers to user B twice, the first transfer is 100 yuan and the second transfer is 300 yuan, the amount of resource transfer is 400 yuan and the number of resource transfers is 2.

In some embodiments, obtaining the set of target users includes: acquiring a user receiving a target risk resource, wherein the target risk resource is an abnormally transferred resource; obtaining a plurality of resource transfer user sequences taking a user receiving the target risk resource as an initial user according to the resource transfer flow direction of the target risk resource; and obtaining a target user set according to the resource transfer user sequence.

In particular, a user receiving a target risk resource may refer to the first user receiving the target risk resource. For example, assuming that user a is fraudulently tricked by user B, transferring a bank card number to user B, user B is the user who receives the target risk resource. The user receiving the target risk resource may be predetermined. For example, after the user a is fraudulently deceived, the user a may perform a resource transfer complaint or alarm through the internet platform, and the server may automatically use the resource receiving account corresponding to the resource transfer complaint as the user for receiving the target risk resource, or may obtain the user for receiving the target risk resource from other channels, for example, obtain the bank card account for receiving the target risk resource from other servers or terminals, as the user account for receiving the target risk resource.

The transfer flow refers to the direction of the resource transfer, the initial user refers to the first user, and the sequence of resource transfer users can be one or more. After receiving the target risk resource, the user receiving the target risk resource further transfers the target risk resource, so that longitudinal expansion is performed, a circulation path of transferring the target risk resource is obtained, and the user receiving the target risk resource is formed into a resource transfer user sequence according to the sequence of receiving the target risk resource. After the resource transfer user sequence is obtained, the users in the resource transfer user sequence can be used as target users, or part of users in the resource transfer user sequence can be selected as target users. The target risk resources are resources for determining existing risks, so that the target user set is obtained by obtaining the users for receiving the target risk resources according to the resource transfer flow direction of the target risk resources, the users for receiving the risk resources can be obtained, then the target user set is determined according to the users, namely, the users related to risk resource transfer can be accurately obtained, malicious users are obtained according to the users, and the accuracy of the users in the target user set as the malicious users is high.

As a practical example, as shown in fig. 3, when a user (victim) is deceived, a bank card is transferred to a fraudster. The bank card of the victim is called a case card, and the card directly transferred by the case card is called a case card, when complaint data of the victim is received, the server obtains the case card (namely, a malicious card of a bad person) and can be longitudinally expanded after obtaining the case card. The transaction flow of the case card can be obtained, and the money in the fraud transaction is tracked. Tracking which cards the victim transfers to the case cards, these cards being referred to as level 1 cards, continuing to acquire level 1 cards transfers the money to which cards, resulting in level 2 cards, and eventually the money may be scattered to many cards until the fraud takes the money from these cards using an ATM (Automatic Teller Machine, automated teller machine), the card from which the money was taken being referred to as the cash card. Multiple resource transfer user sequences are thus available. For example, one of the resource transfer user sequences is case card→1 level card 1→2 level card 1→cash card 1. Through longitudinal expansion, the bank card through which the fraudulent money flows can be obtained. It will be appreciated that the card referred to herein may also be replaced with an electronic payment account, such as a payment account on an internet paymate.

In some embodiments, deriving the set of target users from the sequence of resource transfer users comprises: acquiring resource transfer characteristics corresponding to each resource transfer user in a resource transfer user sequence; and screening the target users from the resource transfer user sequences according to the resource transfer characteristics corresponding to the resource transfer users to form a target user set.

Specifically, the resource transfer characteristic refers to a characteristic related to resource transfer, such as at least one of the frequency of resource transfer or the average number of resource transfers in which resource transfer is performed, and the like. Other user characteristics may of course be included, such as the geographic location of the user, the age or hobbies of the user, etc. A model for determining malicious users, such as an artificial intelligence model for detecting fraud users, can be trained, the resource transfer characteristics of the users are input into the model, whether the users are malicious users or not is detected, if so, the users are taken as target users and added into a target user set, and therefore the target user set formed by the malicious users can be obtained. By further determining the malicious user in combination with the resource transfer characteristics of the user, the accuracy of the malicious user can be determined.

In some embodiments, deriving the set of target users from the sequence of resource transfer users comprises: for a resource transfer user in a resource transfer user sequence, acquiring a resource association user corresponding to the resource transfer user; and obtaining a target user set according to the resource associated user and the resource transfer user.

Specifically, the resource associated user corresponding to the resource transfer user refers to a user having a resource transfer relationship with the resource transfer user, that is, a resource transfer transaction exists between the resource transfer user and the corresponding resource associated user. For example, if user B is a resource transfer user, user C has ever transferred to user B, i.e., user B's resource sources include user C. User C associates user C with user D as user B's resource once user B has transferred to user D, i.e., user B's resource has been directed to include user D. After obtaining the resource-associated user, the resource-associated user may be used as a target user, or the resource-associated user may be screened to obtain the target user. For example, the resource transfer characteristics of the resource-associated user may also be obtained, and whether the resource-associated user is a malicious user may be determined according to the resource transfer characteristics of the resource-associated user. For the resource transfer users in the resource transfer user sequence, the users in the resource transfer user sequence can be all used as target users, or part of the users in the resource transfer user sequence can be selected as target users. In the embodiment of the application, the resource associated user with the resource transfer relationship with the resource transfer user is further obtained, and the target user set is obtained according to the resource associated user, so that more malicious users can be further expanded and obtained.

As a practical example, obtaining a resource-associated user corresponding to a resource transfer user may also be referred to as lateral expansion of the user. Because the resource transfer user receives the abnormally transferred resources, the user can be used as a malicious user, and the bank card of the malicious user becomes a malicious card. The transaction condition with the malicious card can be obtained, the card which is in transaction with the malicious card is taken as the suspected card, and then more malicious cards are screened out from the suspected card according to the characteristics of the malicious card and are taken as target cards. Fig. 4 is a schematic diagram of lateral expansion according to a malicious card. The direction of the arrow represents the flow direction of the resource. And the associated card transferred to the malicious card and the associated card transferred by the malicious card can be obtained as suspected cards. And determining whether the suspected card is a malicious card according to the resource transfer characteristics of the suspected card.

In some embodiments, lateral expansion may also continue for newly discovered malicious cards until no new malicious cards are present. For example, as shown in fig. 5, after obtaining the case cards, a longitudinal expansion may be performed, the cards that receive the resources of the users that are fraudulently used as malicious cards, a lateral expansion is performed based on the malicious cards, so as to obtain the cards that have transactions with the malicious cards as suspected cards, the malicious cards are selected from the suspected cards, and a lateral expansion is performed based on the newly selected malicious cards, so as to obtain new suspected cards until no new malicious cards are generated.

