A knowledge graph maintenance method, device, equipment and storage medium

Abstract

The present application discloses a knowledge graph maintenance method, device, equipment and storage medium, which relates to the field of credit risk control technology, including: obtaining customer operation information on several entities; the several entities are entities related to credit risk control business; using the several entities and the corresponding operation information to update the initial knowledge graph to obtain a target knowledge graph; based on the principle of web page ranking algorithm, the target knowledge graph is quantitatively calculated to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph. In this way, the present application can quickly handle the centrality problem of dynamically changing knowledge graphs, update the relevant knowledge graph according to the customer's operation information on the relevant entities, and with the ability of the web page ranking algorithm to process dynamic graphs, the updated target knowledge graph is quantitatively calculated, and the quantitative results are obtained by standardized processing to complete the maintenance process of the knowledge graph, thereby improving efficiency.

Description

Technical Field

[0001] The present invention relates to the field of credit risk control technology, and in particular to a knowledge graph maintenance method, device, equipment and storage medium. Background Art

[0002] In the field of credit, knowledge graphs are often touted by various institutions as an advanced big data application technology. In the credit business, there are often cases of abnormal applications through intermediaries. In a comprehensive knowledge graph, intermediaries are likely to appear in the core associated position of an abnormal credit application. Therefore, the centrality algorithm can be used to quickly find all core nodes in batches in order to discover intermediaries, such as using the PageRank algorithm. In the scenario of processing dynamically changing knowledge graphs and requiring real-time and rapid feedback of results, the existing PageRank method requires the entire graph to be recalculated each time, which is time-inefficient.

[0003] Therefore, how to efficiently and quickly provide feedback on the dynamic changes of knowledge graphs is a problem that needs to be solved in this field. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a knowledge graph maintenance method, device, equipment and storage medium, which can first quickly process the centrality problem of the dynamically changing knowledge graph, and then use the web page ranking algorithm to perform quantitative calculations on the updated target knowledge graph, which can improve the calculation efficiency. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a knowledge graph maintenance method, comprising:

[0006] Obtaining the customer's operation information on several entities; the several entities are entities related to the credit risk control business;

[0007] Using the entities and the corresponding operation information, the initial knowledge graph is updated to obtain a target knowledge graph;

[0008] Based on the principle of web page ranking algorithm, the target knowledge graph is quantitatively calculated to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph.

[0009] Optionally, the updating the initial knowledge graph using the plurality of entities and the corresponding operation information to obtain a target knowledge graph includes:

[0010] Instantiating nodes of the entities based on a first preset function, so as to add new nodes corresponding to the entities to the initial knowledge graph, thereby obtaining a first knowledge graph;

[0011] Based on a second preset function and using the operation information, disconnecting the relationship between related nodes in the first knowledge graph to obtain a second knowledge graph;

[0012] Based on the third preset function and using the operation information, a relationship between related nodes is established in the second knowledge graph to obtain a target knowledge graph.

[0013] Optionally, the instantiating nodes of the entities based on the first preset function to add new nodes corresponding to the entities to the initial knowledge graph to obtain a first knowledge graph includes:

[0014] According to the asynchronous processing principle and based on the first preset function, a node instantiation operation is performed on the entities to obtain new nodes corresponding to the entities;

[0015] Add the new node to the initial knowledge graph to obtain a first knowledge graph.

[0016] Optionally, based on the second preset function, disconnecting the relationship between related nodes in the first knowledge graph using the operation information to obtain the second knowledge graph includes:

[0017] Determine a current node from the first knowledge graph according to the operation information;

[0018] Based on a second preset function, and using the node relationship information corresponding to the current node in the operation information, the relationship between the current node and other nodes is disconnected to obtain a second knowledge graph.

[0019] Optionally, the establishing a relationship between related nodes in the second knowledge graph based on the third preset function and using the operation information to obtain a target knowledge graph includes:

[0020] Determine a plurality of groups of node relationships to be established according to the node relationship information of the new node corresponding to the operation information;

[0021] Based on the third preset function and using the node relationships to be established, corresponding node relationships are established in the second knowledge graph in sequence to obtain the target knowledge graph.

