Methods for identifying financial fraud gangs, optical quantum computers and media

Abstract

This application provides a method, a quantum computer, and a medium for identifying financial fraud gangs. The method includes: acquiring a first graph structure corresponding to a target transaction network and acquiring a second graph structure corresponding to a financial fraud gang pattern; selecting at least one target subgraph from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure; for each target subgraph in the at least one target subgraph, acquiring the first feature scalar values ​​of all first subgraphs of the target subgraph to form a first set, acquiring the second feature scalar values ​​of all second subgraphs of the second graph structure to form a second set; if the first set and the second set are determined to match based on the first feature scalar values ​​in the first set and the second feature scalar values ​​in the second set, then all nodes corresponding to the target subgraph are confirmed as first candidate financial fraud gangs. The technical solution of this application can improve the accuracy of identifying financial fraud gangs.

Description

Technical Field

[0001] This application relates to the field of financial network transaction risk identification technology, specifically to a method for identifying financial fraud gangs, an optical quantum computer, and a medium. Background Technology

[0002] Financial anti-fraud technology, as a core means of ensuring transaction security, can be used to identify suspicious transactions and prevent financial losses. However, existing financial anti-fraud technologies lack sufficient accuracy in identifying complex and covert organized fraud groups. Summary of the Invention

[0003] In view of this, embodiments of this application provide a method for identifying financial fraud gangs, an optical quantum computer, and a medium, which can improve the accuracy of identifying financial fraud gangs.

[0004] In a first aspect, embodiments of this application provide a method for identifying financial fraud gangs, comprising: obtaining a first graph structure corresponding to a target transaction network and obtaining a second graph structure corresponding to a financial fraud gang pattern; selecting at least one target subgraph from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure; for each target subgraph in the at least one target subgraph, performing the following steps: obtaining first feature scalar values ​​of all first subgraphs of the target subgraph, wherein the first feature scalar values ​​are used to characterize the number of perfect matches or the sum of weights of perfect matches of the first subgraph, constituting a first set; obtaining second feature scalar values ​​of all second subgraphs of the second graph structure, wherein the second feature scalar values ​​are used to characterize the number of perfect matches or the sum of weights of perfect matches of the second subgraph, constituting a second set; if, based on the first feature scalar values ​​in the first set and the second feature scalar values ​​in the second set, it is determined that the first set matches the second set, confirming that all nodes corresponding to the target subgraph are first candidate financial fraud gangs of the first financial fraud gang, and the transaction behavior pattern of the first financial fraud gang matches the financial fraud gang pattern.

[0005] In one embodiment, the financial fraud gang identification method further includes: performing Gaussian boson sampling on the first graph structure to obtain the sample probability distribution corresponding to the subgraphs of the first graph structure; selecting at least one dense subgraph with an occurrence probability higher than a preset threshold from at least one first subgraph of the target subgraph according to the first sample probability distribution; for each dense subgraph in the at least one dense subgraph, if the dense subgraph is not the target subgraph, confirming all nodes corresponding to the dense subgraph as the second candidate financial fraud gang, and the transaction behavior pattern of the second financial fraud gang is inconsistent with the financial fraud gang pattern.

[0006] In one embodiment, obtaining the first feature scalar value of all first subgraphs of the target subgraph includes: performing Gaussian boson sampling on the first graph structure using an optical quantum computer to obtain the first sample probability distribution corresponding to all subgraphs of the first graph structure; and calculating the first feature scalar value of all first subgraphs based on the first sample probability distribution.

[0007] In one embodiment, obtaining the second feature scalar values ​​of all second subgraphs of the second graph structure includes:

[0008] Gaussian boson sampling of the second graph structure is performed using an optical quantum computer to obtain the second sample probability distribution corresponding to all second subgraphs; based on the second sample probability distribution, the second feature scalar value of all second subgraphs is calculated.

[0009] In one embodiment, the financial fraud gang identification method further includes: if the first candidate financial fraud gang is the first financial fraud gang, decoding all nodes corresponding to the first financial fraud gang to obtain the financial entity corresponding to each node; and inputting the financial entity into the financial risk control system to trigger risk disposal operations on the financial entity.

[0010] In one embodiment, selecting at least one target subgraph with the same number of nodes as the second graph structure from the subgraphs of the first graph structure includes:

[0011] From the subgraphs of the first graph structure, select at least one target subgraph that has the same number of nodes as the second graph structure and meets the preset conditions. The vector sample corresponding to each target subgraph that meets the preset conditions contains only the first value and the second value. The vector sample is used to characterize the photon click pattern corresponding to the target subgraph.

[0012] In one embodiment, the first graph structure includes an undirected graph, and the second graph structure includes an undirected graph.

[0013] Secondly, embodiments of this application provide a financial fraud gang identification device, comprising: an acquisition module, configured to acquire a first graph structure corresponding to a target transaction network and a second graph structure corresponding to a financial fraud gang pattern; a selection module, configured to select at least one target subgraph from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure; and an execution module, configured to perform the following steps for each target subgraph in the at least one target subgraph: acquiring first feature scalar values ​​of all first subgraphs of the target subgraph, wherein the first feature scalar values ​​are used to characterize the number of perfect matches or the sum of weights of perfect matches of the first subgraph, constituting a first set; acquiring second feature scalar values ​​of all second subgraphs of the second graph structure, wherein the second feature scalar values ​​are used to characterize the number of perfect matches or the sum of weights of perfect matches of the second subgraph, constituting a second set; if, based on the first feature scalar values ​​in the first set and the second feature scalar values ​​in the second set, it is determined that the first set matches the second set, confirming that all nodes corresponding to the target subgraph are first candidate financial fraud gangs of the first financial fraud gang, and the transaction behavior pattern of the first financial fraud gang matches the financial fraud gang pattern.

[0014] Thirdly, embodiments of this application provide an optical quantum computer, including: a quantum light source, an optical quantum chip, and a single-photon detector. The quantum light source, the optical quantum chip, and the single-photon detector are used to execute the financial fraud gang identification method described in the first aspect.

[0015] Fourthly, embodiments of this application provide an electronic device, including: a processor; and a memory for storing processor-executable instructions, wherein the processor is used to execute the financial fraud gang identification method of the first aspect described above.

[0016] Fifthly, embodiments of this application provide a computer-readable storage medium storing a computer program for performing the financial fraud gang identification method of the first aspect described above.

