Data verification knowledge graph model generation method, data verification method and device

Abstract

The application provides a data verification knowledge graph model generation method, a data verification method and device, and belongs to the technical field of data processing. The data verification method comprises the following steps: acquiring a knowledge graph model; acquiring XDR data of a target to be verified; extracting second information corresponding to a field name in the knowledge graph model from the XDR data of the target to be verified; generating an actual knowledge graph according to the second information and nodes of the knowledge graph model; analyzing values of each node in the actual knowledge graph according to the knowledge graph model, and generating a verification result of the target to be verified according to an analysis result; wherein the knowledge graph model is a knowledge graph model generated by the data verification knowledge graph model generation method provided by an embodiment of the application; XDR verification is performed through the knowledge graph, the logicality and interpretability of the verification result can be improved, and the cost of manual maintenance can be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a data verification knowledge graph model generation method, a data verification method and device. BACKGROUND

[0002] In order to ensure the communication experience of users, analyze the problems that users cannot access the network or have poor network experience, and perform unified DPI (Deep Packet Inspection), it is a very common way. The signaling side of the unified DPI system will record the control and feedback information (original code stream) between the network and the user. Since the amount of these original code stream information is large and the types are various, the unified DPI system will convert each piece of information into an XDR (X Data Recording) information, and check the XDR information.

[0003] In the prior art, XDR verification mainly checks the original code stream. This method can only check the current interface and process, and some fields cannot be checked through the code stream or cannot be checked in the case of missing original code stream. Currently, artificial rule configuration or machine learning is usually used to solve the above problems, but the former has the problem of high labor cost, and the latter has the problem of poor interpretability and lack of convincing conclusions. SUMMARY

[0004] The present application provides a data verification knowledge graph model generation method, a data verification method and device to solve the technical problems existing in the prior art, improve the XDR data verification efficiency, and enhance the logicality of the verification result.

[0005] In order to solve the above technical problems, the present application is implemented as follows:

[0006] In a first aspect, the present application embodiment provides a data verification knowledge graph model generation method, comprising:

[0007] According to the field name in the knowledge graph construction table, the first information corresponding to the field name is extracted from the original code stream and the XDR data, wherein the knowledge graph construction table comprises a plurality of field names preset according to a training scene;

[0008] According to the first information, a plurality of knowledge graph nodes are formed;

[0009] According to all knowledge graph nodes and the first information, the relationship between the all knowledge graph nodes is learned;

[0010] According to the relationship between the all knowledge graph nodes, a knowledge graph model is generated.

[0011] Optionally, before extracting first information corresponding to a field name in the knowledge graph construction table from the original code stream and XDR data, the method further comprises:

[0012] analyzing the original code stream to obtain an analyzed original code stream, the analyzed original code stream being in xml format;

[0013] analyzing the XDR data to obtain an analyzed XDR data, the analyzed XDR data being in csv format.

[0014] Optionally, the method further comprises:

[0015] unifying the field names of the same fields in the analyzed original code stream, the analyzed XDR data and the knowledge graph construction table according to a preset field name mapping table.

[0016] Optionally, after forming a plurality of knowledge graph nodes according to the first information, the method further comprises:

[0017] forming a constraint rule set of a knowledge graph node having a constraint rule according to the all knowledge graph nodes and the first information.

[0018] Optionally, learning relationships between the all knowledge graph nodes according to the all knowledge graph nodes and the first information comprises:

[0019] forming an association relationship of a knowledge graph node having a containing relationship;

[0020] forming an association relationship of a knowledge graph node having a triggering relationship;

[0021] forming an association relationship of a knowledge graph node having a constraint relationship.

[0022] In a second aspect, an embodiment of the present application provides a data verification method, comprising:

[0023] obtaining a knowledge graph model, the knowledge graph model being a knowledge graph model generated by the data verification knowledge graph model generation method provided by an embodiment of the present application;

[0024] obtaining XDR data of a target to be verified;

[0025] extracting second information corresponding to a field name in the knowledge graph model from at least the XDR data of the target to be verified, wherein the second information comprises an information set of each field name;

[0026] generating an actual knowledge graph according to the second information and the knowledge graph model node;

[0027] analyzing values of each node in the actual knowledge graph according to the knowledge graph model, and generating a verification result of the target to be verified according to an analysis result.

