Big data auditing method and system based on knowledge graph

Abstract

The invention discloses a big data auditing method and system based on a knowledge graph, and relates to the technical field of data processing, the structure of the knowledge graph is determined, and the knowledge graph is loaded into an auditing database; setting an audit assistant corresponding to the knowledge graph based on the determined structure of the knowledge graph, the audit assistant comprising a plurality of audit data versions and a plurality of corresponding audit jump points; and obtaining an audit report of the audit target in the knowledge graph based on the audit assistant. According to the method, the temporary patrol point continues to move in the patrol track in the audit range and analyzes the data in the knowledge graph corresponding to the patrol track, so that the audit data of the temporary patrol point which is directly and indirectly started can be obtained, and the audit information is obtained and marked on the corresponding patrol track to be integrated by the audit jump point; the multiple audit data versions are used for data analysis of the knowledge graph, and have the effects of being clear in division of labor, free of mutual interference and capable of conducting efficient audit.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a big data auditing method and system based on knowledge graph. Background Art

[0002] In the power supply operation of power companies, there is a large amount of power-related data that needs to be audited regularly or at any time. For example, the power-related data that needs to be audited includes the audit of electricity bills, the audit of accounts receivable (electricity bills), and the data-based audit process. Data auditing through knowledge graphs can reveal potential risk points and violations through analysis and mining of massive data. At present, in the process of data auditing through knowledge graphs, confusion is prone to occur during the audit due to the large amount of data, making it difficult to audit data efficiently and accurately. Summary of the invention

[0003] The purpose of the present invention is to provide a big data auditing method and system based on knowledge graph to address the deficiencies in the background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solution: a big data audit method based on knowledge graph, comprising the following steps: Determine the structure of the knowledge graph and load the knowledge graph into the audit database; An audit assistant is set up corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; Obtain the audit report of the audit target in the knowledge graph based on the audit assistant.

[0005] In a preferred embodiment, the step of determining the structure of the knowledge graph and loading the knowledge graph into the audit database includes: Collect the nodes of the knowledge graph and the edges between the nodes as the structure of the knowledge graph; Set up an audit database, load the knowledge graph into the audit database, and locate the knowledge graph.

[0006] In a preferred embodiment, the steps of setting up an audit database, loading the knowledge graph into the audit database and locating the knowledge graph include: Setting up a database, setting up multiple data locations in the database, and associating the multiple data locations with each other to obtain an audit database; Multiple data sites are respectively associated with multiple nodes in the knowledge graph to obtain a position association network, the shape of the position association network is determined, and the knowledge graph is located according to the positioning index, where the calculation formula of the positioning index is: ,in, is the positioning index, is the number of nodes whose positions change between the data points and the corresponding nodes in the knowledge graph, is the position difference distance between the data point and the corresponding node, is the number of nodes in the knowledge graph, and are all constants greater than zero; The knowledge graph corresponding to the positioning index that meets the threshold is regarded as a qualified knowledge graph.

[0007] In a preferred embodiment, the step of setting an audit assistant based on the knowledge graph corresponding to the structure of the knowledge graph includes: Set up multiple data layers in the audit database, and connect the multiple data layers to the knowledge graph to obtain multiple audit data versions. The multiple data layers are isolated from each other and connected to the knowledge graph. Multiple audit jump points are set on multiple audit data boards, and multiple audit jump points are connected through audit data board communication; Use multiple audit data versions and corresponding multiple audit jump points as audit assistants.

[0008] In a preferred embodiment, the steps of setting multiple data layers in the audit database, respectively connecting the multiple data layers with the knowledge graph to obtain multiple audit data versions, and isolating the multiple data layers from each other and connecting them with the knowledge graph, include: Set up multiple data layers in the audit database. All of the multiple data layers are connected to the audit database data. The multiple data layers are isolated from each other. The corresponding audit types are marked on the corresponding data layers. Multiple data layers are fitted with the knowledge graph, and the positions where the knowledge graph and the data layer are fitted are connected to obtain the inspection trajectory, wherein the inspection trajectory includes inspection points and inspection lines, and corresponding data space is allocated to the inspection trajectory in the data layer; Based on the data layer, the association between the data space of the inspection trajectory and the knowledge graph is established to obtain multiple audit data versions.