In some embodiments, all users that have a resource transfer relationship with the resource transfer user may be obtained as resource-associated users. In order to improve the efficiency of determining the target user, the user having the resource transfer relationship with the resource transfer user in the first time period may be obtained as the resource association user. The first duration may be a duration that the resource transfer frequency of the resource transfer user is greater than a preset frequency, for example, if the billing rate of the user B is greater than 10 times per day in the last month, the first duration is 1 month, and the user having a transfer relationship with the user B in the last month is acquired as the resource association user.

Step S204, obtaining the target resource transfer relation degree between the target users according to the target resource transfer data between the target users.

Specifically, the resource transfer relationship degree is used to represent the tightness of the resource transfer relationship between users, and the higher the relationship degree is, the higher the closeness of the resource transfer is. The resource transfer relatedness may be derived from at least one of the number of resource transfers or the number of resource transfers. When the target resource transfer relation is obtained according to the number of resource transfers and the number of resource transfers, the resource transfer relation corresponding to the number of resource transfers and the number of resource transfers can be added to obtain the target resource transfer relation.

The number of resource transfers is in positive correlation with the resource transfer relationship, i.e. the larger the number of resource transfers, the larger the resource transfer relationship. The number of resource transfers is in positive correlation with the degree of resource transfer relationship. At least one of the correspondence between the number of resource transfers and the degree of resource transfer relationship or the correspondence between the number of resource transfers and the degree of resource transfer relationship may be preset. To obtain a resource transfer relationship according to at least one of the number of resource transfers or the number of resource transfers. For example, the correspondence of the number of times range and the resource transfer relationship degree, and the correspondence of the number range and the resource transfer relationship degree may be set. When the number of times is 1-3, the corresponding resource transfer relation degree is 2. When the number of times is 4-6, the corresponding resource transfer relation degree is 3. When the number is 1 to 1000 yuan, the corresponding resource transfer relation degree is 1.

In some embodiments, obtaining the target resource transfer relationship between the target users according to the resource transfer data between the target users includes: acquiring first resource transfer data between an ith target user and a jth target user and second resource transfer data between the ith target user and the jth target user, wherein i is less than or equal to n, j is less than or equal to n, i and j are positive integers, n is the number of target users in a target user set, the first resource transfer data is transfer data corresponding to abnormal transfer resources, and the second resource transfer data is transfer data corresponding to non-abnormal transfer resources; obtaining a first relation degree according to the first resource transfer data and the corresponding first target relation weight, and obtaining a second relation degree according to the second resource transfer data and the corresponding second target relation weight; and obtaining the target resource transfer relation degree between the ith target user and the jth target user according to the first relation degree and the second relation degree.

Specifically, the resource transfer data may include transfer data corresponding to a resource (abnormal transfer resource) for which an abnormal transfer exists and transfer data corresponding to a resource for which an abnormal transfer is not determined to exist. . Non-exception-transferring resources refer to resources for which the resource has not been determined to be an exception transfer. For example, if it is determined that the first transaction is 100 yuan for user B to transfer to user C, and it has been determined that the transfer resource, i.e., 100 yuan, is fraud by user B, then the resource corresponding to the transaction is the resource for the abnormal transfer. When the second transaction is a 300-ary transfer from user B to user C, which 300-ary has not been determined whether user B is legitimate or fraudulent, then the 300-ary is the resource for which no abnormal transfer is determined to exist.

i and j may be any positive integer, i.e., the ith target user and the jth target user may be any user in the set of target users, i and j being different. For example, a degree of resource transfer relationship between two users in a set of target users may be calculated. For two users that do not have a resource transfer relationship, the corresponding resource transfer relationship degree may be a preset value, for example, 0.

The first target relation weight and the second target relation weight can be preset, can be empirically set, and can also be calculated through a corresponding weight algorithm. The first target relationship weight may be greater than the second target relationship weight. After the resource transfer data and the corresponding target relation weight are obtained, the resource transfer data and the corresponding weight can be multiplied to obtain the relation degree. For example, the first resource transfer data may be multiplied by a corresponding first target relationship weight to obtain a first degree of relationship. And multiplying the second resource transfer data by a corresponding second target relationship weight to obtain a second relationship degree. That is, the first degree of relationship is obtained from the resource transfer data corresponding to the resource for which the abnormal transfer is determined, and the second degree of relationship may be obtained from the resource transfer data corresponding to the resource for which the abnormal transfer is not determined. The first relationship may be added to the second relationship to obtain a target resource transfer relationship between the ith target user and the jth target user. By distinguishing the relationship weight corresponding to the abnormal resource transfer data and the relationship weight corresponding to the non-abnormal transfer resource, the obtained target resource transfer relationship degree is more accurate.

In some embodiments, resource transfer data between users obtained through longitudinal expansion can be used as transfer data corresponding to resources with abnormal transfer. And taking the resource transfer data among the users obtained through transverse expansion as transfer data corresponding to the non-abnormal transfer resources.

In some embodiments, the target resource transfer relationship between user i and user j may be calculated according to equation (1). Wherein a1 represents a first target relationship weight corresponding to the number of resource transfers corresponding to the abnormal transfer resource, and a represents a first target relationship weight corresponding to the number of resource transfers corresponding to the abnormal transfer resource _， a represents a second target relation weight corresponding to the number of resource transfers corresponding to the non-abnormal transfer resource, and a4 represents a second target relation weight corresponding to the number of resource transfers corresponding to the non-abnormal transfer resource. X1 represents the number of resource transfers corresponding to the abnormal transfer resource, X2 represents the number of resource transfers corresponding to the abnormal transfer resource, X3 represents the number of resource transfers corresponding to the non-abnormal transfer resource, X4 represents the number of resource transfers corresponding to the non-abnormal transfer resource, and Wij represents the target resource transfer relationship between user i and user j. I.e. the first two terms to the right of the equation represent the first degree of relationship and the second two terms to the right of the equation represent the second degree of relationship. Where x1, x2, x3, and x4 may be normalized resource transfer data, and the normalization method may be, for example, a standard score (z-score) method.

Wij＝a1*x1+a2*x2+a3*x3+a4*x4 (1)

Step S206, taking the target users in the target user set as target nodes, and obtaining the connection degree corresponding to the edges between the nodes according to the target resource transfer relation degree between the target users, so as to obtain a target resource transfer relation diagram.

