Resource fusion method, anomaly analysis method, device, equipment and medium

Abstract

The invention provides a resource fusion method, an anomaly analysis method, an anomaly analysis device, equipment and a medium, a semantic network technology is adopted, a unified semantic ontology model fusing power grid and communication network service data resources is constructed based on power grid and communication network service data resource information, and then an anomaly analysis result is obtained based on the unified semantic ontology model. According to the method, service data extracted from a power grid and communication network service system is stored in a triple mode, a service knowledge graph library is constructed, and finally knowledge completion is performed on the service knowledge graph library, so that fusion of power grid and communication network service data resources is realized, and a basis is provided for identification and analysis of resource abnormity.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a resource fusion method, anomaly analysis method, device, equipment and medium. Background Art

[0002] The power grid and communications network constitute two parallel and independent systems, each containing a wealth of complementary resources. These resources support each other at the business level, forming a complex network of connections. In this field, the amount of data generated is enormous. However, the lack of efficient data fusion technology makes it difficult to achieve deep integration and unified application of resources across these two systems. Therefore, despite the abundance of data resources, their potential value is difficult to effectively explore and utilize.

[0003] Secondly, given the complexity and breadth of services related to power grid and communication network resources, abnormal situations (such as resource failures, equipment downtime due to maintenance, and resource alarms) place higher demands on anomaly analysis. Currently, this process relies heavily on manual experience and the retrieval of large amounts of related data, which is not only inefficient but also lacks significant accuracy in problem location and analysis. Summary of the Invention

[0004] The present application proposes a resource fusion method, anomaly analysis method, device, equipment and medium, which can solve one of the problems existing in the background technology.

[0005] To achieve the above objectives, this application adopts the following technical solutions:

[0006] In a first aspect, a resource fusion method is provided, comprising:

[0007] Obtaining power grid business data resource information and communication network business data resource information;

[0008] Based on the power grid business data resource information and the communication network business data resource information, a unified semantic ontology model integrating the power grid business data resources and the communication network business data resources is constructed using semantic network technology, wherein the semantic ontology model defines domain knowledge concepts, object relationships between the domain knowledge concepts, and attribute relationships between the domain knowledge concepts themselves;

[0009] Based on the unified semantic ontology model, extracting business data from the power grid business system and the communication network business system respectively, and storing the business data in a triplet manner to construct a business knowledge graph library; and

[0010] Based on the characteristics of the unified semantic ontology model and the business characteristics of the domain, the business knowledge graph library is supplemented with knowledge.

[0011] Based on the above technical solution, semantic network technology is adopted to build a unified semantic ontology model that integrates the business data resources of the power grid and communication network based on the business data resource information of the power grid and communication network. Then, based on the unified semantic ontology model, the business data extracted from the power grid and communication network business systems are stored in a triplet manner, and a business knowledge graph library is built. Finally, the business knowledge graph library is supplemented with knowledge. In this way, the integration of the business data resources of the power grid and communication network is realized, thereby providing a basis for the identification and analysis of resource anomalies.

[0012] In a possible design method of the first aspect, the domain knowledge concept is defined by the following items: concept identifier, concept name and concept description, the object relationship is defined by the following items: relationship identifier, object relationship name, precondition concept, postcondition concept, transitivity, uniqueness, reflexivity and relationship description, the attribute relationship is defined by the following items: attribute identifier, attribute name, belonging concept and attribute value type, the triple is defined by the following items: subject entity, semantic relationship and object entity or object value, and the semantic relationship is an object relationship or an attribute relationship.

[0013] In a possible design method of the first aspect, based on the characteristics of the unified semantic ontology model, the business knowledge graph library is supplemented with knowledge, specifically: based on the object relationship and the attribute relationship in the unified semantic ontology model, knowledge reasoning is supplemented,

[0014] Based on the domain business characteristics, the business knowledge graph library is supplemented with knowledge, specifically: based on the usage relationship and ownership relationship represented by the domain business characteristics, knowledge reasoning and completion are performed.

[0015] In a second aspect, an anomaly analysis method is provided. The anomaly analysis method is based on the business knowledge graph library constructed as described above, and the anomaly analysis method includes:

[0016] Obtaining current abnormal resource entity information; and

[0017] Using the constructed knowledge graph embedding learning model, the resource entity results corresponding to the current abnormal resource entity are queried from the business knowledge graph library. The knowledge graph embedding learning model is obtained by training using the business knowledge graph library.