[0009] In a preferred embodiment, the step of setting multiple audit jump points on multiple audit data plates, and connecting the multiple audit jump points through the audit data plates, includes: In the inspection track of the audit data version, multiple audit jump points corresponding to the audit type are set, corresponding audit ranges are assigned to the multiple audit jump points respectively, and the audit jump points are bound to the corresponding audit ranges; In the inspection track within the audit scope, inspection points that meet the preset conditions are selected to set up an inspection warehouse, wherein the inspection warehouse includes multiple temporary inspection points and a temporary frame, and the multiple temporary inspection points are all connected to the temporary frame; Multiple audit jump points are connected through audit data communication.

[0010] In a preferred embodiment, the step of obtaining an audit report of the audit target in the knowledge graph based on the audit assistant includes: In the audit data version, based on the audit rules, the audit jump point is moved in the inspection track within the corresponding audit scope, and the data in the knowledge graph corresponding to the inspection track is analyzed; When the audit jump point moves to the position of the inspection warehouse, a connection relationship is established between the audit jump point and the temporary inspection point in the inspection warehouse and the temporary inspection point is enabled. The temporary inspection point moves in the inspection track within the audit range and analyzes the data in the knowledge graph corresponding to the inspection track. When the temporary inspection point moves to the position of the inspection warehouse, it serves as the first temporary inspection point. A connection relationship is established between the first temporary inspection point and the temporary inspection point in the inspection warehouse, and the temporary inspection point is enabled. The temporary inspection point continues to move in the inspection track within the audit scope and analyzes the data in the knowledge graph corresponding to the inspection track, and provides the data audit data of the temporary inspection point to the corresponding audit jump point. After the data audit of the entire audit data version is completed and a single audit report is obtained, the temporary inspection points and the temporary inspection points and the audit jump points are disconnected and returned to the inspection warehouse; The audit report of the knowledge graph is obtained by integrating multiple audit data versions of a single design report through the audit database.

[0011] The present invention also provides a big data audit system based on knowledge graph, comprising: A determination module is used to determine the structure of the knowledge graph and load the knowledge graph into the audit database; A setting module, connected to the determination module, is used to set an audit assistant corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; The audit module, connected with the setting module, is used to obtain the audit report of the audit target in the knowledge graph based on the audit assistant.

[0012] In the above technical solution, the technical effects and advantages provided by the present invention are: The present invention continues to move in the inspection track within the audit scope through temporary inspection points and analyzes the data in the knowledge graph corresponding to the inspection track. It can obtain the audit data of the temporary inspection points directly and indirectly enabled, obtain the audit information and mark it on the corresponding inspection track. The audit jump point integrates it, and multiple audit data versions are used for data analysis of the knowledge graph, which has clear division of labor, does not interfere with each other, and can perform efficient auditing. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0014] Figure 1 The figure is a flow chart of the method of the present invention.

[0015] Figure 2 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0017] Example 1, please refer to Figure 1 As shown, the big data audit method based on knowledge graph described in this embodiment includes the following steps: S1. Determine the structure of the knowledge graph and load the knowledge graph into the audit database; S2. Setting an audit assistant corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; S3, based on the audit assistant, obtain the audit report of the audit target in the knowledge graph; As described in the above steps S1-S3, in the construction of the knowledge graph, it can be constructed through semantic features. First, natural language processing and other technologies are used to automatically extract audit knowledge from text data such as audit documents and audit reports; the text data can be preprocessed, such as word segmentation, part-of-speech tagging, etc., and then semantic analysis technology is used to extract and summarize knowledge; this helps auditors to quickly acquire audit knowledge and reduce the workload of manual input. In addition, semantic analysis technologies, such as predicate logic, semantic networks, etc., are used to perform knowledge reasoning and discover potential connections and rules between audit knowledge; combined with audit objectives, appropriate reasoning algorithms can be designed to automatically reason and analyze audit knowledge, improve the efficiency of audit work, and optimize the knowledge graph for auditing.