Specifically, the resource transfer relationship graph is used for representing the resource transfer relationship among users, and the resource transfer relationship graph comprises nodes and edges. The nodes represent users, edges can exist or not exist among the nodes, and the existing edges represent resource transfer relations among the users. The connectivity of the edges represents the closeness of connection between the nodes, the greater the connectivity, the greater the closeness of connection. The target resource transfer relation degree between the target users can be used as the connection degree corresponding to the edges between the nodes. The connection degree corresponding to the edge can be obtained by combining other connection degrees. For example, the resource transfer relation degree is used as a first connection degree, the connection degree obtained according to the similarity of the geographic positions is used as a second connection degree, and the first connection degree and the second connection degree are added to obtain the connection degree corresponding to the edges between the nodes.

As shown in fig. 6, which is a schematic diagram of a target resource transfer relationship diagram in some embodiments, in fig. 6, a node represents a user, a connection line between nodes represents an edge, and a presence edge represents that a resource transfer relationship degree between nodes is not 0, that is, a resource transfer relationship exists, as can be seen in fig. 6, a target user set includes 13 users.

Step S208, performing group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group.

Specifically, the users in the target resource transfer user group belong to the same community, and the target resource transfer user group may be malicious group partners, for example, the same fraud group partners. Malicious parties are closely related people and composition sets, and the purpose of illegally acquiring resources is achieved through mutual collaboration. The number of users in the target resource transfer user group may be one or more, and the number refers to at least two. The number of the target resource transfer user groups is determined according to the actual situation, and the node group to which each node belongs can be determined according to the target resource transfer relationship diagram, wherein the users corresponding to each node in the node group belong to the same resource transfer user group.

The community division method can be used for carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group, and the community division algorithm can be, for example, fast-unfolding, LPA (Label Propagation Algorithm ), GN algorithm or Newman algorithm. fast-unfolding is a greedy algorithm based on modularity partitioning communities, by which a graph can be partitioned into different communities. The GN algorithm is a classical community discovery algorithm, proposed by Michelle Girvan and Mark Newman, whose basic idea is to continually delete the edges in the network that have the largest edge betweenness with respect to all source nodes, then recalculate the edge betweenness of the remaining edges in the network with respect to all source nodes, and repeat this process until all edges in the network are deleted.

For example, referring back to fig. 6, assuming that nodes 1 to 5 belong to the same node group, nodes 6 to 9 belong to the same node group, and nodes 10 to 13 belong to the same node group after division, users 1 to 5 constitute a target resource transfer user group 1, users 6 to 9 constitute a target resource transfer user group 2, and users 10 to 13 constitute a target resource transfer user group 3.

According to the resource transfer user group determination method, as the target users in the target user set are the resource transfer related users corresponding to the target risk resources, the target resource transfer relation degree between the target users can be obtained through the target resource transfer data between the target users, and group division is carried out according to the target resource transfer relation diagram, wherein the target resource transfer relation diagram takes the target users in the target user set as nodes and takes the target resource transfer relation degree between the target users as the connection degree corresponding to the edges between the nodes. Because the target user is a resource transfer related user corresponding to the target risk resource, and the transfer relation degree determined by the resource transfer data can accurately reflect the transfer relation among the users, the target user set can be accurately subjected to group division according to the target resource transfer relation diagram, and the division accuracy and efficiency of the user group for transferring the risk resource are improved.

In some embodiments, as shown in fig. 7, performing group division on the target user set according to the target resource transfer relationship graph, to obtain at least one target resource transfer user group includes:

step S702, the target resource transfer relationship diagram is used as the current resource transfer relationship diagram.

The step, the current resource transfer relationship graph refers to the current resource transfer relationship graph, and when the division is performed, the resource transfer relationship graph is continuously changed, for example, when the division is performed for the first time, the node 1 and the node 2 can be combined to obtain an updated current resource transfer relationship graph. In the second division, the node obtained by combining the node 1 and the node 2 can be combined with the node 3 to obtain an updated current resource transfer relation diagram.

Step S704, obtaining a current node to be divided in the current resource transfer relation diagram, and calculating a corresponding module degree increment when the current node and a corresponding adjacent node form a current node group.

Specifically, all nodes of the current resource transfer relationship graph can be traversed, and each node in the current resource transfer relationship graph is used as a current node. The adjacent node corresponding to the current node refers to a node having an edge with the current node. The modularity is used to measure the quality of group division, and when the points with dense connection are divided into one group, the value of the modularity is increased, so that the division mode with the maximum modularity can be used as the optimal group division mode. The modularity increment refers to the increment of the modularity of the current node group after the current node and the corresponding adjacent nodes form the current node group relative to the modularity of the current node group. I.e. the benefit of modularity brought by adding the current node to the node group where its neighbor nodes are located.

In some embodiments, calculating the corresponding module degree increment when the current node and the corresponding adjacent node form the current node group includes: calculating the current modularity corresponding to the current node group; calculating according to the node connectivity corresponding to the current node and the connectivity corresponding to the adjacent node to obtain a reference modularity; and obtaining corresponding module degree increment when the current node and the corresponding adjacent node form the current node group according to the difference between the current module degree and the reference module degree.

Specifically, the connectivity of edges between nodes in the current node group may be added to obtain a current modularity corresponding to the current node group, where the modularity reflects the connectivity strength of the current node group. The connectivity corresponding to the current node is calculated according to the connectivity of all the edges connected with the node, for example, the connectivity of all the edges connected with the current node can be added to obtain the connectivity corresponding to the current node. The connection degree corresponding to the adjacent node is calculated from the connection degrees of all the edges connected to the node, and for example, the connection degrees of all the edges connected to the adjacent node may be added to obtain the connection degree corresponding to the adjacent node.

In some embodiments, the connectivity corresponding to the current node may be multiplied by the connectivity corresponding to the neighboring node, and divided by the sum of the connectivity corresponding to the target resource transfer relationship graph to obtain the reference modularity. And calculating the difference value obtained by subtracting the reference modularity from the current modularity, and taking the difference value as a modularity increment. As shown in formula (2), wherein W in the formula _ij Representing the connectivity of an edge between a current node i and its neighboring node j, W _i Representing the connectivity corresponding to the current node, W _j Represents the connection degree corresponding to the adjacent node j, tot represents the sum of the connection degrees in the target resource transfer relation graph, and D _ij Representing module degree increments.