[0022] Optionally, the target knowledge graph is quantitatively calculated based on the principle of the web page ranking algorithm to obtain corresponding quantitative results, including:

[0023] According to the asynchronous processing principle, the node information in the target knowledge graph is copied to obtain the corresponding copied knowledge graph;

[0024] Based on the principle of web page ranking algorithm, several nodes in the copied knowledge graph are calculated to obtain a quantitative result for the copied knowledge graph; the quantitative result has a one-to-one correspondence with each node in the copied knowledge graph.

[0025] Optionally, the process of completing the maintenance of the relevant knowledge graph includes:

[0026] Determining whether the quantization result is not greater than a preset result threshold;

[0027] If so, the node in the target knowledge graph corresponding to the quantization result greater than the preset result threshold is marked as a node to be reviewed so that the node to be reviewed can be manually reviewed; wherein a single node corresponds to a single entity.

[0028] In a second aspect, the present application provides a knowledge graph maintenance device, comprising:

[0029] An information acquisition module, used to acquire the customer's operation information on several entities; the several entities are entities related to the credit risk control business;

[0030] A knowledge graph updating module, used to update the initial knowledge graph using the entities and the corresponding operation information to obtain a target knowledge graph;

[0031] The knowledge graph quantification module is used to perform quantitative calculations on the target knowledge graph based on the principle of the web page ranking algorithm to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph.

[0032] In a third aspect, the present application provides an electronic device, including:

[0033] Memory, used to store computer programs;

[0034] A processor is used to execute the computer program to implement the knowledge graph maintenance method as described above.

[0035] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the knowledge graph maintenance method as described above.

[0036] It can be seen that the present application first obtains the customer's operation information on several entities; the several entities are entities related to the credit risk control business; then the initial knowledge graph is updated using the several entities and the corresponding operation information to obtain the target knowledge graph; then the target knowledge graph is quantitatively calculated based on the principle of the web page ranking algorithm to obtain the corresponding quantitative results to complete the maintenance process of the relevant knowledge graph. In this way, the present application can quickly handle the centrality problem of the dynamically changing knowledge graph, update the relevant knowledge graph according to the customer's operation information on the relevant entities, and with the help of the web page ranking algorithm's ability to process dynamic graphs, the updated target knowledge graph is quantitatively calculated, and the quantitative results are obtained through standardized processing to complete the maintenance process of the knowledge graph, thereby improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0038] Figure 1 A flow chart of a knowledge graph maintenance method disclosed in this application;

[0039] Figure 2 A schematic diagram of an initial knowledge graph disclosed in this application;

[0040] Figure 3 A schematic diagram of a target knowledge graph disclosed in this application;

[0041] Figure 4 A flowchart of a specific knowledge graph maintenance method disclosed in this application;

[0042] Figure 5 A schematic diagram of the structure of a knowledge graph maintenance device disclosed in this application;

[0043] Figure 6 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 As shown, the embodiment of the present invention discloses a knowledge graph maintenance method, including:

[0046] Step S11, obtaining the customer's operation information on several entities; the several entities are entities related to the credit risk control business.

[0047] In this application, we first obtain the customer's operation information on several entities, among which the operation information on related entities can be screened out through the customer's real-time application; it should be pointed out that several entities are entities related to credit risk control business; it can be understood that through the connection between customer-related entities, it can be seen whether there are abnormal application behaviors for credit between different customers and different entities.

[0048] Step S12: Use the entities and the corresponding operation information to update the initial knowledge graph to obtain the target knowledge graph.

[0049] Furthermore, the preset initial knowledge graph can be updated through the customer's operation information on several entities. It is understandable that the present application can set up a knowledge graph related to the previous credit business in advance, so as to use the knowledge graph to judge the relevant information of subsequent customers; in other words, the preset initial knowledge graph can be updated using the current customer's operation information, so that the updated target knowledge graph can be obtained, so as to use the target knowledge graph to judge whether the current customer has abnormal behavior in applying for credit business.