[0018] In a possible design manner of the second aspect, the knowledge graph embedding learning model determines the association between two nodes by calculating the cosine distance between the two nodes.

[0019] In a possible design manner of the second aspect, the resource entity result includes: a first resource entity set affected by the current abnormal resource entity, and a second resource entity set traced back from the current abnormal resource entity.

[0020] In a third aspect, a resource fusion device is provided, the resource fusion device comprising:

[0021] A first acquisition unit is used to obtain power grid business data resource information and communication network business data resource information;

[0022] A first construction unit is configured to construct, based on the power grid service data resource information and the communication network service data resource information, a unified semantic ontology model integrating the power grid service data resources and the communication network service data resources using semantic network technology, wherein the semantic ontology model defines domain knowledge concepts, object relationships between the domain knowledge concepts, and attribute relationships between the domain knowledge concepts themselves;

[0023] A second construction unit is configured to extract business data from the power grid business system and the communication network business system respectively based on the unified semantic ontology model, and store the business data in a triplet manner to construct a business knowledge graph library; and

[0024] The completion unit is used to complete the business knowledge graph library based on the characteristics of the unified semantic ontology model and the domain business characteristics.

[0025] In a fourth aspect, an anomaly analysis device is provided. The anomaly analysis device is based on the business knowledge graph library constructed as described above, and the anomaly analysis device includes:

[0026] A second acquiring unit is configured to acquire current abnormal resource entity information; and

[0027] The query unit is used to use the constructed knowledge graph embedding learning model to query and obtain the resource entity results corresponding to the current abnormal resource entity from the business knowledge graph library. The knowledge graph embedding learning model is obtained by training using the business knowledge graph library.

[0028] In a fifth aspect, an electronic device is provided, comprising: a processor, and a memory coupled to the processor, the memory being used to store a computer program; the processor being used to execute the computer program stored in the memory, so that the electronic device executes the resource fusion method as any possible implementation method in the first aspect, or executes the anomaly analysis method as any possible implementation method in the second aspect.

[0029] In a sixth aspect, a computer-readable storage medium is provided, comprising a computer program or instructions, which, when executed on a computer, enables the computer to execute the resource fusion method as described in any possible implementation of the first aspect, or to execute the anomaly analysis method as described in any possible implementation of the second aspect.

[0030] In the seventh aspect, a computer program product is provided, comprising: a computer program or instructions, which, when the computer program or instructions are run on a computer, enables the computer to execute the resource fusion method as any possible implementation method in the first aspect, or execute the anomaly analysis method as any possible implementation method in the second aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0032] Figure 1 This is a flowchart of the steps for implementing the method provided in the embodiment of the present application;

[0033] Figure 2 This is a semantic ontology model diagram of the integration of power grid and communication network provided by the embodiment of the present application;

[0034] Figure 3 This is a diagram of the impact relationship between the integrated power grid and communication network resources provided in an embodiment of the present application;

[0035] Figure 4 This is the overall framework diagram provided by the embodiment of this application. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0037] It should be noted that although the device schematics illustrate functional module divisions and the flowcharts illustrate logical sequences, in certain circumstances, the steps shown or described may be performed in a sequence that differs from the module divisions in the device or the sequence in the flowcharts. The terms "first," "second," and so on, in the specification, claims, and drawings, are used to distinguish similar items and are not necessarily used to describe a specific sequence or precedence.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0039] The objectives of the embodiments of this application are threefold:

[0040] (1) By adopting semantic network technology, a unified semantic network ontology model of power grid and communication network resources was constructed, and a real-time business knowledge graph was established simultaneously. This initiative aims to manage large-scale business data resources in a more efficient, intuitive and integrated manner, and achieve deep integration and optimized utilization of data.

[0041] (2) To improve the completeness and accuracy of the business knowledge graph, semantic network reasoning technology was used to complete the graph. At the same time, with the help of knowledge graph embedding learning methods, a vectorized representation of entity nodes in the business knowledge graph was achieved. This multi-dimensional vector representation can more accurately capture the characteristics between nodes. Furthermore, cosine distance was used to measure the association of entity nodes. Compared with traditional large-scale data association retrieval methods, this calculation method significantly improves the efficiency of analysis and calculation.

[0042] (3) Based on the business knowledge graph library and the knowledge graph embedding model, a set of resource anomaly impact and traceability reasoning analysis methods are implemented. This method not only has efficient analysis capabilities, but also can accurately identify potential correlation issues, providing strong support for intelligent decision-making in resource anomaly scenarios, and helping to achieve more accurate and faster decision-making analysis.