The calculation principle of the formula (2) is as follows, and the expected value obtained by subtracting the ratio of the edge connecting the nodes in the current node group to the two nodes in the same community structure can be used for representing the increase of the cluster connection strength. As shown in equation (3). Wherein θ _ij And when the node i and the node j have connected edges, the value is 1, otherwise, the value is 0. Since tot is the same for the same target resource transfer relationship graph, equation 3 can be simplified to obtain equation (2) to increase the speed of calculating the modularity increment.

Step S706, judging whether the module degree increment meets the preset condition.

Specifically, the preset increment condition may be that the module degree increment is greater than a preset threshold, for example, 0. If so, step S708 is entered. If not, the update of the current resource transfer relationship diagram is terminated, and the finally obtained current resource transfer relationship diagram is taken as the terminated resource transfer relationship diagram to enter step S710.

Step S708, merging the nodes in the current node group into new nodes to obtain the updated current resource transfer relation graph.

Specifically, when a plurality of adjacent nodes exist in the current node, that is, when a plurality of current node groups can be formed, the node group with the largest module increment can be obtained, and the nodes in the node group are combined into a new node, so that an updated current resource transfer relation diagram is obtained.

It can be understood that, in the updated current resource transfer relationship graph, the connectivity of the edge corresponding to the merged node includes the connectivity of the edge corresponding to the current node and the connectivity of the edge corresponding to the adjacent node. And a new node may have a self-loop, i.e. the node itself also has an internal connectivity that is twice the connectivity of the edges between the current node and the neighboring nodes. For example, if the current node i and the adjacent node j are combined to obtain a new node O, the sum W of connectivity of edges corresponding to the new node O _o Is W _i +W _j . And the internal connectivity W _oo ＝2W _ij . After the updated current resource transfer relation diagram is obtained, the method can continue to enter the step of obtaining the current node to be divided in the current resource transfer relation diagram, and when the current node and the corresponding adjacent node form the current node group, the corresponding module degree increment is calculated until the preset increment condition is met, so that the resource transfer relation diagram is terminated.

Step S710, dividing the users corresponding to the same node in the termination resource transfer relation diagram into the same target resource transfer user group.

Specifically, the final obtained current resource transfer relation diagram is taken as a termination resource transfer relation diagram, and because nodes in the termination resource transfer relation diagram may be obtained by combining, one node can correspond to a plurality of users, and therefore users corresponding to the same node can be divided into the same target resource transfer user group.

For example, assuming that the node 1 and the node 2 in fig. 6 are combined to obtain the current node group 1, a corresponding module degree increment may be calculated, and if the module degree increment meets a preset increment condition, the node 1 and the node 2 may be combined to obtain an updated current resource transfer relationship diagram, as shown in fig. 8, where the node 0 represents a node obtained by combining the node 1 and the node 2. Merging … … of node 0 with node 3 may continue on the updated current resource transfer relationship graph. Assuming that the final obtained termination resource transfer relationship diagram is fig. 9 and includes nodes O, P and Q, assuming that node O is obtained by combining node 1 with node 5, node P is obtained by combining node 6 with node 9, and node Q is obtained by combining node 10 with node 13, users 1 to 5 form a target resource transfer user group 1, users 6 to 9 form a target resource transfer user group 2, and users 10 to 13 form a target resource transfer user group 3.

In some embodiments, as shown in fig. 10, the step of calculating the first target relationship weight and the second target relationship weight through the corresponding weight algorithm may include:

step S1002, obtaining training resource transfer data between training users in a training user set, where the training user set includes a plurality of training users.

Specifically, the training user is a user for training the weighted. The training user may be a malicious user, e.g. a fraudster user, e.g. the training user may also be determined from the resource transfer user to which the target risk resource corresponds. Training resource transfer data refers to resource transfer data between training users, and may include at least one of the number of resource transfers and the number of resource transfers.

Step S1004, third resource transfer data between the kth training user and the h training user and fourth resource transfer data between the kth training user and the h training user are obtained, k is less than or equal to m, h is less than or equal to m, k and h are positive integers, m is the number of training users in the training user set, the third resource transfer data is transfer data corresponding to abnormal transfer resources, and the fourth resource transfer data is transfer data corresponding to non-abnormal transfer resources.

Specifically, h and k may be any positive integers, i.e., the h training user and the k training user are any users in the training user set, where h is different from k. For example, a degree of resource transfer relationship between two users in a training set of users may be calculated. For two users that do not have a resource transfer relationship, the corresponding resource transfer relationship degree may be a preset value, for example, 0.

Step S1006, a third current relation degree is obtained according to the third resource transfer data and the corresponding first current relation weight, and a fourth current relation degree is obtained according to the fourth resource transfer data and the corresponding second current relation weight.

Specifically, the current relationship weight refers to the current relationship weight. Before the first adjustment, the first current relation weight and the second current relation weight may be random values, or may be preset initial weights, for example, may be 1.

And multiplying the third resource transfer data by the corresponding first current relation weight to obtain a third current relation degree. Multiplying the fourth resource transfer data by the corresponding second current relation weight to obtain a fourth current relation degree

And step S1008, obtaining the training resource transfer relation degree between the kth training user and the h training user according to the third current relation degree and the fourth current relation degree.

Specifically, the third current relationship degree and the fourth current relationship degree may be added to obtain a training resource transfer relationship degree between the kth training user and the h training user. For example, the training resource transfer relationship between the kth training user and the h training user can be obtained by referring to formula 1. I.e., the method of calculating the target resource transfer relationships and training the resource transfer relationships is consistent.

Step S1010, taking training users in the training user set as training nodes, and taking the training resource transfer relation degree between the training users as the connection degree of edges between the training nodes to obtain a training resource transfer relation diagram.

Specifically, the training node refers to a node in the training resource transfer graph, and the method for obtaining the training resource transfer relationship graph may refer to the method for obtaining the target resource transfer relationship graph, which is not described herein.

Step S1012, carrying out group division on the training user set according to the training resource transfer relation diagram to obtain at least one current training resource transfer user group.

Specifically, the training resource transfer user group refers to a resource transfer user group obtained by dividing when training the relation weight. The current training resource transfer user group refers to a training resource transfer user group obtained by current division, and if the training is needed for multiple times, the current training resource transfer user group is updated along with the change of the training. The step of performing group division on the training user set according to the training resource transfer relationship diagram to obtain at least one training resource transfer user group may refer to the step of obtaining the target resource transfer user group in step S208, which is not described in detail.

Step S1014, calculating the user similarity corresponding to the training resource transfer user group, so as to adjust the first current relationship weight and the second current relationship weight according to the user similarity corresponding to the current training resource transfer user group, taking the adjusted first current relationship weight as a first target relationship weight, and taking the adjusted second current relationship weight as a second target relationship weight.