[0018] In one embodiment, the step S1 of determining the structure of the knowledge graph and loading the knowledge graph into the audit database includes: S11, collecting nodes of the knowledge graph and edges between the nodes as the structure of the knowledge graph; S12. Set up an audit database, load the knowledge graph into the audit database and locate the knowledge graph; As described in the above steps S11 and S12, the audit database is set up to carry data for the knowledge graph and facilitate the subsequent setting of the audit assistant. Multiple data sites are set in the audit database to locate the storage location of the knowledge graph. After the knowledge graph is stored in the audit database, the nodes of the knowledge graph and the edges between the nodes are obtained as the structure of the knowledge graph. The nodes in the knowledge graph can cover various types such as people, places, organizations, events, etc. The edges between the nodes connect different nodes and represent the relationship between entities. The edges can be directed (indicating a specific flow direction of the relationship) or undirected (indicating only the association between entities). The specific representation here is related to the establishment of the knowledge graph, which can better clarify the relationship between the audit target subjects.

[0019] In one embodiment, the step S12 of setting up an audit database, loading the knowledge graph into the audit database and locating the knowledge graph includes: S121, setting a database, setting multiple data sites in the database, and performing position association between the multiple data sites to obtain an audit database; S122, positionally associate multiple data sites with multiple nodes in the knowledge graph to obtain a position association network, determine the shape of the position association network, and locate the knowledge graph according to the positioning index, wherein the calculation formula of the positioning index is: ,in, is the positioning index, is the number of nodes whose positions change between the data points and the corresponding nodes in the knowledge graph, is the position difference distance between the data point and the corresponding node, is the number of nodes in the knowledge graph, and are all constants greater than zero; S123, taking the knowledge graph corresponding to the positioning index that meets the threshold as a qualified knowledge graph; As described in the above steps S121-S123, a database is set up, and the database is used to store the knowledge graph. Before storing the knowledge graph, preparations need to be made, and multiple data sites are set in the database, wherein the data site is a port set at a storage position in the database, and the multiple data sites are positionally associated, where the position association is associated through a communication connection relationship, and then the multiple data sites are positionally associated with multiple nodes in the knowledge graph, and a port is set on the node in the knowledge graph, and the port of the node on the knowledge graph is connected to the multiple data sites respectively, where the connection is not used for data exchange, but only for position association, so that a position association network can be obtained, and the position is limited by multiple data sites and the knowledge graph, and then the shape of the position association network is determined, so that the position of the knowledge graph in the audit database can be guaranteed to remain unchanged, which is convenient for subsequent data auditing through the audit assistant, and has better data storage and positioning functions. When a positioning index that does not meet the threshold appears, it means that the knowledge graph is unstable and unqualified, and there may be structural changes in network instability, which can monitor the stability of the knowledge graph data.

[0020] In one embodiment, the step S2 of setting an audit assistant based on the knowledge graph corresponding to the structure of the knowledge graph includes: S21. Multiple data layers are set in the audit database, and the multiple data layers are respectively connected to the knowledge graph to obtain multiple audit data versions, wherein the multiple data layers are isolated from each other and connected to the knowledge graph; S22, setting multiple audit jump points on multiple audit data plates, and the multiple audit jump points are connected through audit data plate communication; S23, using multiple audit data versions and corresponding multiple audit jump points as audit assistants; As described in the above steps S21-S23, multiple data layers are set in the audit database, which is a spatial data layer (capable of carrying the operation of subsequent audit jump points, and capable of associating data in the knowledge graph, and is a server with a storage space shape), which can be structurally fitted with the knowledge graph, and can access the knowledge graph through the data layer at the fitted position. The multiple data layers are data isolated, so when executing multiple tasks, they can be performed simultaneously and the data is isolated to avoid data mixing, which has a fast data auditing function. Afterwards, multiple audit jump points are set on the audit data version, which can sort out the structure of the knowledge graph, and use multiple audit data versions and corresponding multiple audit jump points as audit assistants. The data of multiple audit data versions can be integrated through the audit database, which can ensure the stability of the knowledge graph and the accuracy of the data, making it convenient for subsequent auditors to obtain audit reports.