[0017] Sixthly, embodiments of this application provide a computer program product, which includes a computer program that, when executed by a processor of a computer device, enables the computer device to perform the financial fraud gang identification method described in the first aspect.

[0018] In a seventh aspect, embodiments of this application provide a chip, including: a processor; and a memory for storing processor-executable instructions, wherein the processor is used to execute the financial fraud gang identification method of the first aspect described above.

[0019] This application provides a method for identifying financial fraud gangs. In all subgraphs of a first graph structure corresponding to a target transaction network, a target subgraph with the same number of nodes as a second graph structure corresponding to a financial fraud gang pattern is selected. The first feature scalar values ​​of all first subgraphs of the target subgraph are obtained to form a first set. The second feature scalar values ​​of all second subgraphs of the second graph structure are obtained to form a second set. The method then determines whether the first set and the second set match. If they match, all nodes corresponding to the target subgraph are confirmed as first candidate financial fraud gangs. Since when two graph structures are isomorphic, the sets of feature scalar values ​​of all subgraphs corresponding to these two graph structures must be the same, this method can accurately match financial fraud gangs whose transaction behavior is completely consistent with known financial fraud gang patterns, significantly improving the accuracy of identifying financial fraud gangs. Attached Figure Description

[0020] Figure 1 The diagram shown is a schematic representation of the system architecture of a financial fraud gang identification system provided in an exemplary embodiment of this application.

[0021] Figure 2 The diagram shown is a flowchart illustrating a method for identifying financial fraud gangs provided in an exemplary embodiment of this application.

[0022] Figure 3 The diagram shown is a flowchart illustrating a method for identifying financial fraud gangs provided in another exemplary embodiment of this application.

[0023] Figure 4 The diagram shown is a schematic representation of the structure of a financial fraud gang identification device provided in an exemplary embodiment of this application.

[0024] Figure 5 The diagram shown is a block diagram of an optical quantum computer for performing a method for identifying financial fraud gangs, provided in an exemplary embodiment of this application.

[0025] Figure 6 The diagram shown is a block diagram of an electronic device for performing a method for identifying financial fraud gangs, provided in an exemplary embodiment of this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] Exemplary System

[0028] Figure 1 The diagram shown is a schematic representation of the system architecture of a financial fraud gang identification system provided in an exemplary embodiment of this application. Figure 1 As shown, the financial fraud gang identification system 100 may include a classical computer 110 and a quantum computer 120. The quantum computer 120 includes a quantum light source 121, a quantum chip 122, a single-photon detector 123, and a processing unit 124. The classical computer 110 and the quantum computer 120 are connected via an interface 130. The interface 130 may include a high-speed serial computer expansion bus, a high-speed network interconnection interface, or other suitable interfaces. The processing unit 124 is used to process the detection data detected by the single-photon detector 123. For example, the processing unit 124 may include a programmable logic device (such as an FPGA), a microprocessor (such as a CPU, ARM processor), etc.

[0029] In one example, a classical computer 110 can obtain a first graph structure corresponding to the target transaction network and a second graph structure corresponding to the financial fraud gang pattern. Based on the topology of the first graph structure, a first adjacency matrix is ​​generated, and based on the topology of the second graph structure, a second adjacency matrix is ​​generated. The first adjacency matrix is ​​decomposed to obtain a unitary matrix U and a diagonal matrix S. The unitary matrix U and the diagonal matrix S are then sent to the photonic quantum chip 122.

[0030] Furthermore, the optical quantum chip 122 can configure the physical parameters of multiple optical elements (such as beam splitters, phase modulators, etc.) in its integrated linear optical network based on the unitary matrix U and the diagonal matrix S, and send trigger signals to the quantum light source 121.

[0031] Furthermore, steps one through three below can be performed a preset number of times. For example, the preset number of times could be one thousand, ten thousand, or other reasonable numbers.

[0032] Step 1: In response to the trigger signal, the quantum light source 121 can generate a photon beam by laser-excited quantum dots or spontaneous parametric downconversion (SPDC) and inject the photon beam into the input port of the optical quantum chip 122.

[0033] Step 2: The photon beam can propagate and evolve within the linear optical network of the quantum chip 122. During this process, the photons, according to the laws of quantum mechanics, undergo a series of optical network evolutions determined by the unitary matrix U, generating an output light field, which is then guided to the single-photon detector 123.

[0034] Step 3: The single-photon detector 123 can detect photons in the output light field, record the output mode of the photons, generate detection data corresponding to the output mode of the photons, and send the detection data to the processing unit 124.

[0035] Furthermore, after executing steps one to three a preset number of times, the processing unit 124 can summarize and statistically analyze the detection data received in each execution, calculate the frequency of the output pattern obtained in each execution, generate a first sample probability distribution corresponding to the first graph structure, and send the first sample probability distribution to the classical computer 110 via interface 130.

[0036] Similarly, the second sample probability distribution can be obtained using the same method and sent to the classical computer 110 via interface 130.

[0037] Furthermore, the classical computer 110 can determine the number of nodes corresponding to each subgraph of the first graph structure based on the probability distribution of the first sample, and select the subgraph with the same number of nodes as the second graph structure from the subgraphs of the first graph structure to obtain the target subgraph.

[0038] Furthermore, the classical computer 110 can select the probability distributions corresponding to all first subgraphs of the target subgraph from the first sample probability distribution to obtain the third sample probability distribution, calculate the first feature scalar value of all first subgraphs based on the third sample probability distribution, and form a first set of the first feature scalar values ​​of all first subgraphs.

[0039] Furthermore, the classical computer 110 can calculate the second feature scalar values ​​of all second subgraphs based on the second sample probability distribution, and form a second set of the second feature scalar values ​​of all second subgraphs.

[0040] Furthermore, the classical computer 110 can determine whether the first set and the second set match based on the first feature scalar value in the first set and the second feature scalar value in the second set. If the first set and the second set match, all nodes corresponding to the target subgraph are confirmed as the first candidate financial fraud gang of the first financial fraud gang.

[0041] In one example, the classical computer 110 can select at least one dense subgraph with a probability higher than a preset threshold from the subgraphs of the first graph structure based on the probability distribution of the first sample, and determine whether the dense subgraph is the target subgraph. If the dense subgraph is not the target subgraph, it confirms that all nodes corresponding to the dense subgraph are the second candidate financial fraud gang of the second financial fraud gang.