[0028] Optionally, the obtaining of the second information further includes:

[0029] If the original code stream of the target to be verified can be obtained, the second information corresponding to the field name in the knowledge graph model is extracted from the original code stream of the target to be verified and the XDR data of the target to be verified according to the field name.

[0030] Optionally, before the second information corresponding to the field name in the knowledge graph model is extracted from the original code stream of the target to be verified and the XDR data of the target to be verified according to the field name, the method further includes:

[0031] parsing the XDR data to obtain parsed XDR data, the parsed XDR data being in a csv format;

[0032] parsing the original code stream to obtain parsed original code stream, the parsed original code stream being in an xml format.

[0033] Optionally, the analyzing of the values of each node in the actual knowledge graph according to the knowledge graph model, and the generating of the verification result of the target to be verified according to the analysis result include:

[0034] If the actual knowledge graph and the knowledge graph model are compared, and any of the following situations exists, there is a node conflict:

[0035] the number of knowledge graph node categories is inconsistent;

[0036] constraint relationships between corresponding knowledge graph nodes are different;

[0037] corresponding knowledge graph nodes do not conform to constraint rules.

[0038] Optionally, the analyzing of the values of each node in the actual knowledge graph according to the knowledge graph model, and the generating of the verification result of the target to be verified according to the analysis result further include:

[0039] If the node conflict exists, the original code stream of the target to be verified is verified, and field information recorded in the original code stream of the target to be verified is used as a criterion.

[0040] In a third aspect, an embodiment of the present application provides a data verification knowledge graph model generation device, the model generation device comprising:

[0041] an information extraction module configured to extract first information corresponding to a field name from the original code stream and the XDR data according to the knowledge graph construction table, wherein the knowledge graph construction table comprises a plurality of field names preset according to a training scenario;

[0042] a knowledge graph analysis module configured to form a plurality of knowledge graph nodes according to the first information;

[0043] a knowledge graph training module configured to learn relationships between all knowledge graph nodes according to the all knowledge graph nodes and the first information;

[0044] a knowledge graph generation module configured to generate a knowledge graph model according to the relationships between the all knowledge graph nodes.

[0045] In a fourth aspect, an embodiment of the present application provides a data verification device, comprising:

[0046] a model acquisition module configured to acquire a knowledge graph model, wherein the knowledge graph model is generated by the data verification knowledge graph model generation method provided by an embodiment of the present application;

[0047] a data acquisition module configured to acquire XDR data of a target to be verified;

[0048] an information extraction module configured to extract second information corresponding to the field name from the XDR data of the target to be verified, wherein the second information comprises an information set of each field name;

[0049] a knowledge graph generation module configured to generate an actual knowledge graph according to the second information and the knowledge graph model nodes;

[0050] a reasoning analysis module configured to analyze values of nodes in the actual knowledge graph according to the knowledge graph model, and generate a verification result of the target to be verified according to an analysis result.

[0051] In a fifth aspect, an embodiment of the present application provides a network device, comprising a processor, a memory, and a program stored in the memory and capable of running on the processor, wherein the program is executed by the processor to implement steps of the data verification knowledge graph model generation method of the first aspect or the data verification method of the second aspect.

[0052] In a sixth aspect, an embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the data verification knowledge graph model generation method in the first aspect or the steps of the data verification method in the second aspect are implemented.

[0053] The technical scheme provided by the embodiment of the present application has at least the following beneficial effects: the embodiment of the present application provides a data verification method, which obtains a knowledge graph model, obtains XDR data of a target to be verified, extracts second information corresponding to a field name in the knowledge graph model from the XDR data of the target to be verified, generates an actual knowledge graph according to the second information and nodes of the knowledge graph model, analyzes values of the nodes in the actual knowledge graph according to the knowledge graph model, and generates a verification result of the target to be verified according to an analysis result. The knowledge graph model is generated by any one of the data verification knowledge graph model generation methods provided by the embodiment of the present application. The XDR verification is performed through the knowledge graph of the target to be verified, the verification result is obtained through the relationship between the nodes, the logicality and the explainability of the verification result are improved, and the artificial maintenance cost is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0054] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not intended to limit the scope of the application. Moreover, like reference numerals designate similar parts throughout the several views in the drawings. In the drawings:

[0055] Figure 1 A data verification knowledge graph model generation method flowchart is provided for the embodiment of the present application;

[0056] Figure 2 A data verification knowledge graph model generation method schematic diagram is provided for the embodiment of the present application;

[0057] Figure 3 A knowledge graph model schematic diagram is provided for the embodiment of the present application;

[0058] Figure 4 A data verification method flowchart is provided for the embodiment of the present application;

[0059] Figure 5 A data verification knowledge graph model generation device structural schematic diagram is provided for the embodiment of the present application;

[0060] Figure 6A structural schematic diagram of a data verification method device provided by an embodiment of the present application is provided.

[0061] Figure 7 A structural schematic diagram of a network device provided by an embodiment of the present application is provided.

[0062] Figure 8 A structural schematic diagram of a data verification system based on a knowledge graph provided by an embodiment of the present application is provided. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0064] Please refer to Figure 1 , Figure 1 A flow chart of a data verification knowledge graph model generation method provided by an embodiment of the present application is shown, and the method comprises:

[0065] Step 11: According to the field name in the knowledge graph construction table, the first information corresponding to the field name is extracted from the original code stream and the XDR data, wherein the knowledge graph construction table comprises a plurality of field names preset according to the training scene.

[0066] Specifically, in the training scene, a plurality of field names can be preset based on the training scene, a knowledge graph construction table is created, and a model training device, equipment or system, etc. receives the knowledge graph construction table, the original code stream of the training scene and the XDR of the training scene, and performs data analysis and model training on these, for outputting the knowledge graph model in the later stage.

[0067] Specifically, the knowledge graph construction table can contain a certain field name, and the associated information related to the certain field name that needs to be extracted. For example, the field name is “measurement report”, and the information associated with the field name “measurement report” in the knowledge graph construction table includes: the measured service cell, frequency, RSRP, etc. For example, the first information about the field name “measurement report” is a set composed of all associated information about “measurement report” extracted from the original code stream and the XDR data.

[0068] Step 12: According to the first information, a plurality of knowledge graph nodes are formed.

[0069] Specifically, by step 11, the first information corresponding to a plurality of field names can be obtained based on the knowledge graph construction table, the original code stream and the XDR data, such as the first information of the field name "measurement report", the first information of the field name "measurement configuration", the first information of the field name "message type", and the like.

[0070] Specifically, by the first information, a plurality of nodes of the knowledge graph can be generated, such as generating respective node labels according to the field names and the associated information, such as generating nodes of "measurement report", "measurement cell", "frequency point", "measurement configuration", and the like.

[0071] Step 13, learning the relationships between all the knowledge graph nodes according to all the knowledge graph nodes and the first information.

[0072] Specifically, in step 12, the respective nodes of the knowledge graph are generated, and at this time, the generated knowledge graph nodes are generated according to the information set about the respective field names in the first information, and the relationships between the nodes are not clear, only the respective nodes existing in the knowledge graph are generated, and the nodes generated by the first information corresponding to different field names do not form relationships between them; by the first information, the relationships between all the knowledge graph nodes are learned, so as to obtain the relationships between all the knowledge graph nodes; these relationships can be a containing relationship, a superior-inferior attribute relationship, a constraint relationship, a numerical correlation relationship, a logical attribute relationship, and the like.

[0073] Step 14, generating a knowledge graph model according to the relationships between all the knowledge graph nodes.

[0074] According to the relationships between the knowledge graph nodes learned, a complete knowledge graph model can be obtained, and the specific relationships between the nodes can be between 2 nodes, between 1-N nodes (such as in a containing relationship, 1 parent node contains N child node types), between N-1 nodes, or between N-N nodes, such as there can be a triggering relationship between two nodes, for example, the node "user UE" triggers the "RRC establishment" node field through the RRC connection action; there can also be a constraint relationship, for example, the value of the node "service cell ID" in the node "service cell ID" contained in the node "measurement configuration" and the node "service cell ID" contained in the node "measurement report" must be the same;

[0075] For example, in the containing relationship of 1-N nodes, such as the node "measurement report configuration (field name ReportConfigToAdd)", which contains the node "reporting event type (such as eventA2)" and the node "measurement value (such as rsrp)", and the like, in the RRC reconfiguration process.