[0050] In a specific embodiment, the use of the several entities and the corresponding operation information to update the initial knowledge graph to obtain the target knowledge graph may include: based on the first preset function, instantiating the several entities to add the new nodes corresponding to the several entities to the initial knowledge graph to obtain the first knowledge graph; based on the second preset function, disconnecting the relationship between the related nodes in the first knowledge graph using the operation information to obtain the second knowledge graph; based on the third preset function, using the operation information to establish the relationship between the related nodes in the second knowledge graph to obtain the target knowledge graph. Specifically, in the process of updating the initial knowledge graph, firstly, based on the first preset function, instantiating the several entities corresponding to the operation information, and adding these instantiated nodes to the initial knowledge graph, so that the first knowledge graph can be obtained. After that, based on the second preset function, the relationship between the related nodes in the first knowledge graph can be disconnected according to the operation information, that is, the relationship between the related nodes in the first knowledge graph can be disconnected according to the operation information corresponding to the old node; after the relationship between the related nodes in the first knowledge graph is disconnected according to the operation information, the second knowledge graph can be obtained. Correspondingly, there will be relationships that need to be established between the new nodes instantiated previously, and between the new nodes and the old nodes, and even between the old nodes; that is, after obtaining the second knowledge graph, the relationship between the relevant nodes can be established in the second knowledge graph based on the third preset function and using the operation information, so that the target knowledge graph can be obtained.

[0051] In a specific embodiment, the node instantiation of the several entities based on the first preset function to add the new nodes corresponding to the several entities to the initial knowledge graph to obtain the first knowledge graph may include: performing node instantiation operations on the several entities according to the principle of asynchronous processing and based on the first preset function to obtain new nodes corresponding to the several entities; adding the new nodes to the initial knowledge graph to obtain the first knowledge graph. Specifically, in the process of node instantiation of the several entities based on the first preset function, firstly, according to the principle of asynchronous processing and based on the first preset function, the node instantiation operation can be performed on the several entities, so that the instantiated new nodes corresponding to the several entities (the entities corresponding to the operation information) can be obtained. Furthermore, the instantiated new nodes are added to the preset initial knowledge graph, so that the first knowledge graph after adding the new nodes can be obtained.

[0052] In another specific embodiment, the method of disconnecting the relationship between related nodes in the first knowledge graph based on the second preset function and using the operation information to obtain the second knowledge graph may include: determining the current node from the first knowledge graph according to the operation information; disconnecting the relationship between the current node and other nodes based on the second preset function and using the node relationship information corresponding to the current node in the operation information to obtain the second knowledge graph. Specifically, in the process of disconnecting the relationship between nodes in the knowledge graph, first determine a number of old nodes that need to be processed from the first knowledge graph according to the operation information, and disconnect the relationship between these nodes in turn; further, first determine the current node that needs to be processed, and determine the relationship with another node that needs to be disconnected, and then disconnect the relationship between the current node and the other node based on the second preset function. It can be understood that in this way, the relationship between related nodes in the first knowledge graph can be disconnected according to the operation information, and when all the relationships between the nodes that need to be processed are disconnected, the second knowledge graph can be obtained.

[0053] In another specific embodiment, the process of establishing the relationship between related nodes in the second knowledge graph based on the third preset function and using the operation information to obtain the target knowledge graph may include: determining several groups of node relationships to be established based on the node relationship information of the new node corresponding to the operation information; establishing corresponding node relationships in the second knowledge graph in sequence based on the third preset function and using the node relationships to be established to obtain the target knowledge graph. Specifically, in the process of establishing node relationships, several groups of node relationships to be established are first determined based on the node relationship information of the new node corresponding to the operation information; it can be understood that the node relationship information of the new node includes new nodes and new nodes, new nodes and old nodes; in special cases, the operation information may include node relationships that need to be established between old nodes. Afterwards, corresponding node relationships can be established in the second knowledge graph in sequence based on the third preset function and using several node relationships to be established.