[0021] In one embodiment, the step S21 of setting multiple data layers in the audit database, respectively connecting the multiple data layers to the knowledge graph to obtain multiple audit data versions, wherein the multiple data layers are isolated from each other and connected to the knowledge graph, includes: S211. Multiple data layers are set in the audit database. The multiple data layers are all connected to the audit database data. The multiple data layers are isolated from each other. The corresponding audit types are marked on the corresponding data layers. S212, multiple data layers are all fitted with the knowledge graph, and the positions where the knowledge graph and the data layer are fitted are connected to obtain a patrol track, wherein the patrol track includes patrol points and patrol lines, and corresponding data spaces are allocated in the data layer corresponding to the patrol track; S213, establishing a correlation relationship between the data space of the inspection trajectory and the knowledge graph based on the data layer, and obtaining multiple audit data versions; In one embodiment, the step S22 of setting multiple audit jump points on multiple audit data boards, wherein the multiple audit jump points are communicatively connected via the audit data boards, includes: S221. Set multiple audit jump points corresponding to the audit type in the inspection track of the audit data version, assign corresponding audit ranges to the multiple audit jump points, and bind the audit jump points to the corresponding audit ranges; S222, selecting inspection points that meet preset conditions in the inspection track within the audit scope to set up an inspection warehouse, wherein the inspection warehouse includes multiple temporary inspection points and a temporary frame, and the multiple temporary inspection points are all connected to the temporary frame; S223. Multiple audit jump points are connected through audit data communication.

[0022] As described in the above steps S211-S213 and S221-S223, the multiple data layers set in the audit database are servers with data storage and operation functions. The multiple data layers are isolated from each other and there is no data interaction relationship. The knowledge graph can be audited separately without affecting each other. The multiple data layers are connected to the audit database data. Finally, the data layer obtains the data audit report which can be summarized through the audit database. After that, the multiple data layers are all fitted with the knowledge graph. Since the knowledge graph is composed of nodes and edges, which represents the information of multiple audit subjects and the relationship between them, the inspection tracks with the same structure as the knowledge graph can be formed at the positions where the multiple data layers are fitted. The inspection track here is used for subsequent audit jump point mobile inspection. The inspection track includes inspection points and inspection lines, where the inspection points are nodes that fit the corresponding knowledge graph, and the inspection lines are edges that fit the corresponding knowledge graph. In order to enable the subsequent audit jump points to move on the inspection track, it is necessary to rely on the data layer. The corresponding data space is allocated to the inspection track on the data layer for the subsequent audit jump point to run. The association relationship between the data space of the inspection track and the knowledge graph is established based on the data layer. The data in the knowledge graph can be accessed through the inspection track. Multiple audit data versions can independently perform the corresponding data audit goals. For example, when the audit subject is an enterprise, there are two audit data versions, one of which is an audit data version. The audit jump point in the audit data version is used to audit financial statements, and the audit jump point in another audit data version is used to audit transactions and other information between audit subjects, so as to obtain multiple audit data versions. Multiple audit jump points are set in the inspection track of each audit data version, and corresponding audit scopes are respectively assigned to the multiple audit jump points. The audit jump points are bound to the corresponding audit scopes. The audit jump points are only responsible for collecting audit data within the bound audit scopes. In the inspection track within the audit scope, inspection points that meet the preset conditions are selected to set up inspection warehouses, wherein the inspection warehouse includes multiple temporary inspection points, and the preset conditions are inspection points with multiple connected edges. The inspection points represent the audit subjects. When the inspection points have multiple edges, it means that their relationships are complex. Therefore, an inspection warehouse is set on the inspection points. Audit jump points can use temporary inspection points to conduct data audits more efficiently. Multiple audit jump points are connected through the audit data version. The audit data of multiple audit jump points can be collected by the audit data version. The audit data of the audit data version can be collected by the audit database and can be displayed to auditors for understanding. Since there are many types of audit projects, one data layer corresponds to one audit type. The audit jump points in the data layer only have data audit actions corresponding to the audit type. For example, if the audit data version corresponds to financial audit, the audit jump points in the audit data version only audit the data in finance, and do not analyze other parts. It has clear division of labor, does not interfere with each other, and can perform efficient audits.