[0042] It should be understood that the above application scenario examples are only shown to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited thereto. Rather, the embodiments of this application can be applied to any applicable scenario.

[0043] Exemplary methods

[0044] Figure 2The diagram shown is a flowchart illustrating a method for identifying financial fraud gangs provided in an exemplary embodiment of this application. Figure 2 As shown, the method for identifying financial fraud gangs may include the following:

[0045] 210: Obtain the first graph structure corresponding to the target transaction network and the second graph structure corresponding to the financial fraud gang model.

[0046] In one embodiment, the graph structure may include a topological graph consisting of nodes and edges connecting the nodes. For example, the type of graph structure may include directed graphs, undirected graphs, hypergraphs, etc.

[0047] In one embodiment, the target transaction network may include a network of financial transaction data that is needed to identify financial fraud groups.

[0048] In one embodiment, all RMB transfer transaction records occurring within the most recent 24 hours can be extracted from the core transaction system of a commercial bank. These transaction records constitute a target transaction network. Further, a first graph structure can be constructed based on the following method: each financial entity in the transaction records is defined as a node, where a financial entity may include personal bank accounts and / or corporate accounts participating in the transaction. For two financial entities that have had at least one transfer transaction within the 24-hour window, an undirected edge is established between the nodes of the two financial entities. Further still, an adjacency matrix can be generated based on the first graph structure. The values ​​of the matrix elements A_{ij} are determined as follows: if there is an edge connecting node i and node j, then the value of A_{ij} represents the transfer amount; if there is no edge connecting node i and node j, then A_{ij} = 0; simultaneously, the diagonal element A_{ii} is also specified as 0.

[0049] In another embodiment, the target transaction network can be an internal fund transfer network of a multinational corporation. Further, the first graph structure can be constructed based on the following method: each account in the network is defined as a node, where accounts include operating accounts of subsidiaries in various countries, regional cash pool accounts, and dedicated foreign exchange trading accounts, etc. Directed edges can be established between account nodes with fund transfer relationships based on fund transfer records within a preset period. The direction of the edge represents the direction of fund flow, and the weight of the edge is defined as the net amount of funds transferred along the direction of the edge within the most recent complete financial period. Further, an adjacency matrix can be generated based on the first graph structure. The value of the matrix element A_{ij} of the adjacency matrix follows these rules: if there is a net fund transfer from node i to node j within the period, then the value of A_{ij} is the net amount of funds; if there is no fund transfer, then A_{ij} = 0.

[0050] In one embodiment, a financial fraud gang model may include a network of typical fraudulent transaction behaviors summarized from historical fraud cases.

[0051] In one embodiment, the financial fraud ring model can be a known "cross-border gambling money laundering" network. Further, a second graph structure can be constructed based on the following method: each account involved in the "cross-border gambling money laundering" network is defined as a node. Based on the confirmed fund flows in historical case data, directed edges are established between account nodes with direct transfer relationships. The direction of the directed edge represents the actual flow of funds. The weight of the directed edge can be defined as the estimated frequency and typical amount of funds transferred along that direction within a typical crime cycle. Further, an adjacency matrix can be generated based on the second graph structure. The values ​​of the matrix elements a_{ij} are determined as follows: if there is a fund transfer path from node i to node j, then a_{ij} = 1 (for a binary matrix) or a set weight value; if there is no fund transfer path from node i to node j, then a_{ij} = 0.

[0052] In another embodiment, the financial fraud gang model can be the operational network of a known "credit card cash-out" gang. Further, a second graph structure can be constructed based on the following method: Individual credit card accounts, fake merchant accounts, and core fund collection accounts in the operational network of the "credit card cash-out" gang are all defined as nodes. Based on the cash-out behavior pattern of the "credit card cash-out" gang, directed edges are established from individual credit card account nodes to fake merchant account nodes, and directed edges are established from fake merchant account nodes to core fund collection account nodes. Furthermore, an adjacency matrix can be generated based on the second graph structure, where the values ​​of matrix elements a_{ij} follow the following rules: if there is an edge from node i to node j, then a_{ij} = 1; if there is no edge from node i to node j, then a_{ij} = 0.

[0053] 220: Select at least one target subgraph from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure.

[0054] In one embodiment, the target subgraph may include a subgraph selected from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure. When the two graph structures are isomorphic, the number of nodes in the two graph structures must be the same; therefore, the number of nodes in the target subgraph must be the same as the number of nodes in the second graph structure.

[0055] In one embodiment, a first adjacency matrix can be generated based on a first graph structure. Gaussian boson sampling is then performed on the first adjacency matrix to obtain a first sample probability distribution. A first sample set can be determined based on this first sample probability distribution. Since the first sample set contains information about the number of nodes corresponding to each subgraph of the first graph structure, the number of nodes corresponding to each subgraph of the first graph structure can be determined based on the first sample set. Subgraphs with the same number of nodes as the second graph structure are then selected from the subgraphs of the first graph structure to obtain the target subgraph. The first sample set can include a set of subgraphs of the first graph structure. For example, the value rule for the matrix element A_{ij} of the first adjacency matrix can include: if there is an edge connecting node i and node j, then the value of A_{ij} is 1; if there is no edge connecting node i and node j, then A_{ij} = 0. For example, the first sample probability distribution can be obtained by Gaussian boson sampling of the first adjacency matrix using the following method: decompose the first adjacency matrix to obtain a unitary matrix U and a diagonal matrix S, map the unitary matrix U to the beam splitter and phase shifter parameters of the Gaussian boson sampling device, map the diagonal matrix S to the compression parameter of the Gaussian boson sampling device, inject an initial quantum state into the Gaussian boson sampling device, run the Gaussian boson sampling device, and perform multiple measurements at the output of the Gaussian boson sampling device using a single-photon detector array. Each measurement can obtain a vector sample representing the photon click mode, and the probability of each vector sample is counted to obtain the first sample probability distribution.

[0056] In another embodiment, the selection of the target subgraph can be achieved using classical graph theory algorithms. For example, based on the first graph structure, a subgraph sampling algorithm can be used to efficiently generate candidate subgraphs with the same number of nodes as the second graph structure, which can then be used as the target subgraph. Furthermore, the subgraph sampling algorithm can include random walk-based sampling algorithms, topology-based sampling algorithms, community-based sampling algorithms, etc.