[0076] For example, referring to Figure 2The application discloses a data checking knowledge graph model generation method and device, and a knowledge graph model generation method and device.

[0077] Optionally, the method further comprises:

[0078] The original code stream is parsed to obtain parsed original code stream, and the parsed original code stream is in xml format.

[0079] The XDR data is parsed to obtain parsed XDR data, and the parsed XDR data is in csv format.

[0080] Specifically, the original code stream is processed, so that the original code stream can be converted into a format convenient for reading data, for example, xml format can facilitate field data query and extraction; the XDR data is processed, so that the XDR data can be converted into a format convenient for reading data, for example, csv format can facilitate field data query and extraction.

[0081] Optionally, the method further comprises:

[0082] The field names of the same fields in the parsed original code stream, the parsed XDR data and the knowledge graph construction table are unified according to a preset field name mapping table.

[0083] For example, table 1 is a field name mapping table of XDR data and indicates an intention.

[0084] Table 1: Field name mapping table of XDR data

[0085]

[0086] For example, table 2 is a field name mapping table of decoded original code stream and indicates an intention.

[0087] Table 2: Field name mapping table of decoded original code stream

[0088]

[0089] Note: XDR data and original code stream have fields related to knowledge graph, and the present application takes both as input sources, learns the types and relationships of the fields in the knowledge graph construction table, establishes the knowledge graph in the training scene, and establishes the knowledge graph in the reasoning scene (to be checked target). Among them, the same field name (for example, C-RNTI field in XDR and code stream in RRC_STP process) is the same field.

[0090] It is worth noting that the so-called "mapping table" means that the various data in the original code stream and / or XDR data and the specific field types in the knowledge graph construction table are corresponded, for example, the C-RNTI field in the above table can be obtained through the XDR or code stream related to RRC_STP.

[0091] The above mapping table is only a schematic, and in actual use, the content of the mapping table is adjusted according to the application scene, and the related data of one field name in the mapping table is corresponded.

[0092] Optionally, after forming a plurality of knowledge graph nodes according to the first information, it further includes:

[0093] According to all knowledge graph nodes and the first information, for the knowledge graph nodes with constraint rules, a constraint rule set of the knowledge graph nodes is formed.

[0094] For example, the constraint rule of node A is the value range (0-100), the constraint rule of node B is to contain the attribute of a certain type, and the constraint rule of node C is the position range belongs to a certain city.

[0095] For example, reference Figure 3 A knowledge graph model generated by the data verification knowledge graph model generation method provided by the embodiment of the present application is shown. If there is a relationship between each node in the knowledge graph, they are connected, and the relationship type between them is displayed on the connection line. Specifically, the connection line between the nodes containing the relationship directly displays the containing relationship, and the connection line between the nodes triggering the relationship directly displays the type of the triggering action, such as "reconfiguration", "RRC access", "measurement", etc. If there is a constraint relationship between the nodes, the connection line relationship may not be directly displayed. Only when the user selects the node, the associated constraint node or constraint rule is displayed.

[0096] Optionally, learning the relationships between all knowledge graph nodes according to all knowledge graph nodes and the first information includes:

[0097] For the knowledge graph nodes with containing relationships, the associated relationships of the knowledge graph nodes with containing relationships are formed.

[0098] For the knowledge graph node with a trigger relationship, an association relationship of the knowledge graph node with the trigger relationship is formed.

[0099] For the knowledge graph node with a constraint relationship, an association relationship of the knowledge graph node with the constraint relationship is formed.

[0100] Specifically, the containing relationship is, for example, node A is contained in node B; or N nodes, for example, nodes BCDE are all contained in node A. The containing relationship can be a containing relationship between upper and lower nodes, or a containing relationship between attribute relationships, for example, the node "measurement report" contains the nodes "measured cell", "measurement configuration", "measurement data", "neighborhood of the measured cell", and the like. For example, the node "measurement parameter" contains various specific parameter types such as "frequency point" on the attribute.

[0101] Specifically, the trigger relationship, also known as the action relationship, for example, the "measurement relationship" node with a trigger relationship connected to the "user (UE)" node has a "configuration" trigger relationship between the two. For example, there is an "access" trigger relationship between the "user (UE)" node and the "serving cell" node. The trigger relationship refers to the action interaction between nodes, thereby generating a relationship. Such a trigger action can be a message prompt, access, configuration, measurement, and the like.