[0054] In a specific embodiment, Figure 2 The figure shows an initial knowledge graph, including various information of the customer. When the customer's application information 3 is obtained, the initial knowledge graph needs to be updated, such as Figure 3 Shown is the updated target knowledge graph.

[0055] Step S13: perform quantitative calculations on the target knowledge graph based on the principle of the web page ranking algorithm to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph.

[0056] In the present application, the operation information can be used to update the initial knowledge graph, add nodes, disconnect old node relationships, and establish new node relationships through the aforementioned steps, so that the processed target knowledge graph can be obtained. After that, the traffic of each node in the target knowledge graph can be quantitatively calculated based on the web page ranking algorithm, and the corresponding quantitative results can be obtained. It can be understood that the relevant nodes in the target knowledge graph can be maintained according to the quantitative results to complete the maintenance process of the relevant knowledge graph. In a specific embodiment, the target knowledge graph is quantitatively calculated based on the principle of the web page ranking algorithm to obtain the corresponding quantitative results, which may include: according to the principle of asynchronous processing, the node information in the target knowledge graph is copied to obtain the corresponding copied knowledge graph; based on the principle of the web page ranking algorithm, several nodes in the copied knowledge graph are calculated to obtain the quantitative results for the copied knowledge graph; the quantitative results have a one-to-one correspondence with each node in the copied knowledge graph. Specifically, before performing quantitative calculations on the target knowledge graph, it is necessary to copy the node information of each node in the target knowledge graph to obtain the corresponding copied knowledge graph, and then the quantitative calculations can be performed on several nodes in the copied knowledge graph based on the web page ranking algorithm, so that the quantitative results for the copied knowledge graph can be obtained. It should be pointed out that the present application first copies the non-standard processed node information corresponding to each node in the target knowledge graph; then uses the web page ranking information to quantify the node information of each node in the copied knowledge graph to obtain the corresponding quantitative results, and each node corresponds to its own quantitative result.

[0057] In a specific embodiment, the process of completing the maintenance of the relevant knowledge graph may include: determining whether the quantification result is not greater than a preset result threshold; if so, marking the node in the target knowledge graph corresponding to the quantification result greater than the preset result threshold as a node to be reviewed, so as to perform manual review on the node to be reviewed; wherein a single node corresponds to a single entity. Specifically, in the process of maintaining the knowledge graph, first determine whether the quantification result of each node is greater than the preset result threshold. If the quantification result corresponding to a single node is greater than the preset result threshold, it means that the node, that is, the single entity, may be the core associated position of the abnormal credit application. At this time, the node can be marked as a node to be reviewed, so as to facilitate manual review of the node to be reviewed.

[0058] It can be seen that this application can quickly handle the centrality problem of dynamically changing knowledge graphs, update relevant knowledge graphs based on customer operation information on related entities, and use the ability of web page ranking algorithms to process dynamic graphs to perform quantitative calculations on the updated target knowledge graphs, obtain quantitative results through standardized processing, complete the maintenance process of the knowledge graph, and improve efficiency. It is understandable that in credit-related businesses, timely and comprehensive discovery of entities with abnormal properties is of great significance for reducing the abnormality rate of the entire credit business and ensuring asset quality.

[0059] See also Figure 4 As shown, an embodiment of the present invention discloses a knowledge graph maintenance method, including:

[0060] In this embodiment, the real-time application of the customer is first obtained, that is, the operation information of the above embodiment; the relevant information can be updated in real time to the preset knowledge graph according to the real-time application of the customer. After that, the relationship between the nodes such as the nodes with association relationships in the knowledge graph (which can be called association classes) and the nodes with more association relationships (which can be called cluster classes) is disconnected or established, so as to use the web page ranking algorithm to calculate the association classes and cluster classes in the knowledge graph in real time; in this way, the quantified variables (quantitative results) of the knowledge graph can be calculated; the results are then fed back to the preset approval risk model or approval strategy, so as to maintain the relevant nodes based on the preset strategy and according to the quantified results of each node in the target knowledge graph finally obtained.