[0043] The resource fusion method and the abnormality analysis method based on the resource fusion method in the embodiment of the present application are exemplarily described below with reference to an application example. Figure 1 and Figure 4 shown.

[0044] (1) Constructing a unified semantic ontology model of power grid and communication network resources based on the integration of semantic network technology

[0045] Using semantic network technology to build a unified and integrated semantic ontology model for power grid and communication network resources: This mainly includes defining and constructing domain knowledge concepts, object relationships between concepts, and describing the attribute relationships of the concepts themselves. In combination with the implementation requirements of this method and the large amount of business data resource information of the power grid and communication network, the domain knowledge concepts are defined as shown in Table 1 below:

[0046] Table 1 Definition of domain concepts for the integration of power grid and communication network

[0047]

[0048] As shown in the table above, a total of 14 core domain concepts are defined. By analyzing the business data models of power grid and communication network resources and the relationships between data, the object relationships between concepts are defined as shown in Table 2 below:

[0049] Table 2 Definition of conceptual object relationships for the integration of power grid and communication network

[0050]

[0051] As shown in the table above, this method defines a total of 16 types of object relationships between domain concepts. In combination with the construction requirements of semantic ontology model technology, the directionality of object relationships is defined by clarifying the preceding concept (Domain) and the following concept (Range); each object relationship is also configured with corresponding characteristics: transitivity, uniqueness, and reflexivity. For example, the "belongs to" relationship is a transitive object relationship. If A belongs to B, and B belongs to C, then through the transitivity of this relationship, it can be inferred that A belongs to C; the "parent slot" relationship is unique, indicating that any instance entity of a "slot" can only have one "parent slot" relationship at most; reflexivity represents the bidirectionality of object relationships. For example, the "association" relationship is a reflexive object relationship. If A is associated with B, then B is also associated with A.

[0052] In addition to describing the object relationships between domain concepts, it is also necessary to further describe the concepts themselves through attribute relationships. In combination with the requirements of this method, the common attribute relationships of each concept are mainly defined as shown in Table 3 below:

[0053] Table 3 Definition of attribute relationship of power grid and communication network integration

[0054]

[0055] As shown in the table above, a total of four attribute relationships are defined, among which the attribute relationships whose concept is "Thing" represent the attribute relationships common to all concepts.

[0056] Combining the above concepts with the definitions of semantic relationships (including object relationships and attribute relationships), a semantic ontology model diagram of the integration of power grid and communication network resources is constructed and displayed in a visual way, such as Figure 2 shown.

[0057] (2) Real-time collection of power grid and communication network business data to build a fused business knowledge graph

[0058] Combined with the semantic ontology model of the power grid and communication network integration built in the previous step, business data is extracted from the power grid and communication network related business systems respectively, and stored in the form of triples, thereby building a large-scale business knowledge graph library. In this knowledge graph library, each piece of knowledge is represented by<subject,predicate,object> The data is stored in triples. Subject represents the subject entity, predicate represents the semantic relationship (object relationship / attribute relationship), and object represents the object entity or value (when the predicate is an object relationship, object is the object entity; when the predicate is an attribute relationship, object is the object value).

[0059] The data collection method needs to follow certain rules and sequence. Priority should be given to collecting business object information corresponding to the concepts in the semantic ontology model to construct instance entities under the corresponding concepts, and then the relationship between these entities should be constructed based on the association relationships in the business data.

[0060] For example, to collect fiber core information, we obtain the basic data structure and information list of fiber core minimization from the communication network system as shown in Table 4 below:

[0061] Table 4 Basic data structure and information list of core minimization

[0062] Core ID name describe 10001 Fiber core No. 1 in section xx Not yet enabled 10002 xx section No. 2 fiber core Not yet enabled

[0063] After obtaining the basic data of the fiber core, we can construct instance entities corresponding to the "fiber core" concept, namely: 10001 and 10002.

[0064] However, considering that different business objects from multiple business systems may have duplicate business object IDs after being merged, which will lead to duplicate and non-unique entities under different concepts during the construction of the business knowledge graph, the present invention formulates a generation strategy for the ID of each instance entity in the business knowledge graph as: "[concept identifier]-[business id]", so even if the business object ids under different concepts are the same, as long as the concept identifiers are different, the uniqueness of the entity ID in the business knowledge graph can be guaranteed.