Specifically, the user similarity is used to represent the degree of similarity of users, and the higher the similarity, the more similar the representation. The calculation method of the user similarity can be set according to the requirement. The similarity between users may be determined based on user characteristics, such as at least one of user attribute information or user behavior. The user similarity is obtained by calculation according to at least one of the online time period of the user account, the geographical position of the user or the ip address used for logging in the account. For example, it may be set that the similarity is 1 when the online time period belongs to the same time period, and is 2 when the ip address belongs to the same lan. The user characteristics can also be acquired, the user characteristics are input into a user vector determination model, the user vectors corresponding to the users are obtained, and the user similarity is obtained through calculation according to the similarity between the user vectors. The user vector determination model may be a machine learning model.

The user similarity corresponding to the training resource transfer user group may be a statistical value of the similarity between users in the training resource transfer user group, for example, may be an average value or a sum of the similarities. For example, suppose that the training resource user group includes users B, C and D. And if the similarity between the user B and the user C is 2, the similarity between the user B and the user D is 3, and the similarity between the user C and the user D is 4, the similarity of the users corresponding to the training resource transfer user group is (2+3+4)/3=3.

After the user similarity is obtained, the first current relation weight and the second current relation weight can be adjusted according to the user similarity corresponding to the current training resource transfer user group, and the adjustment times can be one or more times. The first current relation weight and the second current relation weight can be adjusted according to the user similarity corresponding to the training resource transfer user group, and the adjustment can be performed manually or automatically.

In some embodiments, the first current relationship weight and the second current relationship weight may be adjusted toward a direction in which the user similarity corresponding to the user group is transferred to the training resource, so as to obtain an updated first current relationship weight and second current relationship weight. And returning to the step S1006, namely repeatedly executing the step S1006 until the user similarity is greater than the preset similarity, wherein the step is used for adjusting the first current relation weight and the second current relation weight according to the user similarity corresponding to the training resource transfer user group. The preset similarity may be set as required, for example, may be 0.9.

For practical examples, assuming that the initial values of a1, a2, a3 and a4 in equation 1 are all set to be 1, the training resource transfer user group is obtained by dividing according to the initial values, the user similarity corresponding to the training resource transfer user group is calculated, and assuming that the average user similarity between users in the training resource transfer user group obtained by calculation is 0.5 and does not satisfy the preset similarity of 0.9, a1, a2, a3 and a4 can be adjusted, wherein the adjustment principle is that the user similarity corresponding to the training resource transfer user group is increased. After the adjusted a1, a2, a3 and a4 are obtained, updating the connectivity of the edges in the training resource transfer relation graph, carrying out group division based on the updated connectivity to obtain an updated current training resource transfer user group, calculating the user similarity corresponding to the updated training resource transfer user group, and taking the updated a1, a2, a3 and a4 as final relation weights if the similarity is greater than 0.9. If the similarity is less than 0.9, the adjustments a1, a2, a3, and a4 may be continued.

The method provided by the embodiment of the application can be used for detecting a partner with malicious resources, such as a fraud partner, and the method provided by the embodiment of the application is described below by taking the detection of the fraud partner as an example, and comprises the following steps:

1. And acquiring a user receiving the target risk resource, wherein the target risk resource is an abnormally transferred resource.

In particular, the target risk resource may refer to an illegally acquired resource, such as a fraudulent money. The user receiving the target risk resource is an illegal user, such as a fraudulent user.

2. And obtaining a resource transfer user sequence taking the user receiving the target risk resource as an initial user according to the resource transfer flow direction of the target risk resource.

Specifically, after the fraud user receives the target risk resource, the transfer condition of the target risk resource can be longitudinally expanded, the flow direction of the money which is fraudulently received can be obtained, namely, the money which is fraudulently received is forwarded to which cards, the first-stage descending tracing is performed until the cash card is traced, namely, the cash card is taken out, and the resource transfer user sequence is obtained.

3. Acquiring a resource associated user corresponding to a resource transfer user; and obtaining a target user set according to the resource associated user and the resource transfer user sequence.

Specifically, the card longitudinally expanded in the step 3 is used as a malicious card, namely a card for money laundering. And acquiring the transaction conditions of the malicious cards, obtaining the cards which are transacted with the malicious cards, and acquiring the resource transfer characteristics of the suspected cards, such as the transaction times and the transaction amount, as the suspected cards. And determining whether the card is a malicious card according to the resource transfer characteristics of the suspected card, and if so, taking the card as a target card. And finally, the users corresponding to all the obtained malicious cards form a target user set.

4. And obtaining the target resource transfer relation degree between the target users according to the target resource transfer data between the target users.

Specifically, the target resource transfer relationship degree between users can be obtained according to the transaction data, such as the transaction times and the transaction amounts, between the malicious cards.

5. And taking the target users in the target user set as target nodes, and obtaining the connection degree corresponding to the edges between the nodes according to the target resource transfer relation degree between the target users to obtain a target resource transfer relation diagram.

Specifically, the users are taken as nodes, and the target resource transfer relation between the users is taken as the weight of the edge connected with the nodes, namely the corresponding connection degree of the edge, so as to obtain a target resource transfer relation diagram.

6. And obtaining the current node to be divided in the current resource transfer relation diagram, and calculating the corresponding module degree increment when the current node and the corresponding adjacent nodes form a current node group.

Specifically, when dividing for the first time, the target resource transfer relationship graph is used as the current resource transfer relationship graph, and all nodes of the current resource transfer relationship graph can be used as current nodes and combined with adjacent nodes to obtain the current node group. And obtaining the modularity gain caused by connecting the current node and the adjacent nodes to form the node group relative to any two nodes to form the node group, and taking the modularity gain as the modularity increment.

7. Judging whether the module degree increment meets the preset increment condition.

Specifically, the preset increment condition may be, for example, greater than 0. If yes, go to step 8, if no, go to step 9.

8. And merging the nodes in the current node group into new nodes to obtain an updated current resource transfer relation diagram.

Specifically, after the updated current resource transfer relation diagram is obtained, the step 6 is continued.

9. And dividing the users corresponding to the same node in the termination resource transfer relation diagram into the same target resource transfer user group.

Specifically, when the modularity increment does not meet the preset increment condition, the current resource transfer relation diagram obtained by final updating is used as a termination resource transfer relation diagram, and the users corresponding to the same node in the termination resource transfer relation diagram belong to the same resource transfer user group, for example, belong to the same fraud group.