[0102] Specifically, the constraint relationship refers to the constraint between at least two nodes in terms of attributes, values, and logic. For example, the parameter relationship between the nodes contained in the "measurement parameter" node cannot have logical contradictions. For example, the parameters of the nodes "measurement frequency point" and "event" reported by the node "UE" in the node "measurement report" must be within the range of the measurement configuration. If the range is exceeded, there is a logical conflict. For example, the range of the node "RSRP" contained in the node "measurement report" must be 0-97, and the node "measurement frequency point" must be within the range of several enumerated values.

[0103] Reference Figure 4 It is shown that the embodiment of the application provides a data verification method flowchart, which comprises:

[0104] Step 41, acquiring a knowledge graph model, wherein the knowledge graph model is a knowledge graph model generated by the data verification knowledge graph model generation method provided by the embodiment of the application;

[0105] Step 42, acquiring XDR data of a target to be verified;

[0106] Step 43, extracting second information corresponding to a field name in the knowledge graph model from at least the XDR data of the target to be verified according to the field name, wherein the second information comprises an information set of each field name.

[0107] In step 44, an actual knowledge graph is generated according to the second information and the knowledge graph model node.

[0108] In step 45, the values of each node in the actual knowledge graph are analyzed according to the knowledge graph model, and a verification result of the to-be-verified target is generated according to the analysis result.

[0109] Specifically, the data verification method is based on the knowledge graph model provided by the embodiments of the present application, reasoning is performed, an actual knowledge graph is generated according to actual scene parameters to be verified, and the actual knowledge graph is compared and verified with the knowledge graph model, so that a verification result is obtained. Since the generation of the knowledge graph model follows the logical relationship between the nodes, a more accurate, more logical and more interpretable verification result can be obtained through this comparison method.

[0110] Optionally, the obtaining of the second information further includes:

[0111] If the original code stream of the to-be-verified target can be obtained, the second information corresponding to the field name in the knowledge graph model is extracted from the original code stream of the to-be-verified target and the XDR data of the to-be-verified target according to the field name.

[0112] Optionally, before the second information corresponding to the field name in the knowledge graph model is extracted from the original code stream of the to-be-verified target and the XDR data of the to-be-verified target according to the field name, the method further includes:

[0113] The XDR data is parsed to obtain parsed XDR data, and the parsed XDR data is in a csv format.

[0114] The original code stream is parsed to obtain parsed original code stream, and the parsed original code stream is in an xml format.

[0115] Optionally, the analyzing of the values of each node in the actual knowledge graph according to the knowledge graph model and the generating of the verification result of the to-be-verified target according to the analysis result include:

[0116] If the actual knowledge graph and the knowledge graph model are compared, any of the following situations exists, and there is a node conflict:

[0117] The number of knowledge graph node categories is inconsistent;

[0118] The constraint relationship between the corresponding knowledge graph nodes is different;

[0119] The corresponding knowledge graph nodes do not conform to the constraint rule.

[0120] Exemplarily, in an actual application scenario, for example, the interfaces of the to-be-checked target have only XDR data, without original code stream data, the knowledge graph model in the Figure 3 is taken as an example to generate an actual knowledge graph, at this time, the reconfiguration process and the measurement process are all without original code stream, but according to the node relationship in the knowledge graph model, the frequency points of the two should be consistent, and if the two are inconsistent in the actual knowledge graph, the system determines that there is a node conflict.

[0121] Optionally, the method further includes: analyzing values of each node in the actual knowledge graph according to the knowledge graph model, and generating a checking result of the to-be-checked target according to an analysis result.

[0122] If there is a node conflict, the original code stream of the to-be-checked target is verified, and field information recorded in the original code stream of the to-be-checked target is used as a criterion.

[0123] Exemplarily, in an actual application scenario, for example, the interfaces of the to-be-checked target have only XDR data, without original code stream data, the knowledge graph model in the Figure 3 is taken as an example to generate an actual knowledge graph, assuming that the reconfiguration process has original code stream and the measurement process has no original code stream, if the frequency points are inconsistent, since the reconfiguration process has original code stream to provide evidence, the system determines that the frequency point field in the XDR data of the measurement process is filled in error.