[0065] Combined with the entity ID generation strategy, the IDs of the two fiber core instance entities are: FIBERCORE-1001, FIBERCORE-1002. At the same time, the business data in the above table is converted into a business knowledge graph triple:

[0066] <FIBERCORE-1001,instanceOf,FIBERCORE>

[0067] <FIBERCORE-1001, name, "Fiber Core No. 1 of xx Section">

[0068] <FIBERCORE-1001, description, "Not yet enabled">

[0069] <FIBERCORE-1002, instanceOf_of, FIBERCORE>

[0070] <FIBERCORE-1002, name, "Fiber Core No. 2 of xx Section">

[0071] <FIBERCORE-1002, description, "Not yet enabled">

[0072] Among them, "instanceOf" represents an instance relationship, which comes from the OWL ontology modeling language. For example, the specific fiber core represented by "FIBERCORE-1001" is an instance of the "FIBERCORE (fiber core)" concept.

[0073] Similarly, referring to the above, we can collect the optical cable section information in the business system data table, as shown in Table 5 below:

[0074] Table 5 Optical Cable Section Information in the Business System Data Table

[0075] Cable segment ID name describe 012314 xx optical cable segment 15km 010203 yy optical cable segment 24km

[0076] Combined with the example data in the above table, two instance entities corresponding to the "optical cable section" concept can be constructed, namely: FIBERCABLESEG-012314, FIBERCABLESEG-010203. Similarly, the property triple knowledge of these two entities can also be constructed.

[0077] After completing the construction of the instance entities of "fiber core" and "optical cable section", the knowledge between these two types of entities can be further constructed by combining the semantic ontology model and the association relationship in the business data. For example, the example data of the association relationship between the fiber core and the optical cable section obtained from the business system is as shown in Table 6 below:

[0078] Table 6 Example Data of the Association Relationship between the Fiber Core and the Optical Cable Section

[0079] Core ID Optical cable segment ID 10001 012314 10002 012314

[0080] Based on the above instance data, combined with the "belong_to" object relationship in the semantic ontology model, the triple knowledge of the fiber core entity and the optical cable section entity can be constructed as follows:

[0081] <FIBERCORE-1001,belong_to,FIBERCABLESEG-012314>

[0082] <FIBERCORE-1002,belong_to,FIBERCABLESEG-012314>

[0083] To sum up, combined with the above examples and methods, this method can complete the construction of a large-scale business knowledge graph based on the constructed semantic ontology model and combined with the business data of the power grid and communication network. The business knowledge graph is stored in the form of triples.

[0084] (3) Based on semantic network technology and knowledge graph embedding method, business knowledge graph completion and knowledge graph embedding representation learning are realized respectively

[0085] After the above two steps, we can obtain a large-scale business knowledge graph that integrates power grid and communication network resources. In order to further apply the knowledge graph for reasoning and analysis, it is necessary to complete the knowledge graph.

[0086] This method completes knowledge from the perspectives of semantic ontology model characteristics and domain business characteristics.

[0087] a. Knowledge completion based on the characteristics of semantic ontology model:

[0088] In the semantic network ontology model constructed by this method, the object relationship "belong to" is transitive. Therefore, we can perform knowledge completion based on this feature. The completion rules are:

[0089] like<a,belong_to,b> and<b,belong_to,c> Then we can infer that<a,belong_to,c>

[0090] For example, if chassis a belongs to network element b, and network element b belongs to site c, then we can infer that chassis a also belongs to site c. Similarly, we can combine the characteristics of this relationship to perform knowledge reasoning and completion on similar semantic relationships in the constructed knowledge graph.

[0091] b. Knowledge completion based on domain business characteristics:

[0092] Combining domain business characteristics can be used to complete knowledge and build "relate_with" relationships between knowledge graph entities:

[0093] For example: complete the "relate_with" relationship between the "business" entity and other conceptual entities.

[0094] like Figure 2The semantic model in the data structure builds the relationship between services and channels, but does not build the direct relationship between services and resources such as routing, ports, and boards. Completion rules can be built.

[0095] If business a uses channel b, and channel b uses route c, then the knowledge of business a and route c can be completed;

[0096] Furthermore, if route c uses port d, then the knowledge can be completed: service a is associated with port d;

[0097] Furthermore, if port d belongs to board e, then the knowledge can be completed: service a is associated with board e;

[0098] Furthermore, if board e belongs to slot f, then the knowledge can be completed: business a is associated with slot f;

[0099] Furthermore, if slot f belongs to chassis g, then the knowledge can be completed: service a is associated with chassis g;

[0100] Furthermore, if chassis g belongs to network element f, then the knowledge can be completed: service a is associated with network element f;

[0101] Furthermore, if network element f belongs to site h, then the knowledge can be completed: service a is associated with site h.