[0023] In one embodiment, the step S3 of obtaining an audit report of the audit target in the knowledge graph based on the audit assistant includes: S31. In the audit data version, based on the audit rules, the audit jump point is moved in the inspection track within the corresponding audit scope, and the data in the knowledge graph corresponding to the inspection track is analyzed; S32. When the audit jump point moves to the position of the inspection warehouse, a connection relationship between the audit jump point and the temporary inspection point in the inspection warehouse is established and the temporary inspection point is enabled. The temporary inspection point moves in the inspection track within the audit range and analyzes the data in the knowledge graph corresponding to the inspection track. S33, when the temporary inspection point moves to the position of the inspection warehouse, it serves as the first temporary inspection point, establishes a connection relationship between the first temporary inspection point and the temporary inspection point in the inspection warehouse, and enables the temporary inspection point, continues to move in the inspection track within the audit scope through the temporary inspection point, analyzes the data in the knowledge graph corresponding to the inspection track, and provides the data audit data of the temporary inspection point to the corresponding audit jump point; S34, until the data audit of the entire audit data version is completed and a single audit report is obtained, the temporary inspection points and the temporary inspection points and the audit jump points are disconnected and returned to the inspection warehouse; S35. Integrate the single design reports of multiple audit data versions through the audit database to obtain the audit report of the knowledge graph; As described in the above steps S31-S35, during the use process, after obtaining the knowledge graph, since there may be multiple audit types, for example, when conducting financial and contract or transaction audits at the same time, the mutual influence of audit data makes it difficult to ensure the efficiency and accuracy of data audits. Here, data audits of corresponding audit types are performed respectively through multiple audit data versions. First, in the audit data version, the audit jump point is moved in the inspection track within the corresponding audit range based on the audit rules, and the data in the knowledge graph corresponding to the inspection track is analyzed. The data rules here include: data query and retrieval rules to quickly locate and analyze target data; data association and mining rules to discover potential relationships and patterns between data; risk warning and identification rules, when the data meets specific conditions, the warning mechanism is triggered and the standards and methods for risk identification are formulated, such as identifying potential risks by setting risk indicators and thresholds, etc. The audit rules corresponding to the audit type of the data are different;Afterwards, when the audit jump point moves to the position of the inspection warehouse, a connection relationship between the audit jump point and the temporary inspection point in the inspection warehouse is established and the temporary inspection point is enabled. The process of enabling the temporary inspection point is: the audit jump point is placed in the temporary frame, where the temporary frame is a communication point (blank and has no function as a point that only serves as a communication relationship). After the audit jump point is placed in the temporary frame, it can overlap with the temporary frame, and the communication function of the temporary frame can be used for subsequent data transmission. Afterwards, there is no need to temporarily build a connection relationship between the audit jump point and the temporary inspection point, and it can be used directly, ensuring the efficiency of enabling the temporary inspection point. Afterwards, the audit jump point can remain in the current inspection warehouse, and other temporary inspection points are released from the inspection warehouse for The audit of data places the audit jump point in the temporary frame in the following way: the temporary frame consists of a connection layer and a content layer, wherein the content layer is located inside the connection layer, and a plurality of corresponding ports connected to the temporary inspection points are arranged on the connection layer. The content layer is a fake server inside the connection layer (without any functional role, and only serves as a temporary server that can carry the function of connecting to the temporary inspection point in cooperation with the connection layer). The audit jump point is replaced with the content layer to complete the action of placing the audit jump point in the temporary frame. When the fake server is replaced, it will be automatically hidden and not enabled. After the entire audit report is completed, the audit jump point will leave the temporary frame, and the fake server will be enabled again, continuing to be within the audit scope through the temporary inspection point. The number of temporary inspection points is not less than the number of edges of the inspection point. Similarly, when the temporary inspection point moves to the position of the inspection warehouse, it is used as the first temporary inspection point. The connection relationship between the first temporary inspection point and the temporary inspection point in the inspection warehouse is established and the temporary inspection point is enabled. The inspection trajectory within the audit range continues to move through the temporary inspection point and the data in the knowledge graph corresponding to the inspection trajectory is analyzed. The audit data of the temporary inspection points directly and indirectly enabled can be obtained, and the audit information is obtained and marked on the corresponding inspection trajectory for integration by the audit jump point. The audit jump point here is the data analyzer. The temporary inspection point is the same as the audit jump point, but The functions performed are different. The audit jump point is bound to the corresponding audit scope and has the function of data integration. The temporary inspection point is equivalent to the incarnation of the audit jump point. It can enable the temporary inspection point and connect the audit jump point. Finally, the audit data of the temporary inspection point is provided to the audit jump point. The audit rules of the temporary inspection point are the same as those of the audit jump point. Finally, after the data audit of the entire audit data version is completed and a single audit report is obtained, the temporary inspection points and the temporary inspection points and the audit jump points are disconnected and returned to the inspection warehouse. The disconnection here means that the audit jump point or the temporary inspection point is separated from the temporary frame. Finally, the single design reports of multiple audit data versions are integrated through the audit database to obtain the audit report of the knowledge graph. ;