With the development of technology, money laundering is increasingly transferred to a network, so are the water house partners of the counterfeit public inspection method, master a large number of bank cards of different banks on hand, distribute large amount of black money to a plurality of small amount of cash-taking cards through a multi-stage transfer network, and after the special cash-taking partners obtain the cash-taking cards, quick batch cash taking can be carried out, and after cash taking is finished, tracking is difficult, so that the purpose of money laundering is achieved. By applying the method provided by the embodiment of the application to detection of fraud partners, the relation diagram among the water-house partner accounts can be obtained by analyzing the fund flow of fraud cases of the counterfeit public inspection method, and the water-house partners can be rapidly and automatically detected through a community division algorithm, so that timeliness and accuracy are high, related personnel are released from the originally tedious and time-consuming partner division work, the labor cost is greatly reduced, and the public security police can conveniently collect evidence, locate and capture the partners, thereby reducing the loss of the user when being cheated.

After the target resource transfer user group is obtained, group information of the target resource transfer user group, such as at least one of the number of users, the activity condition, the number of login devices or the activity area, can be obtained and sent to the corresponding terminal, so as to provide criminals for police, and can perform accurate striking for certain groups.

As shown in fig. 11, in one embodiment, a resource transfer user group determining apparatus is provided, and the resource transfer user group determining apparatus may be integrated in the server 120, and specifically may include a target resource transfer data obtaining module 1102, a target resource transfer relationship degree obtaining module 1104, a target resource transfer relationship diagram obtaining module 1106, and a target resource transfer user group dividing module 1108.

And a target resource transfer data acquisition module 1102, configured to acquire target resource transfer data between target users in a target user set, where the target users are resource transfer related users corresponding to target risk resources.

The target resource transfer relationship obtaining module 1104 is configured to obtain a target resource transfer relationship between target users according to the target resource transfer data between the target users.

The target resource transfer relationship diagram obtaining module 1106 is configured to obtain a connection degree corresponding to edges between nodes according to the target resource transfer relationship degree between target users by using target users in the target user set as target nodes, and obtain a target resource transfer relationship diagram.

The target resource transfer user group dividing module 1108 is configured to divide the target user set into groups according to the target resource transfer relationship graph, so as to obtain at least one target resource transfer user group.

In some embodiments, the means for obtaining the set of target users comprises: the receiving user acquisition unit is used for acquiring a user receiving target risk resources, wherein the target risk resources are abnormal transfer resources; a resource transfer user sequence obtaining unit, configured to obtain a resource transfer user sequence using a user receiving the target risk resource as an initial user according to a resource transfer flow direction of the target risk resource; and the target user set obtaining unit is used for obtaining the target user set according to the resource transfer user sequence.

In some embodiments, the target user set derivation unit is to: acquiring resource transfer characteristics corresponding to each resource transfer user in a resource transfer user sequence; and screening the target users from the resource transfer user sequences according to the resource transfer characteristics corresponding to the resource transfer users to form a target user set.

In some embodiments, the target user set derivation unit is to: for a resource transfer user in a resource transfer user sequence, acquiring a resource association user corresponding to the resource transfer user; and obtaining a target user set according to the resource associated user and the resource transfer user sequence.

In some embodiments, the target resource transfer relatedness derivation module 1106 includes:

the resource transfer data acquisition unit is used for acquiring first resource transfer data between the ith target user and the jth target user and second resource transfer data between the ith target user and the jth target user, wherein i is less than or equal to n, j is less than or equal to n, i and j are positive integers, n is the number of target users in the target user set, the first resource transfer data is transfer data corresponding to abnormal transfer resources, and the second resource transfer data is transfer data corresponding to non-abnormal transfer resources.

The first relation degree and the second relation degree obtaining unit are used for obtaining the first relation degree according to the first resource transfer data and the corresponding first target relation weight, and obtaining the second relation degree according to the second resource transfer data and the corresponding second target relation weight.

The target resource transfer relation obtaining unit is used for obtaining the target resource transfer relation between the ith target user and the jth target user according to the first relation and the second relation.

In some embodiments, the obtaining of the first target relationship weight and the second target relationship weight includes:

the training user acquisition unit is used for acquiring training resource transfer data among training users in the training user set, and the training user set comprises a plurality of training users.

The training resource transfer data acquisition unit is used for acquiring third resource transfer data between a kth training user and a h training user and fourth resource transfer data between the kth training user and the h training user, wherein k is less than or equal to m, h is less than or equal to m, k and h are positive integers, m is the number of training users in the training user set, the third resource transfer data is transfer data corresponding to abnormal transfer resources, and the fourth resource transfer data is transfer data corresponding to non-abnormal transfer resources.

The second relation degree and third relation degree obtaining unit is used for obtaining a third current relation degree according to the third resource transfer data and the corresponding first current relation weight, and obtaining a fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight.

The training resource transfer relation obtaining unit is used for obtaining the training resource transfer relation between the kth training user and the h training user according to the third current relation and the fourth current relation.

The training resource transfer relation diagram obtaining unit is used for obtaining the training resource transfer relation diagram by taking training users in the training user set as training nodes and taking the training resource transfer relation degree between the training users as the connection degree of edges between the training nodes.

The training resource transfer user group dividing unit is used for carrying out group division on the training user set according to the training resource transfer relation diagram to obtain at least one current training resource transfer user group.

The adjusting unit is used for calculating the user similarity corresponding to the current training resource transfer user group so as to adjust the first current relation weight and the second current relation weight according to the user similarity corresponding to the current training resource transfer user group; and taking the adjusted first current relation weight as a first target relation weight and taking the adjusted second current relation weight as a second target relation weight.

In some embodiments, the adjustment unit is to: the first current relation weight and the second current relation weight are adjusted towards the direction of increasing the user similarity corresponding to the current training resource transfer user group, and updated first current relation weight and second current relation weight are obtained;

The apparatus further comprises: and the return module is used for returning to the step of obtaining the third current relation degree according to the third resource transfer data and the corresponding first current relation weight and obtaining the fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight until the similarity of the corresponding users of the current training resource transfer user group is larger than the preset similarity.

In some embodiments, the target resource transfer user group partitioning module 1108 is configured to: taking the target resource transfer relation diagram as a current resource transfer relation diagram; obtaining a current node to be divided in a current resource transfer relation diagram, and calculating a corresponding module degree increment when the current node and a corresponding adjacent node form a current node group; when the module degree increment meets the preset increment condition, merging the nodes in the current node group into new nodes to obtain an updated current resource transfer relation diagram; and returning to obtain the current node to be divided in the current resource transfer relation diagram based on the updated current resource transfer relation diagram, and calculating the corresponding module degree increment when the current node and the corresponding adjacent node form the current node group until the update of the current resource transfer relation diagram is terminated, thereby obtaining the termination resource transfer relation diagram. And dividing the users corresponding to the same node in the termination resource transfer relation diagram into the same target resource transfer user group.