[0061] In a specific embodiment, for the convenience of description, various variables are configured: G: the full set of the entire graph; A: the probability that the viewer continues browsing on a page (point) instead of stopping at the page (point); Mass: the unstandardized probability of a point; OutEdges: all outgoing edges of a point; Destination: the end point of an edge. After obtaining the customer's operation information, asynchronous technology can be used to avoid the synchronous stage to accelerate convergence; the asynchronous model used can be shown in the following algotithm0:

[0062]

[0063]

[0064] In addition, each vertex (node ​​corresponding to the entity) triggers the callback function InitializeVertex(), that is, the first preset function, when it is instantiated for the first time; each time the callback function SendMessage() is triggered when the vertex receives at least one message, the vertex reallocates the traffic it receives, that is, the unstandardized probability amount.

[0065] Furthermore, when the relationship between related nodes needs to be disconnected, that is, when the edge between the nodes needs to be deleted, a callback function as shown in Algor ithm1 can be triggered:

[0066]

[0067] Specifically, first calculate the total outflow flow distAllPrev of point src; then update the flow between other neighbors: calculate the old flow distOld, the new flow distNew, the flow difference distDelta, and trigger the callback function SendMessage to other neighbors. Then update and delete the flow between neighbors: calculate the opposite number of the original flow toOldEdge, and trigger the callback function SendMessage to the new neighbor to delete the original flow.

[0068] Correspondingly, when it is necessary to establish a relationship between related nodes, that is, when it is necessary to establish an edge between nodes, a callback function as shown in Algorithm 2 can be triggered:

[0069]

[0070]

[0071] Specifically, first calculate the total outflow flow distAllPrev from point src; then update the flow with the old neighbor: calculate the old flow distOld, the new flow distNew, the flow difference distDelta, and trigger the callback function SendMessage to the old neighbor. Then update the flow with the new neighbor: calculate the new flow toNewEdge, and trigger the callback function SendMessage to the new neighbor.

[0072] It is understandable that the simple asynchronous SendMessage method cannot correctly handle the sinks of dynamic graphs because it will cause traffic leakage at the sinks, so special processing is required. Specifically, after executing the callback function for adding / deleting edges, before the calculation of this step begins, it is necessary to copy u.Mass of all points u (which can be executed in parallel) to obtain a new copy u._Mass. All subsequent standardization (quantization calculation) and sink processing modification operations are performed on _Mass instead of Mass, and the final standardized probability distribution result (i.e. pagerank) is also given by _Mass. Furthermore, when querying PageRank, the function for normalization and sink processing to obtain the final result can be shown in Algorithm3:

[0073]

[0074]

[0075] In this way, the intermediate results (Mass of each point) obtained previously can be standardized and the sinks can be specially processed to obtain the final quantitative results; subsequently, the risks of related nodes can be evaluated based on the quantitative results to screen out the nodes that require manual review.

[0076] It can be seen that the present application proposes an asynchronous model for calculating PageRank, which is convenient for online maintenance of dynamic graphs; the PageRank scores of dynamic graphs are maintained in real time, and standardized processing and special processing of sinks are performed when users query to quickly obtain corresponding quantitative results, so as to manually review abnormal nodes.

[0077] like Figure 5 As shown, the embodiment of the present application discloses a knowledge graph maintenance device, including:

[0078] The information acquisition module 11 is used to obtain the customer's operation information on several entities; the several entities are entities related to the credit risk control business;

[0079] A knowledge graph updating module 12 is used to update the initial knowledge graph using the entities and the corresponding operation information to obtain a target knowledge graph;

[0080] The knowledge graph quantization module 13 is used to perform quantitative calculations on the target knowledge graph based on the principle of the web page ranking algorithm to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph.

[0081] It can be seen that the present application can quickly handle the centrality problem of dynamically changing knowledge graphs, update relevant knowledge graphs according to the customer's operation information on related entities, and use the web page ranking algorithm's ability to process dynamic graphs to perform quantitative calculations on the updated target knowledge graph, obtain quantitative results through standardized processing, complete the knowledge graph maintenance process, and improve efficiency.