[0024] Example 2, please refer to Figure 2As shown, the big data audit system based on knowledge graph described in this embodiment includes: A determination module is used to determine the structure of the knowledge graph and load the knowledge graph into the audit database; A setting module, connected to the determination module, is used to set an audit assistant corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; The audit module, connected with the setting module, is used to obtain the audit report of the audit target in the knowledge graph based on the audit assistant.

[0025] By continuing to move in the inspection track within the audit scope through temporary inspection points and analyzing the data in the knowledge graph corresponding to the inspection track, it is possible to obtain the audit data of the temporary inspection points directly and indirectly enabled, obtain the audit information and mark it on the corresponding inspection track. The audit jump points are integrated and multiple audit data versions are used for data analysis of the knowledge graph, which has clear division of labor, does not interfere with each other, and can perform efficient auditing.

[0026] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A big data audit method based on knowledge graph, characterized in that: The following steps are involved: Determine the structure of the knowledge graph and load the knowledge graph into the audit database; An audit assistant is set up corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; Obtain the audit report of the audit target in the knowledge graph based on the audit assistant.

2. According to the big data audit method based on knowledge graph according to claim 1, it is characterized by: The step of determining the structure of the knowledge graph and loading the knowledge graph into the audit database includes: Collect the nodes of the knowledge graph and the edges between the nodes as the structure of the knowledge graph; Set up an audit database, load the knowledge graph into the audit database, and locate the knowledge graph.

3. According to the big data audit method based on knowledge graph according to claim 2, it is characterized by: The steps of setting up an audit database, loading the knowledge graph into the audit database and locating the knowledge graph include: Setting up a database, setting up multiple data locations in the database, and associating the multiple data locations with each other to obtain an audit database; Multiple data sites are respectively associated with multiple nodes in the knowledge graph to obtain a position association network, the shape of the position association network is determined, and the knowledge graph is located according to the positioning index, where the calculation formula of the positioning index is: ,in, is the positioning index, is the number of nodes whose positions change between the data points and the corresponding nodes in the knowledge graph, is the position difference distance between the data point and the corresponding node, is the number of nodes in the knowledge graph, and are all constants greater than zero; The knowledge graph corresponding to the positioning index that meets the threshold is regarded as a qualified knowledge graph.