[0057] 230: For each target subgraph in at least one target subgraph, perform the following steps:

[0058] Obtain the first feature scalar values ​​of all first subgraphs of the target subgraph. The first feature scalar values ​​are used to characterize the number of perfect matches or the sum of the weights of perfect matches in the first subgraph, forming the first set.

[0059] Obtain the second feature scalar values ​​of all second subgraphs in the second graph structure. The second feature scalar values ​​are used to characterize the number of perfect matches or the sum of the weights of perfect matches in the second subgraph, forming a second set.

[0060] If, based on the first feature scalar value in the first set and the second feature scalar value in the second set, it is determined that the first set matches the second set, then all nodes corresponding to the target subgraph are confirmed as the first candidate financial fraud gang, and the transaction behavior pattern of the first financial fraud gang matches the financial fraud gang pattern.

[0061] In one embodiment, the first feature scalar value may include the feature scalar value corresponding to the first subgraph, and the second feature scalar value may include the feature scalar value corresponding to the second subgraph.

[0062] In one embodiment, a perfect match can be a set of edges in a graph structure that covers all nodes in the graph structure and where no two edges share a common node. Further, when the graph structure is unweighted, the feature scalar value can characterize the number of perfect matches in the graph structure. Still further, when the graph structure is weighted, the feature scalar value can characterize the sum of the weights of the perfect matches in the graph structure, which can include the sum of the products of the weights of the edges in the perfect match.

[0063] In one embodiment, the characteristic scalar value may include Pfaffian, determinant, Tutte polynomial, etc.

[0064] In one embodiment, the feature scalar value may include a Hafnian value, which can be a mathematical function defined on the adjacency matrix, where the Hafnian is defined as 0 for a 2n × 2n adjacency matrix (if the size of the adjacency matrix is ​​odd). Further, the Hafnian can characterize the number of perfect matches in the graph. The Hafnian of the adjacency matrix is ​​defined as:

[0065]

[0066] Where M represents a perfect matching of the set {1,2,...,2n}, Represents the set of all perfect matches, each product term Matrix elements The product of.

[0067] In one embodiment, the Hafnian values ​​of all first subgraphs of the target subgraph can be calculated using an exact algorithm based on recursive decomposition, an approximate algorithm based on random sampling, or other reasonable algorithms. In another embodiment, the Hafnian values ​​of all second subgraphs of the second graph structure can be calculated using an exact algorithm based on recursive decomposition, an approximate algorithm based on random sampling, or other reasonable algorithms.

[0068] In one embodiment, matching the first set with the second set may include: the similarity between the first set and the second set being greater than a preset similarity. Further, the preset similarity may include 80%, 90%, or other reasonable values.

[0069] In one embodiment, the similarity between the first set and the second set can be calculated by: identifying the number of target first feature scalar values ​​in the first set, and using the percentage of the number of target first feature scalar values ​​relative to the number of elements in the first set (or the second set) as the similarity, wherein the target first feature scalar values ​​may include first feature scalar values ​​present in the second set. In another embodiment, the similarity between the first set and the second set can be calculated based on other reasonable methods.

[0070] In one embodiment, determining that the first set matches the second set may include: determining that the first set is equal to the second set.

[0071] In one embodiment, the equality of the first set and the second set can be determined by checking whether every first feature scalar value in the first set exists in the second set, and checking whether every second feature scalar value in the second set exists in the first set. If every first feature scalar value in the first set exists in the second set, and every second feature scalar value in the second set exists in the first set, then the first set and the second set are determined to be equal. In another embodiment, the equality of the first set and the second set can be determined by other reasonable methods.

[0072] In one embodiment, confirming that all nodes corresponding to the target subgraph are the first candidate financial fraud gangs may include: confirming that the account holders (such as natural persons or legal persons) corresponding to all nodes corresponding to the target subgraph are the first candidate financial fraud gangs, wherein each node corresponding to the target subgraph represents an account holder in the first candidate financial fraud gang, and all account holders corresponding to all nodes corresponding to the target subgraph constitute the first candidate financial fraud gang.

[0073] In one embodiment, the first candidate financial fraud gang can be a high-probability candidate of the first financial fraud gang. Further, the probability that the first candidate financial fraud gang is the first financial fraud gang is higher than the preset probability. Further still, the preset probability can be determined based on experience. For example, the preset probability can include 90%, 95%, etc.

[0074] This application provides a method for identifying financial fraud gangs. In all subgraphs of a first graph structure corresponding to a target transaction network, a target subgraph with the same number of nodes as a second graph structure corresponding to a financial fraud gang pattern is selected. The first feature scalar values ​​of all first subgraphs of the target subgraph are obtained to form a first set. The second feature scalar values ​​of all second subgraphs of the second graph structure are obtained to form a second set. The first set and the second set are then compared. If they match, all nodes corresponding to the target subgraph are confirmed as first candidate financial fraud gangs. Since when two graph structures are isomorphic, the sets of feature scalar values ​​of all subgraphs corresponding to each graph structure are necessarily the same, this method can accurately match financial fraud gangs whose transaction behavior is completely consistent with known financial fraud gang patterns, significantly improving the accuracy of financial fraud gang identification. Furthermore, the method provided in this application can also improve the efficiency of financial fraud gang identification.

[0075] According to one embodiment of this application, the method for identifying financial fraud gangs further includes: performing Gaussian boson sampling on the first graph structure to obtain a first sample probability distribution corresponding to the subgraphs of the first graph structure; selecting at least one dense subgraph from at least one first subgraph whose occurrence probability is higher than a preset threshold based on the first sample probability distribution; and for each dense subgraph in the at least one dense subgraph, if the dense subgraph is not the target subgraph, confirming all nodes corresponding to the dense subgraph as a second candidate financial fraud gang, and the transaction behavior pattern of the second financial fraud gang is inconsistent with the financial fraud gang pattern.

[0076] The method described in the above embodiments of this application can be referred to to perform Gaussian boson sampling on the first graph structure to obtain the sample probability distribution corresponding to the subgraph of the first graph structure, which will not be elaborated here.

[0077] In one embodiment, since the sample probability distribution contains the occurrence probability information of each subgraph, subgraphs with occurrence probabilities higher than a preset threshold can be found in the sample probability distribution and designated as dense subgraphs.