[0124] In summary, the embodiment of the present application provides a data checking method, which includes: acquiring a knowledge graph model; acquiring XDR data of a to-be-checked target; extracting second information corresponding to a field name in the knowledge graph model from decoded XDR data; generating an actual knowledge graph according to the second information and nodes of the knowledge graph model; analyzing values of each node in the actual knowledge graph according to the knowledge graph model, and generating a checking result of the to-be-checked target according to an analysis result; wherein the knowledge graph model is a knowledge graph model generated by any one of the data checking knowledge graph model generation methods provided by the embodiment of the present application; XDR checking is performed through a knowledge graph, the logicality and explainability of the checking result are improved, and the artificial maintenance cost is reduced.

[0125] For reference, Figure 5 the embodiment of the present application provides a data checking knowledge graph model generation device 50, which includes

[0126] an information extraction module 51, configured to extract first information corresponding to a field name in a knowledge graph construction table from original code stream and XDR data, wherein the knowledge graph construction table includes a plurality of field names preset according to a training scenario;

[0127] The knowledge graph analysis module 52 is configured to form a plurality of knowledge graph nodes according to the first information.

[0128] The knowledge graph training module 53 is configured to learn relationships between all knowledge graph nodes according to the all knowledge graph nodes and the first information.

[0129] The knowledge graph generation module 54 is configured to generate a knowledge graph model according to the relationships between the all knowledge graph nodes.

[0130] Optionally, the knowledge graph analysis module 52 is further configured to, before extracting the first information corresponding to the field name from the original code stream and the XDR data according to the field name in the knowledge graph construction table:

[0131] analyze the original code stream to obtain an analyzed original code stream, the analyzed original code stream being in an xml format;

[0132] analyze the XDR data to obtain an analyzed XDR data, the analyzed XDR data being in a csv format.

[0133] Optionally, the knowledge graph analysis module 52 is further configured to:

[0134] unify the field names of the same fields in the analyzed original code stream, the analyzed XDR data, and the knowledge graph construction table according to a preset field name mapping table.

[0135] Optionally, the knowledge graph training module 53 is further configured to:

[0136] for a knowledge graph node that exists a constraint rule, form a constraint rule set of the knowledge graph node according to the all knowledge graph nodes and the first information.

[0137] Optionally, the knowledge graph training module 53 is further configured to:

[0138] for a knowledge graph node that exists a containing relationship, form an association relationship of the knowledge graph node that exists the containing relationship.

[0139] for a knowledge graph node that exists a trigger relationship, form an association relationship of the knowledge graph node that exists the trigger relationship.

[0140] for a knowledge graph node that exists a constraint relationship, form an association relationship of the knowledge graph node that exists the constraint relationship.

[0141] The data verification knowledge graph model generation apparatus 50 can realize each process of the embodiment of the data verification knowledge graph model generation method provided by the application and achieve the same technical effects. To avoid repetition, details are not described herein.

[0142] Please refer to Figure 6 The embodiment of the application provides a data verification apparatus 60, comprising:

[0143] A model acquisition module 61 is configured to acquire a knowledge graph model, wherein the knowledge graph model is a knowledge graph model generated by any one of the data verification knowledge graph model generation methods provided by the embodiments of the application.

[0144] A data acquisition module 62 is configured to acquire XDR data of a target to be verified.

[0145] An information extraction module 63 is configured to extract second information corresponding to a field name in the knowledge graph model from at least the XDR data of the target to be verified, wherein the second information comprises a set of information of each field name.

[0146] A knowledge graph generation module 64 is configured to generate an actual knowledge graph according to the second information and nodes of the knowledge graph model.

[0147] A reasoning analysis module 65 is configured to analyze values of each node in the actual knowledge graph according to the knowledge graph model, and generate a verification result of the target to be verified according to an analysis result.

[0148] Optionally, the information extraction module 63 is further configured to:

[0149] If the original code stream of the target to be verified can be acquired, the second information corresponding to the field name in the knowledge graph model is extracted from the original code stream of the target to be verified and the XDR data of the target to be verified.