4. According to the big data audit method based on knowledge graph according to claim 1, it is characterized by: The step of setting an audit assistant based on the knowledge graph corresponding to the structure of the knowledge graph includes: Set up multiple data layers in the audit database, and connect the multiple data layers to the knowledge graph to obtain multiple audit data versions. The multiple data layers are isolated from each other and connected to the knowledge graph. Multiple audit jump points are set on multiple audit data boards, and the multiple audit jump points are connected through audit data board communication; Use multiple audit data versions and corresponding multiple audit jump points as audit assistants.

5. According to the big data audit method based on knowledge graph according to claim 4, it is characterized by: The step of setting multiple data layers in the audit database, fitting and connecting the multiple data layers with the knowledge graph to obtain multiple audit data versions, and isolating the multiple data layers from each other and connecting them with the knowledge graph, includes: Set up multiple data layers in the audit database. All of the multiple data layers are connected to the audit database data. The multiple data layers are isolated from each other. The corresponding audit types are marked on the corresponding data layers. Multiple data layers are fitted with the knowledge graph, and the positions where the knowledge graph and the data layer are fitted are connected to obtain the inspection trajectory, wherein the inspection trajectory includes inspection points and inspection lines, and corresponding data space is allocated to the inspection trajectory in the data layer; Based on the data layer, the association between the data space of the inspection trajectory and the knowledge graph is established to obtain multiple audit data versions.

6. According to claim 5, a big data audit method based on knowledge graph is characterized in that: The step of setting multiple audit jump points on multiple audit data plates, and connecting the multiple audit jump points through the audit data plates, includes: In the inspection track of the audit data version, multiple audit jump points corresponding to the audit type are set, corresponding audit ranges are assigned to the multiple audit jump points respectively, and the audit jump points are bound to the corresponding audit ranges; In the inspection track within the audit scope, inspection points that meet the preset conditions are selected to set up an inspection warehouse, wherein the inspection warehouse includes multiple temporary inspection points and a temporary frame, and the multiple temporary inspection points are all connected to the temporary frame; Multiple audit jump points are connected through audit data communication.

7. The big data audit method based on knowledge graph according to claim 1 is characterized in that: The step of obtaining an audit report of the audit target in the knowledge graph based on the audit assistant includes: In the audit data version, based on the audit rules, the audit jump point is moved in the inspection track within the corresponding audit scope, and the data in the knowledge graph corresponding to the inspection track is analyzed; When the audit jump point moves to the position of the inspection warehouse, a connection relationship is established between the audit jump point and the temporary inspection point in the inspection warehouse and the temporary inspection point is enabled. The temporary inspection point moves in the inspection track within the audit range and analyzes the data in the knowledge graph corresponding to the inspection track. When the temporary inspection point moves to the position of the inspection warehouse, it serves as the first temporary inspection point. A connection relationship is established between the first temporary inspection point and the temporary inspection point in the inspection warehouse, and the temporary inspection point is enabled. The temporary inspection point continues to move in the inspection track within the audit scope and analyzes the data in the knowledge graph corresponding to the inspection track, and provides the data audit data of the temporary inspection point to the corresponding audit jump point. After the data audit of the entire audit data version is completed and a single audit report is obtained, the temporary inspection points and the temporary inspection points and the audit jump points are disconnected and returned to the inspection warehouse; The audit report of the knowledge graph is obtained by integrating multiple audit data versions of a single design report through the audit database.

8. A big data audit system based on knowledge graph, used to implement a big data audit method based on knowledge graph as described in any one of claims 1 to 7, characterized in that: include: A determination module is used to determine the structure of the knowledge graph and load the knowledge graph into the audit database; A setting module, connected to the determination module, is used to set an audit assistant corresponding to the knowledge graph based on the structure of the determined knowledge graph, wherein the audit assistant includes multiple audit data versions and corresponding multiple audit jump points; The audit module, connected with the setting module, is used to obtain the audit report of the audit target in the knowledge graph based on the audit assistant.

Images