[0082] In a specific embodiment, the knowledge graph updating module 12 may include:

[0083] A node adding submodule, configured to instantiate nodes of the entities based on a first preset function, so as to add new nodes corresponding to the entities to the initial knowledge graph to obtain a first knowledge graph;

[0084] A node relationship disconnection submodule, used to disconnect the relationship between related nodes in the first knowledge graph based on a second preset function and using the operation information to obtain a second knowledge graph;

[0085] A node relationship establishment submodule is used to establish relationships between related nodes in the second knowledge graph based on a third preset function and using the operation information to obtain a target knowledge graph.

[0086] In a specific embodiment, the node adding submodule may include:

[0087] An entity instantiation unit, configured to perform node instantiation operations on the entities according to an asynchronous processing principle and based on a first preset function to obtain new nodes corresponding to the entities;

[0088] A node adding unit is used to add the new node to the initial knowledge graph to obtain a first knowledge graph.

[0089] In a specific embodiment, the node relationship disconnection submodule may include:

[0090] A current node determination unit, configured to determine a current node from the first knowledge graph according to the operation information;

[0091] A node relationship disconnection unit is used to disconnect the relationship between the current node and other nodes based on a second preset function and using the node relationship information corresponding to the current node in the operation information to obtain a second knowledge graph.

[0092] In a specific embodiment, the node relationship establishment submodule may include:

[0093] a node relationship determination unit, configured to determine a plurality of groups of node relationships to be established according to the node relationship information of the new node corresponding to the operation information;

[0094] A node relationship establishing unit is used to establish corresponding node relationships in the second knowledge graph in sequence based on a third preset function and using the node relationships to be established to obtain a target knowledge graph.

[0095] In a specific embodiment, the knowledge graph quantization module 13 may include:

[0096] A knowledge graph replication unit, used to replicate the node information in the target knowledge graph according to the asynchronous processing principle to obtain a corresponding replicated knowledge graph;

[0097] A node quantization unit is used to calculate several nodes in the copied knowledge graph based on the principle of the web page ranking algorithm to obtain a quantization result for the copied knowledge graph; the quantization result has a one-to-one correspondence with each node in the copied knowledge graph.

[0098] In a specific embodiment, the knowledge graph quantization module 13 may include:

[0099] A result judging unit, used to judge whether the quantization result is not greater than a preset result threshold;

[0100] A node marking unit is used to mark the node in the target knowledge graph corresponding to the quantification result greater than the preset result threshold as a node to be reviewed when the quantization result is greater than the preset result threshold, so as to perform manual review on the node to be reviewed; wherein a single node corresponds to a single entity.

[0101] Furthermore, the present application also discloses an electronic device. Figure 6 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram cannot be regarded as any limitation on the scope of use of the present application.

[0102] Figure 6 A schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the knowledge graph maintenance method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0103] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0104] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0105] The operating system 221 is used to manage and control the hardware devices and computer programs 222 on the electronic device 20, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the knowledge graph maintenance method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0106] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned disclosed knowledge graph maintenance method is implemented. The specific steps of the method can be referred to the corresponding contents disclosed in the aforementioned embodiments, and will not be repeated here.