[0078] In one embodiment, the preset threshold can be determined based on actual needs. For example, the preset threshold could be 95%. As another example, the preset threshold could be 90%.

[0079] In one embodiment, the second candidate financial fraud gang can be a high-probability candidate of the second financial fraud gang. Further, the probability that the second candidate financial fraud gang is the second financial fraud gang is higher than the preset probability. Further still, the preset probability can be determined based on experience. For example, the preset probability can include 90%, 95%, etc.

[0080] In one embodiment, the second candidate financial fraud group can be further analyzed to confirm whether it is indeed a second financial fraud group. For example, the internal correlation within the second candidate financial fraud group can be analyzed. If there are abnormally high closed-loop fund transactions among nodes, or if they share high-risk equipment or addresses, it can be confirmed whether the second candidate financial fraud group is indeed a second financial fraud group. Another example is analyzing the historical risk scores of members within the group. If the historical risk scores of members within the group are higher than a preset score, it can be confirmed whether the second candidate financial fraud group is indeed a second financial fraud group.

[0081] In one example, the first financial fraud gang has a higher risk level than the second financial fraud gang.

[0082] In this embodiment, the sample probability distribution corresponding to the subgraph of the first graph structure is obtained by Gaussian boson sampling, and dense subgraphs with a probability of occurrence higher than a preset threshold are selected from the subgraphs of the first graph structure. When the dense subgraph is not the target subgraph, all nodes corresponding to the dense subgraph are marked as the second candidate financial fraud gangs of the second financial fraud gang whose transaction behavior pattern does not match the financial fraud gang pattern. In this way, financial fraud gangs with dense structure but not in line with the known financial fraud gang pattern can be captured, thereby avoiding the omission of identification of possible risky financial fraud gangs.

[0083] According to one embodiment of this application, obtaining the first feature scalar value of all first subgraphs of the target subgraph includes: performing Gaussian boson sampling on the structure of the first graph using an optical quantum computer to obtain the first sample probability distribution corresponding to all subgraphs of the first graph structure; and calculating the first feature scalar value of all first subgraphs based on the first sample probability distribution.

[0084] In one embodiment, the probability of the first subgraph appearing in the probability distribution of the first sample is proportional to the square of the feature scalar value of the first subgraph.

[0085] In one embodiment, the first subgraph includes m nodes. The first feature scalar value of all first subgraphs can be calculated based on the following method: From the first sample probability distribution, select the third sample probability distribution of the target subgraph, and calculate the first feature scalar value of all first subgraphs based on the third sample probability distribution of the target subgraph. For example, the first feature scalar value of the first subgraph can be calculated based on the following formula:

[0086]

[0087] in, This represents the first feature scalar value corresponding to the first subgraph. This represents the probability of the first subgraph appearing in the probability distribution of the first sample. is the probability of vacuum sampling, c is a constant related to the compressibility parameter, and k is the total number of photons detected. This is the number of photons detected in the first output mode during Gaussian boson sampling. It represents the number of photons detected in the m-th output mode of Gaussian boson sampling. Each node in the first subgraph corresponds one-to-one with each output mode in Gaussian boson sampling.

[0088] In this embodiment, the probability distribution of samples is obtained through Gaussian boson sampling, and the first feature scalar value of the first sub-graph is calculated. Utilizing the global sampling capability of the optical quantum computer, the problem of getting trapped in local optima is avoided. Simultaneously, through precise comparison of feature scalar values, a rigorous judgment criterion is provided for identifying financial fraud gangs, thereby significantly improving the accuracy of financial fraud gang identification. Furthermore, the method provided in this application embodiment can also improve the efficiency of identifying financial fraud gangs.

[0089] According to one embodiment of this application, obtaining the second feature scalar value of all second subgraphs of the second graph structure includes: performing Gaussian boson sampling on the second graph structure using an optical quantum computer to obtain the second sample probability distribution corresponding to all second subgraphs; and calculating the second feature scalar value of all second subgraphs based on the second sample probability distribution.

[0090] The method described in the above embodiments of this application can be referred to to perform Gaussian boson sampling on the second graph structure to obtain the second sample probability distribution corresponding to the second subgraph, which will not be elaborated here.

[0091] The method described in the above embodiments of this application can be referred to to calculate the second feature scalar value of all second subgraphs based on the second sample probability distribution, which will not be elaborated here.

[0092] In this embodiment, the sample probability distribution is obtained by Gaussian boson sampling and the second feature scalar value of the second subgraph is calculated. By utilizing the global sampling capability of the optical quantum computer, it avoids getting trapped in local optima. At the same time, through precise comparison of the second feature scalar value, a strict judgment basis is provided for the identification of financial fraud gangs, thereby significantly improving the accuracy of identifying financial fraud gangs.

[0093] According to one embodiment of this application, the financial fraud gang identification method further includes: if the first candidate financial fraud gang is the first financial fraud gang, decoding all nodes corresponding to the first financial fraud gang to obtain the financial entity corresponding to each node; inputting the financial entity into the financial risk control system to trigger risk disposal operations on the financial entity.

[0094] In one embodiment, further analysis can be conducted on the first candidate financial fraud group to confirm whether it is indeed the first financial fraud group. For example, the internal correlation within the first candidate financial fraud group can be analyzed. If there are abnormally high closed-loop fund transactions among nodes, or if they share high-risk devices or addresses, it can be confirmed whether the first candidate financial fraud group is indeed the first financial fraud group. Another example is analyzing the historical risk scores of members within the group. If the historical risk scores of members within the group are higher than a preset score, it can be confirmed whether the first candidate financial fraud group is indeed the first financial fraud group.

[0095] In one embodiment, all nodes corresponding to the first financial fraud gang can be decoded to obtain the financial entity corresponding to each node based on the following method: Based on the mapping relationship between nodes and financial entities, each node in the target subgraph is converted into information of the corresponding financial entity. Further, the financial entity may include personal bank accounts, corporate accounts, credit card accounts, etc. Further still, the information of the financial entity may include account number, account holder name, account opening institution code, etc.

[0096] In one embodiment, risk management operations for a financial entity may include: intercepting abnormal transactions in an account, reducing the account's transaction limits, and freezing the account's fund transfer function.