In some embodiments, the target resource transfer user group partitioning module is to: calculating the current modularity corresponding to the current node group; calculating according to the node connectivity corresponding to the current node and the node connectivity corresponding to the adjacent node to obtain a reference modularity; and obtaining corresponding module degree increment when the current node and the corresponding adjacent node form the current node group according to the difference between the current module degree and the reference module degree.

FIG. 12 illustrates an internal block diagram of a computer device in one embodiment. The computer device may be specifically the server 120 of fig. 1. As shown in fig. 12, the computer device includes a processor, a memory, and a network interface connected by a system bus. The memory includes a nonvolatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system, and may also store a computer program that, when executed by a processor, causes the processor to implement a resource transfer user group determination method. The internal memory may also have stored therein a computer program which, when executed by the processor, causes the processor to perform a resource transfer user group determination method.

It will be appreciated by those skilled in the art that the structure shown in fig. 12 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, the resource transfer user group determining apparatus provided herein may be implemented in the form of a computer program that is executable on a computer device as shown in fig. 12. The memory of the computer device may store various program modules constituting the resource transfer user group determining apparatus, for example, a target resource transfer data acquisition module 1102, a target resource transfer relationship degree obtaining module 1104, a target resource transfer relationship diagram obtaining module 1106, and a target resource transfer user group dividing module 1108 shown in fig. 11. The computer program constituted by the respective program modules causes the processor to execute the steps in the resource transfer user group determination method of the respective embodiments of the present application described in the present specification.

For example, the computer device shown in fig. 12 may obtain, by using the target resource migration data obtaining module 1102 in the resource migration user group determining apparatus shown in fig. 11, target resource migration data between target users in the target user set, where the target users are resource migration related users corresponding to the target risk resources. The target resource transfer relationship between the target users is obtained by the target resource transfer relationship obtaining module 1104 according to the target resource transfer data between the target users. And the target resource transfer relation diagram obtaining module 1106 obtains the connection degree corresponding to the edges between the nodes according to the target resource transfer relation degree between the target users by taking the target users in the target user set as target nodes. And performing group division on the target user set according to the target resource transfer relationship diagram through a target resource transfer user group division module 1108 to obtain at least one target resource transfer user group.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the resource transfer user group determination method described above. The steps of the resource transfer user group determination method herein may be the steps in the resource transfer user group determination method of the above-described respective embodiments.

In one embodiment, a computer readable storage medium is provided, storing a computer program which, when executed by a processor, causes the processor to perform the steps of the resource transfer user group determination method described above. The steps of the resource transfer user group determination method herein may be the steps in the resource transfer user group determination method of the above-described respective embodiments.

It should be understood that, although the steps in the flowcharts of the embodiments of the present application are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in various embodiments may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of the sub-steps or stages of other steps or other steps.

Those skilled in the art will appreciate that implementing all or part of the above-described methods may be accomplished by way of a computer program, which may be stored on a non-transitory computer readable storage medium and which, when executed, may comprise the steps of the above-described embodiments of the methods. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.

The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.

Claims

1. A method of resource transfer user group determination, the method comprising:

acquiring target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources;

performing weighted calculation on the target resource transfer data among the target users to obtain the target resource transfer relation degree among the target users;

Taking target users in the target user set as target nodes, taking target resource transfer relation between the target users as first connection degree, obtaining second connection degree according to the similarity of geographic positions, adding the first connection degree and the second connection degree to obtain connection degree corresponding to edges between the corresponding target nodes, or taking the target resource transfer relation between the target users as connection degree corresponding to edges between the corresponding target nodes to obtain a target resource transfer relation graph;

and carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group.

2. The method of claim 1, wherein the step of obtaining the set of target users comprises:

acquiring a user receiving a target risk resource, wherein the target risk resource is an abnormally transferred resource;

obtaining a resource transfer user sequence taking the user receiving the target risk resource as an initial user according to the resource transfer flow direction of the target risk resource;

and obtaining a target user set according to the resource transfer user sequence.

3. The method of claim 2, wherein the obtaining the set of target users from the sequence of resource transfer users comprises:

acquiring resource transfer characteristics corresponding to each resource transfer user in the resource transfer user sequence;

and screening the resource transfer user sequence according to the resource transfer characteristics corresponding to the resource transfer users to obtain target users to form a target user set.

4. The method of claim 2, wherein the obtaining the set of target users from the sequence of resource transfer users comprises:

for the resource transfer users in the resource transfer user sequence, acquiring resource association users corresponding to the resource transfer users;

and obtaining a target user set according to the resource associated user and the resource transfer user sequence.

5. The method of claim 1, wherein weighting the target resource transfer data between the target users to obtain a target resource transfer relationship between the target users comprises:

acquiring first resource transfer data between an ith target user and a jth target user, and second resource transfer data between the ith target user and the jth target user, wherein i is less than or equal to n, j is less than or equal to n, i and j are positive integers, n is the number of target users in a target user set, the first resource transfer data is transfer data corresponding to abnormal transfer resources, and the second resource transfer data is transfer data corresponding to non-abnormal transfer resources;

Obtaining a first relationship degree according to the first resource transfer data and the corresponding first target relationship weight, and obtaining a second relationship degree according to the second resource transfer data and the corresponding second target relationship weight;

and obtaining the target resource transfer relation degree between the ith target user and the jth target user according to the first relation degree and the second relation degree.

6. The method of claim 5, wherein the deriving of the first target relationship weight and the second target relationship weight comprises:

acquiring training resource transfer data among training users in a training user set, wherein the training user set comprises a plurality of training users;

acquiring third resource transfer data between a kth training user and an h training user and fourth resource transfer data between the kth training user and the h training user, wherein k is less than or equal to m, h is less than or equal to m, k and h are positive integers, m is the number of training users in a training user set, the third resource transfer data is transfer data corresponding to abnormal transfer resources, and the fourth resource transfer data is transfer data corresponding to non-abnormal transfer resources;

Obtaining a third current relation degree according to the third resource transfer data and the corresponding first current relation weight, and obtaining a fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight;

obtaining a training resource transfer relation degree between the kth training user and the h training user according to the third current relation degree and the fourth current relation degree;

training users in the training user set are used as training nodes, and the training resource transfer relation degree between the training users is used as the connection degree of edges between the training nodes, so that a training resource transfer relation diagram is obtained;

performing group division on the training user set according to the training resource transfer relation diagram to obtain at least one current training resource transfer user group;

calculating the user similarity corresponding to the current training resource transfer user group so as to adjust the first current relation weight and the second current relation weight according to the user similarity corresponding to the current training resource transfer user group; and taking the adjusted first current relation weight as the first target relation weight, and taking the adjusted second current relation weight as the second target relation weight.