[0150] Optionally, before the information extraction module 63 extracts the second information corresponding to the field name in the knowledge graph model from the original code stream of the target to be verified and the XDR data of the target to be verified, the information extraction module 63 is further configured to:

[0151] The XDR data is parsed to acquire parsed XDR data, wherein the parsed XDR data is in a csv format.

[0152] The original code stream is parsed to acquire parsed original code stream, wherein the parsed original code stream is in an xml format.

[0153] Optionally, the knowledge graph model is used to analyze the values of the nodes in the actual knowledge graph, and a verification result of the target to be verified is generated according to an analysis result.

[0154] If the actual knowledge graph and the knowledge graph model are compared, and any of the following conditions exists, a node conflict exists:

[0155] The number of knowledge graph node categories is inconsistent.

[0156] The constraint relationship between the corresponding knowledge graph nodes is different.

[0157] The corresponding knowledge graph nodes do not conform to the constraint rules.

[0158] Optionally, the knowledge graph model is used to analyze the values of the nodes in the actual knowledge graph, and a verification result of the target to be verified is generated according to an analysis result.

[0159] If the node conflict exists, the original code stream of the target to be verified is verified, and the field information recorded in the original code stream of the target to be verified is used as a reference.

[0160] The data verification device 60 can implement each process of the embodiment of the data verification method provided by the application and achieve the same technical effects. To avoid repetition, details are not described here.

[0161] The embodiment of the application also provides a knowledge graph-based data verification system 80, which can simultaneously implement the functions of the data verification knowledge graph model generation device 50 and the data verification device 60 provided by the embodiment of the application, and reference Figure 8 The knowledge graph-based data verification system 80 comprises:

[0162] A knowledge graph analysis module 81 is configured to construct a knowledge graph model based on a training scenario and an inference scenario (a target to be verified scenario).

[0163] An original code stream analysis module 82 is configured to parse the original code stream from a binary bit string into an xml format based on the training scenario and the inference scenario (the target to be verified scenario).

[0164] An XDR analysis module 83 is configured to parse XDR into a csv format based on the training scenario and the inference scenario (the target to be verified scenario).

[0165] An information extraction module 84 is configured to extract first information and second information of the original code stream and / or the original code stream based on the training scenario and the inference scenario (the target to be verified scenario).

[0166] The model training module 85 is configured to train a scene to generate a knowledge graph model.

[0167] The inference output module 86 is configured to infer a scene (a target scene to be checked) to generate a checking result.

[0168] It should be noted that Figure 7 The embodiments of the present application also provide a network device 70, which comprises a processor 71, a memory 72, and a computer program stored in the memory 72 and capable of running on the processor 71. When the computer program is executed by the processor 71, each process of the embodiments of the above data checking method or the data checking knowledge graph model generation method is implemented, and the same technical effects can be achieved. To avoid repetition, details are not described herein.

[0169] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program. When the computer program is executed by a processor, each process of the above checking method or the data checking knowledge graph model generation method is implemented, and the same technical effects can be achieved. To avoid repetition, details are not described herein. The computer readable storage medium may, for example, be a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0170] It should be noted that, in this document, the terms “comprising” and “including” or any other variant thereof are intended to cover non-exclusive inclusions, such that processes, methods, articles or devices including a series of elements not only include those elements, but also include other elements not explicitly listed, or include elements inherent to such processes, methods, articles or devices. Without more limitations, the element defined by the statement “comprising a” does not exclude the presence of another identical element in the process, method, article or device comprising the element.

[0171] From the above description of the embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, an optical disk) and includes a plurality of instructions for causing a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the methods described in the embodiments of the present application.

[0172] The embodiments of the present application are described above with reference to the accompanying drawings, but the present application is not limited to the above-described specific embodiments, and the above-described specific embodiments are merely illustrative, but not restrictive, and a person of ordinary skill in the art can make many forms under the inspiration of the present application without departing from the purpose of the present application and the scope protected by the claims, and all of them belong to the protection of the present application.

Claims

1. A data verification method, characterized in that, include: Obtain a knowledge graph model; Obtain the XDR data of the target to be verified; Based on the field names in the knowledge graph model, at least the second information corresponding to the field names is extracted from the XDR data of the target to be verified, wherein the second information includes a collection of information for each field name; Based on the second information and the knowledge graph model nodes, an actual knowledge graph is generated; Based on the knowledge graph model, the values ​​of each node in the actual knowledge graph are analyzed, and the verification results of the target to be verified are generated based on the analysis results.