[0097] In this embodiment, by decoding all nodes corresponding to the first financial fraud gang, a mapping relationship between nodes and financial entities is established to trigger risk disposal measures against the financial entities, effectively blocking the continuation of fraudulent transactions and improving the accuracy and efficiency of financial risk management.

[0098] According to one embodiment of this application, selecting at least one subgraph with the same number of nodes as the second graph structure from the subgraphs of the first graph structure as a target subgraph includes: selecting at least one target subgraph with the same number of nodes as the second graph structure from the subgraphs of the first graph structure and satisfying a preset condition, wherein the vector sample corresponding to each target subgraph satisfying the preset condition contains only a first value and a second value, and the vector sample is used to characterize the photon click pattern corresponding to the target subgraph.

[0099] In one embodiment, the preset conditions that the target sub-image needs to meet may include: the photon click pattern corresponding to the target sub-image includes vector samples containing only the first and second values.

[0100] In one embodiment, the first and second values ​​can be set based on actual needs. For example, the first value can be 0 and the second value can be 1. As another example, the first value can be 2 and the second value can be 3.

[0101] In one embodiment, the first value is 0 and the second value is 1. When selecting a target subgraph from the subgraph of the first graph structure, the node corresponding to the position with a value of 1 in the vector sample is the node of the target subgraph. For example, when the vector sample of the first graph structure is (1,1,1,1,0,0), the target subgraph corresponding to the first graph structure is a subgraph composed of four nodes in the first graph structure, and the corresponding value of these four nodes in the vector sample of the first graph structure is 1.

[0102] In this embodiment, the photon click pattern corresponding to the target sub-image includes vector samples containing only the first and second values. This preset condition filters out the interference of multi-value or abnormal samples, reduces the risk of misjudgment caused by noise, and significantly improves the identification accuracy of financial fraud gangs.

[0103] According to one embodiment of this application, the first graph structure includes an undirected graph, and the second graph structure includes an undirected graph.

[0104] In one embodiment, an undirected graph may include a graph structure consisting of nodes and edges, where nodes represent financial entities, edges represent transaction relationships between financial entities, and edges are not directional.

[0105] In this embodiment, both the first graph structure and the second graph structure include undirected graphs. By utilizing the stability of the topological structure of undirected graphs, the essential characteristics of the transaction relationship between financial entities can be effectively represented, simplifying the computational complexity. At the same time, it avoids the directional misjudgment that may be caused by directed graphs, thereby improving the accuracy of identifying financial fraud gangs.

[0106] Figure 3 The diagram shown is a flowchart illustrating a method for identifying financial fraud gangs provided in another exemplary embodiment of this application. Figure 3 The example is Figure 2 Examples of the embodiments are provided below; to avoid repetition, the similarities can be referred to the descriptions in the above embodiments, and will not be repeated here. For example... Figure 3 As shown, the method for identifying financial fraud gangs may include the following.

[0107] 310: Obtain the first graph structure corresponding to the target transaction network and the second graph structure corresponding to the financial fraud gang model.

[0108] Specifically, the relevant content regarding obtaining the first graph structure corresponding to the target transaction network and the second graph structure corresponding to the financial fraud gang model can be found in the relevant descriptions in the above embodiments. To avoid repetition, they will not be repeated here.

[0109] 320: Generate the first adjacency matrix based on the first graph structure, and generate the second adjacency matrix based on the second graph structure.

[0110] Specifically, the details of generating the first adjacency matrix based on the first graph structure and generating the second adjacency matrix based on the second graph structure can be found in the relevant descriptions in the above embodiments. To avoid repetition, they will not be repeated here.

[0111] 330: Based on the first adjacency matrix, Gaussian boson sampling is performed on the first graph structure to obtain the first sample probability distribution corresponding to all subgraphs of the first graph structure. Based on the second adjacency matrix, Gaussian boson sampling is performed on the second graph structure to obtain the second sample probability distribution corresponding to all second subgraphs of the second graph structure.

[0112] Specifically, the relevant content regarding Gaussian boson sampling of the first graph structure based on the first adjacency matrix to obtain the first sample probability distribution corresponding to all subgraphs of the first graph structure, and Gaussian boson sampling of the second graph structure to obtain the second sample probability distribution corresponding to all second subgraphs of the second graph structure, can be found in the relevant descriptions in the above embodiments. To avoid repetition, it will not be repeated here.

[0113] 340: Based on the probability distribution of the first sample, determine the number of nodes corresponding to each subgraph of the first graph structure, and select the subgraphs with the same number of nodes as the second graph structure from the subgraphs of the first graph structure to obtain the target subgraph.

[0114] Specifically, the number of nodes corresponding to each subgraph of the first graph structure is determined based on the probability distribution of the first sample, and the subgraphs with the same number of nodes as the second graph structure are selected from the subgraphs of the first graph structure to obtain the target subgraph. For the relevant content of obtaining the target subgraph, please refer to the relevant description in the above embodiments. To avoid repetition, it will not be repeated here.

[0115] 350: In the first sample probability distribution, select the third sample probability distribution corresponding to all first subgraphs of the target subgraph, calculate the first feature scalar value of all first subgraphs based on the third sample probability distribution, and form the first feature scalar value of all first subgraphs into a first set.

[0116] Specifically, in the first sample probability distribution, the third sample probability distribution corresponding to all first subgraphs of the target subgraph is selected, and the first feature scalar value of all first subgraphs is calculated based on the third sample probability distribution. The relevant content of forming a first set of the first feature scalar values ​​of all first subgraphs can be found in the relevant description in the above embodiments. To avoid repetition, it will not be repeated here.

[0117] 360: Based on the probability distribution of the second sample, calculate the second feature scalar value of all second subgraphs, and form a second set of the second feature scalar values ​​of all second subgraphs.

[0118] Specifically, the calculation of the second feature scalar values ​​of all second subgraphs and the formation of a second set of the second feature scalar values ​​of all second subgraphs can be found in the relevant descriptions in the above embodiments. To avoid repetition, they will not be repeated here.

[0119] 370: Based on the first feature scalar value in the first set and the second feature scalar value in the second set, determine whether the first set and the second set match. If the first set and the second set match, confirm that all nodes corresponding to the target subgraph are the first candidate financial fraud gangs of the first financial fraud gang.