[0107] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0108] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0109] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0110] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0111] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A knowledge graph maintenance method, characterized in that: include: Obtaining the customer's operation information on several entities; the several entities are entities related to the credit risk control business; Using the entities and the corresponding operation information, the initial knowledge graph is updated to obtain a target knowledge graph; Based on the principle of web page ranking algorithm, the target knowledge graph is quantitatively calculated to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph; Wherein, the updating of the initial knowledge graph by using the several entities and the corresponding operation information to obtain the target knowledge graph includes: Instantiating nodes of the entities based on a first preset function, so as to add new nodes corresponding to the entities to the initial knowledge graph, thereby obtaining a first knowledge graph; Based on a second preset function and using the operation information, disconnecting the relationship between related nodes in the first knowledge graph to obtain a second knowledge graph; Based on a third preset function, and using the operation information, a relationship between related nodes is established in the second knowledge graph to obtain a target knowledge graph; The target knowledge graph is quantitatively calculated based on the principle of the web page ranking algorithm to obtain the corresponding quantitative results, including: According to the asynchronous processing principle, the node information in the target knowledge graph is copied to obtain the corresponding copied knowledge graph; Based on the principle of web page ranking algorithm, several nodes in the copied knowledge graph are calculated to obtain a quantitative result for the copied knowledge graph; the quantitative result has a one-to-one correspondence with each node in the copied knowledge graph.

2. The knowledge graph maintenance method according to claim 1, characterized in that: The instantiating nodes of the entities based on the first preset function to add new nodes corresponding to the entities to the initial knowledge graph to obtain a first knowledge graph includes: According to the asynchronous processing principle and based on the first preset function, a node instantiation operation is performed on the entities to obtain new nodes corresponding to the entities; Add the new node to the initial knowledge graph to obtain a first knowledge graph.

3. The knowledge graph maintenance method according to claim 1, characterized in that: The method of disconnecting the relationship between related nodes in the first knowledge graph based on the second preset function and using the operation information to obtain the second knowledge graph includes: Determine a current node from the first knowledge graph according to the operation information; Based on a second preset function, and using the node relationship information corresponding to the current node in the operation information, the relationship between the current node and other nodes is disconnected to obtain a second knowledge graph.

4. The knowledge graph maintenance method according to claim 1, characterized in that: The step of establishing a relationship between related nodes in the second knowledge graph based on the third preset function and using the operation information to obtain a target knowledge graph includes: Determine a plurality of groups of node relationships to be established according to the node relationship information of the new node corresponding to the operation information; Based on the third preset function and using the node relationships to be established, corresponding node relationships are established in the second knowledge graph in sequence to obtain the target knowledge graph.

5. The knowledge graph maintenance method according to any one of claims 1 to 4, characterized in that: The process of completing the maintenance of the relevant knowledge graph includes: Determining whether the quantization result is not greater than a preset result threshold; If so, the node in the target knowledge graph corresponding to the quantization result greater than the preset result threshold is marked as a node to be reviewed so that the node to be reviewed can be manually reviewed; wherein a single node corresponds to a single entity.

6. A knowledge graph maintenance device, characterized in that: include: An information acquisition module, used to acquire the customer's operation information on several entities; the several entities are entities related to the credit risk control business; A knowledge graph updating module, used to update the initial knowledge graph using the entities and the corresponding operation information to obtain a target knowledge graph; The knowledge graph quantification module is used to perform quantitative calculations on the target knowledge graph based on the principle of the web page ranking algorithm to obtain corresponding quantitative results to complete the maintenance process of the relevant knowledge graph; Wherein, the knowledge graph updating module includes: A node adding submodule, configured to instantiate nodes of the entities based on a first preset function, so as to add new nodes corresponding to the entities to the initial knowledge graph to obtain a first knowledge graph; A node relationship disconnection submodule, used to disconnect the relationship between related nodes in the first knowledge graph based on a second preset function and using the operation information to obtain a second knowledge graph; A node relationship establishment submodule, used to establish a relationship between related nodes in the second knowledge graph based on a third preset function and using the operation information to obtain a target knowledge graph; Wherein, the knowledge graph quantization module includes: A knowledge graph replication unit, used to replicate the node information in the target knowledge graph according to the asynchronous processing principle to obtain a corresponding replicated knowledge graph; A node quantization unit is used to calculate several nodes in the copied knowledge graph based on the principle of the web page ranking algorithm to obtain a quantization result for the copied knowledge graph; the quantization result has a one-to-one correspondence with each node in the copied knowledge graph.

7. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, used to execute the computer program to implement the knowledge graph maintenance method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, implements the knowledge graph maintenance method as described in any one of claims 1 to 5.

Images