2. The data verification method according to claim 1, characterized in that, Also includes: If the original bitstream of the target to be verified can be obtained, then the second information corresponding to the field name is extracted from the original bitstream of the target to be verified and the XDR data of the target to be verified according to the field name in the knowledge graph model.

3. The data verification method according to claim 2, characterized in that, Before extracting the second information corresponding to the field name from the original bitstream of the target to be verified and the XDR data of the target to be verified based on the field name in the knowledge graph model, the method further includes: The XDR data is parsed to obtain parsed XDR data, which is in CSV format. The original bitstream is parsed to obtain the parsed original bitstream, which is in XML format.

4. The data verification method according to any one of claims 1 to 3, characterized in that, The step of analyzing the values ​​of each node in the actual knowledge graph based on the knowledge graph model, and generating the verification result of the target to be verified based on the analysis results, includes: If, when comparing the actual knowledge graph and the knowledge graph model, any of the following situations exist, then a node conflict exists: The number of different types of nodes in the knowledge graph is inconsistent. The corresponding constraint relationships between knowledge graph nodes are different; The corresponding knowledge graph nodes do not conform to the constraint rules.

5. The data verification method according to claim 4, characterized in that, The step of analyzing the values ​​of each node in the actual knowledge graph based on the knowledge graph model, and generating the verification result of the target to be verified based on the analysis result, further includes: If node conflicts exist, the original bitstream of the target to be verified shall be verified, and the field information recorded in the original bitstream of the target to be verified shall prevail.

6. The data verification method according to claim 4, characterized in that, Based on the field names in the knowledge graph construction table, first information corresponding to the field names is extracted from the original bitstream and XDR data, wherein the knowledge graph construction table includes multiple field names preset according to the training scenario; Based on the first information, multiple knowledge graph nodes are formed; Based on all knowledge graph nodes and the first information, learn the relationships between all knowledge graph nodes; A knowledge graph model is generated based on the relationships between all the knowledge graph nodes.

7. The data verification method according to claim 6, characterized in that, Before obtaining the knowledge graph model, the process also includes: Generating the knowledge graph model includes: Before extracting the first information corresponding to the field name from the original bitstream and XDR data based on the field name in the table constructed according to the knowledge graph, the method further includes: The original bitstream is parsed to obtain the parsed original bitstream, which is in XML format; The XDR data is parsed to obtain parsed XDR data, which is in CSV format.

8. The data verification method according to claim 7, characterized in that, The method further includes: Based on the preset field name mapping table, the field names of the same fields in the parsed original code stream, the parsed XDR data, and the knowledge graph construction table are unified.

9. The data verification method according to claim 6, characterized in that, After forming multiple knowledge graph nodes based on the first information, the process further includes: Based on all the knowledge graph nodes and the first information, a set of constraint rules is formed for the knowledge graph nodes that have constraint rules.

10. The data verification method according to claim 6, characterized in that, The step of learning the relationships between all knowledge graph nodes based on all knowledge graph nodes and the first information includes: For knowledge graph nodes that have an inclusion relationship, an association relationship is formed between the knowledge graph nodes that have an inclusion relationship; For knowledge graph nodes that have triggering relationships, an association relationship is formed between the knowledge graph nodes that have triggering relationships; For knowledge graph nodes that have constraints, an association relationship is formed between the knowledge graph nodes that have constraints.

11. A data verification device, characterized in that, include: The model acquisition module is used to acquire knowledge graph models; The data acquisition module is used to acquire the XDR data of the target to be verified; The information extraction module is used to extract second information corresponding to the field name from the XDR data of the target to be verified, based on the field name in the knowledge graph model, wherein the second information includes an information set for each field name; The knowledge graph generation module is used to generate an actual knowledge graph based on the second information and the knowledge graph model nodes. The reasoning and analysis module is used to analyze the values ​​of each node in the actual knowledge graph based on the knowledge graph model, and generate the verification result of the target to be verified based on the analysis result.

12. A network device, characterized in that, include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the data verification method as described in any one of claims 1 to 10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the data verification method as described in any one of claims 1 to 10.

Images