[0120] Specifically, the determination of whether the first set and the second set match based on the first feature scalar value in the first set and the second feature scalar value in the second set, and the confirmation that all nodes corresponding to the target subgraph are the first candidate financial fraud gangs of the first financial fraud gang, can be found in the relevant descriptions in the above embodiments. To avoid repetition, they will not be repeated here.

[0121] 380: Based on the probability distribution of the first sample, select at least one dense subgraph from the subgraphs of the first graph structure whose occurrence probability is higher than a preset threshold, and for each dense subgraph in the at least one dense subgraph, determine whether the dense subgraph is the target subgraph. If the dense subgraph is not the target subgraph, confirm that all nodes corresponding to the dense subgraph are the second candidate financial fraud gang of the second financial fraud gang.

[0122] Specifically, from the subgraphs of the first graph structure, at least one dense subgraph with a probability of occurrence higher than a preset threshold is selected, and it is determined whether the dense subgraph is the target subgraph. If the dense subgraph is not the target subgraph, the relevant content of confirming that all nodes corresponding to the dense subgraph are the second candidate financial fraud gangs of the second financial fraud gang can be found in the relevant descriptions in the above embodiments. To avoid repetition, it will not be repeated here.

[0123] 390: If the first candidate financial fraud gang is the first financial fraud gang, decode all the first nodes corresponding to the first financial fraud gang to obtain the first financial entity corresponding to each first node. If the second candidate financial fraud gang is the second financial fraud gang, decode all the second nodes corresponding to the second financial fraud gang to obtain the second financial entity corresponding to each second node. Input the first financial entity and the second financial entity into the financial risk control system to trigger risk disposal operations for the first financial entity and the second financial entity.

[0124] Specifically, the first node can be the node corresponding to the first financial fraud gang, and the second node can be the node corresponding to the second financial fraud gang.

[0125] Specifically, the decoding of all first nodes corresponding to the first financial fraud gang yields the first financial entity corresponding to each first node, the decoding of all second nodes corresponding to the second financial fraud gang yields the second financial entity corresponding to each second node, and the input of the first and second financial entities into the financial risk control system to trigger risk disposal operations for the first and second financial entities can be found in the relevant descriptions in the above embodiments. To avoid repetition, they will not be repeated here.

[0126] Exemplary device

[0127] Figure 4 The diagram shown is a schematic representation of a financial fraud gang identification device provided in an exemplary embodiment of this application. This financial fraud gang identification device can be applied to electronic devices. Figure 4 As shown, the financial fraud gang identification device 400 includes: an acquisition module 410, a selection module 420, and an execution module 430.

[0128] The acquisition module 410 is used to acquire the first graph structure corresponding to the target transaction network and the second graph structure corresponding to the financial fraud gang model. The selection module 420 is used to select at least one target subgraph from the subgraphs of the first graph structure that has the same number of nodes as the second graph structure. The execution module 430 is used to perform the following steps for each target subgraph in the at least one target subgraph: acquire the first feature scalar values ​​of all first subgraphs of the target subgraph to form a first set; acquire the second feature scalar values ​​of all second subgraphs of the second graph structure to form a second set; if the first set and the second set are determined to match based on the first feature scalar values ​​in the first set and the second feature scalar values ​​in the second set, confirm that all nodes corresponding to the target subgraph are the first candidate financial fraud gangs of the first financial fraud gang, and the transaction behavior pattern of the first financial fraud gang matches the financial fraud gang model.

[0129] This application provides a device for identifying financial fraud gangs. In all subgraphs of a first graph structure corresponding to a target transaction network, a target subgraph with the same number of nodes as a second graph structure corresponding to a financial fraud gang pattern is selected. The first feature scalar values ​​of all first subgraphs of the target subgraph are obtained to form a first set. The second feature scalar values ​​of all second subgraphs of the second graph structure are obtained to form a second set. The device then determines whether the first set and the second set match. If they match, all nodes corresponding to the target subgraph are considered first candidate financial fraud gangs. Since when two graph structures are isomorphic, the sets of feature scalar values ​​of all subgraphs corresponding to each graph structure must be the same, this method can accurately match financial fraud gangs whose transaction behavior is completely consistent with known financial fraud gang patterns, significantly improving the accuracy of identifying financial fraud gangs.

[0130] According to one embodiment of this application, the execution module 430 is used to perform Gaussian boson sampling on the first graph structure using an optical quantum computer to obtain the first sample probability distribution corresponding to all subgraphs of the first graph structure; and to calculate the first feature scalar value of all first subgraphs based on the first sample probability distribution.

[0131] According to one embodiment of this application, the execution module 430 is used to perform Gaussian boson sampling on the second graph structure using an optical quantum computer to obtain the second sample probability distribution corresponding to all second subgraphs; and to calculate the second feature scalar value of all second subgraphs based on the second sample probability distribution.

[0132] According to one embodiment of this application, the financial fraud gang identification device 400 further includes a decoding module 440, which is used to decode all nodes corresponding to the first financial fraud gang to obtain the financial entity corresponding to each node, and input the financial entity into the financial risk control system to trigger risk disposal operations on the financial entity.

[0133] According to an embodiment of this application, the selection module 420 is used to select at least one target subgraph from the subgraph of the first graph structure that has the same number of nodes as the second graph structure and satisfies a preset condition. The vector sample corresponding to each target subgraph that satisfies the preset condition contains only a first value and a second value. The vector sample is used to characterize the photon click pattern corresponding to the target subgraph.

[0134] According to one embodiment of this application, the first graph structure includes an undirected graph, and the second graph structure includes an undirected graph.

[0135] According to one embodiment of this application, the execution module 430 is further configured to perform Gaussian boson sampling on the first graph structure to obtain a first sample probability distribution corresponding to the subgraphs of the first graph structure; based on the first sample probability distribution, select at least one dense subgraph from the subgraphs of the first graph structure whose occurrence probability is higher than a preset threshold; if the dense subgraph is not the target subgraph, confirm that all nodes corresponding to the dense subgraph are the second candidate financial fraud gang of the second financial fraud gang, and the transaction behavior pattern of the second financial fraud gang is inconsistent with the financial fraud gang pattern.

[0136] It should be understood that the operations and functions of the acquisition module 410, selection module 420, execution module 430, and decoding module 440 in the above embodiments can be referred to the above. Figure 2 or Figure 3 The description of the financial fraud gang identification method provided in the embodiments will not be repeated here to avoid repetition.