7. The method of claim 6, wherein adjusting the first current relationship weight and the second current relationship weight according to the user similarity corresponding to the current training resource transfer user group comprises:

the first current relation weight and the second current relation weight are adjusted towards the direction of increasing the user similarity corresponding to the current training resource transfer user group, and updated first current relation weight and second current relation weight are obtained;

the method further comprises the steps of:

and returning to the step of obtaining a third current relation degree according to the third resource transfer data and the corresponding first current relation weight and obtaining a fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight until the similarity of the corresponding users of the current training resource transfer user group is greater than a preset similarity.

8. The method of claim 1, wherein the grouping the set of target users according to the target resource transfer relationship graph to obtain at least one target resource transfer user group comprises:

taking the target resource transfer relation diagram as a current resource transfer relation diagram;

Obtaining a current node to be divided in a current resource transfer relation diagram, and calculating a corresponding module degree increment when the current node and a corresponding adjacent node form a current node group;

when the modularity increment meets a preset increment condition, merging the nodes in the current node group into new nodes to obtain an updated current resource transfer relation diagram;

returning to obtain a current node to be divided in the current resource transfer relation diagram based on the updated current resource transfer relation diagram, and calculating a corresponding module degree increment when the current node and a corresponding adjacent node form a current node group until the update of the current resource transfer relation diagram is terminated, so as to obtain a termination resource transfer relation diagram;

and dividing the users corresponding to the same node in the termination resource transfer relation diagram into the same target resource transfer user group.

9. The method of claim 8, wherein calculating the corresponding module degree increment when the current node and the corresponding neighboring node are grouped into the current node group comprises:

calculating the current modularity corresponding to the current node group;

calculating according to the node connectivity corresponding to the current node and the node connectivity corresponding to the adjacent node to obtain a reference modularity;

And obtaining corresponding module degree increment when the current node and the corresponding adjacent node form a current node group according to the difference between the current module degree and the reference module degree.

10. A resource transfer user group determination apparatus, the apparatus comprising:

the target resource transfer data acquisition module is used for acquiring target resource transfer data among target users in a target user set, wherein the target users are resource transfer related users corresponding to target risk resources;

the target resource transfer relation degree obtaining module is used for carrying out weighted calculation on target resource transfer data among the target users to obtain target resource transfer relation degree among the target users;

the target resource transfer relation diagram obtaining module is used for obtaining a second connection degree according to the similarity of geographic positions by taking target users in the target user set as target nodes, taking the target resource transfer relation degree between the target users as a first connection degree, adding the first connection degree and the second connection degree to obtain the connection degree corresponding to edges between the corresponding target nodes, or taking the target resource transfer relation degree between the target users as the connection degree corresponding to edges between the corresponding target nodes to obtain a target resource transfer relation diagram;

And the target resource transfer user group dividing module is used for carrying out group division on the target user set according to the target resource transfer relation diagram to obtain at least one target resource transfer user group.

11. The apparatus of claim 10, wherein the means for obtaining the set of target users comprises:

a receiving user obtaining unit, configured to obtain a user who receives a target risk resource, where the target risk resource is an abnormally transferred resource;

a resource transfer user sequence obtaining unit, configured to obtain a resource transfer user sequence using the user receiving the target risk resource as an initial user according to a resource transfer flow direction of the target risk resource;

and the target user set obtaining unit is used for obtaining the target user set according to the resource transfer user sequence.

12. The apparatus of claim 10, wherein the target resource transfer relationship obtaining module comprises:

the resource transfer data acquisition unit is used for acquiring first resource transfer data between an ith target user and a jth target user and second resource transfer data between the ith target user and the jth target user, wherein i is less than or equal to n, j is less than or equal to n, i and j are positive integers, n is the number of target users in a target user set, the first resource transfer data is transfer data corresponding to abnormal transfer resources, and the second resource transfer data is transfer data corresponding to non-abnormal transfer resources;

The first relation degree and second relation degree obtaining unit is used for obtaining a first relation degree according to the first resource transfer data and the corresponding first target relation weight, and obtaining a second relation degree according to the second resource transfer data and the corresponding second target relation weight;

13. The apparatus of claim 12, wherein the means for deriving the first target relationship weight and the second target relationship weight comprises:

the training user acquisition unit is used for acquiring training resource transfer data among training users in a training user set, wherein the training user set comprises a plurality of training users;

the training resource transfer data acquisition unit is used for acquiring third resource transfer data between a kth training user and an h training user and fourth resource transfer data between the kth training user and the h training user, wherein k is less than or equal to m, h is less than or equal to m, k and h are positive integers, m is the number of training users in a training user set, the third resource transfer data is transfer data corresponding to abnormal transfer resources, and the fourth resource transfer data is transfer data corresponding to non-abnormal transfer resources;

The third relation degree and fourth relation degree obtaining unit is used for obtaining a third current relation degree according to the third resource transfer data and the corresponding first current relation weight, and obtaining a fourth current relation degree according to the fourth resource transfer data and the corresponding second current relation weight;

the training resource transfer relation degree obtaining unit is used for obtaining the training resource transfer relation degree between the kth training user and the h training user according to the third current relation degree and the fourth current relation degree;

the training resource transfer relation diagram obtaining unit is used for obtaining a training resource transfer relation diagram by taking training users in the training user set as training nodes and taking the training resource transfer relation degree between the training users as the connection degree of edges between the training nodes;

the training resource transfer user group dividing unit is used for dividing the training user set into groups according to the training resource transfer relation diagram to obtain at least one current training resource transfer user group;

the adjusting unit is used for calculating the user similarity corresponding to the current training resource transfer user group so as to adjust the first current relation weight and the second current relation weight according to the user similarity corresponding to the current training resource transfer user group; and taking the adjusted first current relation weight as the first target relation weight, and taking the adjusted second current relation weight as the second target relation weight.

14. A computer device comprising a memory and a processor, the memory having stored therein a computer program which, when executed by the processor, causes the processor to perform the steps of the resource transfer user group determination method of any of claims 1 to 9.

15. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, causes the processor to perform the steps of the resource transfer user group determination method of any of claims 1 to 9.