[0137] Figure 5 The diagram shown is a block diagram of an optical quantum computer 500 for performing a method for identifying financial fraud gangs, provided in an exemplary embodiment of this application.

[0138] Reference Figure 5 The optical quantum computer 500 includes a quantum light source 510, an optical quantum chip 520, and a single-photon detector 530. The quantum light source 510 generates high-quality photons as qubit carriers by exciting quantum dots with lasers or through spontaneous parametric down-conversion (SPDC). The optical quantum chip 520 is composed of optical components such as optical fibers, waveguides, beam splitters, phase modulators, and mirrors to achieve optical transmission and logic operations (such as Hadamard gates and CNOT gates). The single-photon detector 530 can measure the final state of photons (such as polarization or path) and output the calculation results. The quantum light source 510, optical quantum chip 520, and single-photon detector 530 are configured to execute instructions to perform the aforementioned financial fraud gang identification method.

[0139] Figure 6 The diagram shown is a block diagram of an electronic device 600 for performing a financial fraud gang identification method, provided in an exemplary embodiment of this application. Specifically, the electronic device 600 may be a server, a mobile device, a control device for the mobile device, a server interacting with the mobile device, a controller, or other devices.

[0140] Reference Figure 6 The electronic device 600 includes a processing component 610, which further includes one or more processors, and memory resources represented by memory 620 for storing instructions executable by the processing component 610, such as application programs. The application programs stored in memory 620 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 610 is configured to execute instructions to perform the aforementioned financial fraud gang identification method.

[0141] Electronic device 600 may also include a power supply component configured to perform power management of electronic device 600, a wired or wireless network interface configured to connect electronic device 600 to a network, and an input / output (I / O) interface. Electronic device 600 can be operated based on an operating system stored in memory 620, such as Windows Server. TM Mac OSX TM Unix TM Linux TM FreeBSD TM Or similar.

[0142] A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by the processor of the aforementioned electronic device 600, enables the electronic device 600 to perform a method for identifying financial fraud gangs.

[0143] A computer program product includes a computer program that, when executed by a processor of a computer device, enables the computer device to perform the financial fraud gang identification method provided in any of the above embodiments.

[0144] All of the above-mentioned optional technical solutions can be combined in any way to form optional embodiments of this application, and will not be described in detail here.

[0145] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0146] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0147] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0148] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0149] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0150] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program verification codes, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0151] It should be noted that in the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0152] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0153] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications or equivalent substitutions made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method of financial fraud ring identification, the method comprising: The method comprises: obtaining a first graph structure corresponding to a target transaction network, and obtaining a second graph structure corresponding to a financial fraud gang mode; from the subgraphs of the first graph structure, selecting at least one target subgraph having the same number of nodes as the second graph structure; for each target subgraph in the at least one target subgraph, the following steps are performed: obtaining first characteristic scalar values of all first subgraphs of the target subgraph, the first characteristic scalar values being used to represent the number of perfect matches or the weight sum of perfect matches of the first subgraph, constituting a first set; obtaining second characteristic scalar values of all second subgraphs of the second graph structure, the second characteristic scalar values being used to represent the number of perfect matches or the weight sum of perfect matches of the second subgraph, constituting a second set; if it is determined that the first set and the second set are consistent based on the first characteristic scalar values in the first set and based on the second characteristic scalar values in the second set, confirming that all nodes corresponding to the target subgraph are a first candidate financial fraud gang of a first financial fraud gang, the first financial fraud gang being consistent with a transaction behavior mode of the financial fraud gang mode, wherein the determination that the first set and the second set are consistent comprises determining that the similarity of the first set and the second set is greater than a preset similarity.

2. The method of claim 1, wherein, The method further comprises: performing Gaussian Bose sampling on the first graph structure to obtain a first sample probability distribution corresponding to the subgraphs of the first graph structure; from the at least one first subgraph of the target subgraph, selecting at least one dense subgraph having an occurrence probability higher than a preset threshold according to the first sample probability distribution; for each dense subgraph in the at least one dense subgraph, if the dense subgraph is not a target subgraph, confirming that all nodes corresponding to the dense subgraph are a second candidate financial fraud gang of a second financial fraud gang, the second financial fraud gang being inconsistent with the transaction behavior mode of the financial fraud gang mode.

3. The method of claim 1, wherein, The obtaining of the first characteristic scalar values of all first subgraphs of the target subgraph comprises: performing Gaussian Bose sampling on the first graph structure by using a photonic quantum computer to obtain a first sample probability distribution corresponding to the subgraphs of the first graph structure; based on the first sample probability distribution, calculating the first characteristic scalar values of the all first subgraphs.

4. The method of claim 1, wherein, The obtaining of the second characteristic scalar values of all second subgraphs of the second graph structure comprises: performing Gaussian Bose sampling on the second graph structure by using a photonic quantum computer to obtain a second sample probability distribution corresponding to the all second subgraphs; based on the second sample probability distribution, calculating the second characteristic scalar values of the all second subgraphs.

5. The method of claim 1, wherein, The method further comprises: if the first candidate financial fraud gang is the first financial fraud gang, decoding all nodes corresponding to the first financial fraud gang to obtain a financial entity corresponding to each node; inputting the financial entity into a financial risk control system to trigger a risk disposal operation on the financial entity.

6. The method of claim 1, wherein, The selecting of at least one target subgraph having the same number of nodes as the second graph structure from the subgraphs of the first graph structure comprises: From the sub-graphs of the first graph structure, at least one target sub-graph with the same number of nodes as the second graph structure and satisfying a preset condition is selected, each target sub-graph satisfying the preset condition corresponds to a vector sample containing only a first value and a second value, and the vector sample is used to represent a photon click pattern corresponding to the target sub-graph.

7. The method according to any one of claims 1 to 6, characterized in that, The first graph structure includes an undirected graph, and the second graph structure includes an undirected graph.

8. A photonic quantum computer, characterized by The optical quantum computer is used to execute the financial fraud ring identification method in any one of claims 1 to 7, and the optical quantum computer includes a quantum light source, an optical quantum chip, and a single-photon detector.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and the computer program is used to execute the financial fraud ring identification method in any one of claims 1 to 7.

10. A computer program product, characterised in that, The computer program product includes a computer program, and the computer program is executed by a processor of a computer device, so that the computer device can execute the financial fraud ring identification method in any one of claims 1 to 7.

Images