[0102] …

[0103] Similarly, combining the business characteristics of power grid and communication network resources can complete and enrich the relationships between various entities.

[0104] After completing the knowledge graph reasoning and completion, the knowledge graph network structure and scale will become more complex and large. Considering that the spatial complexity of graph-based search will also increase, it will seriously affect the efficiency of subsequent reasoning and analysis. This method uses the knowledge graph embedding learning method (Node2Vec) to represent the nodes in the knowledge graph. Knowledge graph embedding is a model that uses supervised learning to learn embeddings and vector representations of nodes and edges. They project "knowledge" into a continuous low-dimensional space. These low-dimensional space vectors generally have only a few hundred dimensions (used to represent the memory efficiency of knowledge storage). In the vector space, each point represents an entity in a knowledge graph, and the position of each point in space has semantic meaning. By calculating the cosine distance between two nodes, it is determined whether the two entities are related. Assuming two entity node vectors A and B are given, the cosine distance between them is calculated as follows:

[0105]

[0106] Here, (A·B) represents the dot product of vectors A and B, and ||A|| and ||B|| represent the moduli (lengths) of vectors A and B, respectively. The cosine distance ranges from [0, 2]. When two vectors are identical, the cosine distance is 0; when the two vectors are in opposite directions, the cosine distance is 2; and when the two vectors are perpendicular, the cosine distance is 1. Therefore, if the cosine distance between two entities is between 0 and 1, they are related; the closer it is to 0, the stronger the relationship.

[0107] This method uses Node2Vec to iteratively train and verify the completed knowledge graph to obtain the embedding model of the knowledge graph, which is expressed as KGEV. From this embedding model, each node in the knowledge graph has a corresponding vector, expressed as V entity .

[0108] (4) Combine the constructed business knowledge graph and its embedded model to realize the impact and traceability analysis method of abnormal power grid and communication network resources

[0109] Method input:

[0110] resId: abnormal resource ID, resource service ID in the power grid and communication network system.

[0111] resType: resource type (the value is the concept identifier in Table 1, such as BUSINESS, STATION, CHANNEL, etc.).

[0112] threshold: cosine distance threshold, the default value is 0.5. If it is less than or equal to the threshold, it is considered that there is a correlation between the two resources, otherwise there is no correlation.

[0113] Output of the method:

[0114] Influenced resource entity collection: influenceRes is empty by default

[0115] Traceable resource entity collection: tracedRes is empty by default

[0116] result: (success – represents success / fail – represents failure)

[0117] msg: (prompt message, usually provided when result is fail)

[0118] Method implementation steps:

[0119] Before describing the method implementation steps, we first define some variables and methods:

[0120] The constructed business knowledge graph is defined as KG; the model trained based on the knowledge graph embedding method is KGEV;

[0121] Method definition:

[0122] (1) Method query to obtain the specified entity from the knowledge graph:

[0123] entity = query(KG,entityId)

[0124] Where entityId is the given entity id, and KG represents the current business knowledge graph library.

[0125] (2) According to the given entity, the method of obtaining the resource entity set entities whose cosine distance is less than or equal to the threshold from the knowledge graph embedding model is:

[0126] entities=extract(entity,threshold,KGEV)

[0127] Where entity is the current entity; threshold is the cosine distance threshold; KGEV is the trained knowledge graph embedding model. The main implementation of this method is to calculate the cosine distance (V) between the entity entity (entity∈KG) and any other entity e (e∈KG) entity ,V e ), if the distance <= threshold, then e is included in the set entities.

[0128] (3) Based on the given resource entity and its associated resource collection entities, combined with Figure 3 The resource impact relationship diagram shown defines the resource entity collection acquisition method influence affected by the resource entity entity, and the traceable resource entity collection acquisition method trace as follows:

[0129] influenceEntities=influence(entity,entities)

[0130] tracedEntities=trace(entity,entities)

[0131] That is Figure 3 As shown in the figure, if the currently given resource entity is a fiber core resource, then the resource entities it affects can be optical paths, channels, services, and unit entities; and the resource entities it can trace can be ports, boards, slots, chassis, network elements, optical cable segments, and site entities.

[0132] Based on the above definition, the analysis method is implemented as follows:

[0133] Step 1: Initialize variables: influencedRes = {}; tracedRes = {}; result = 'fail'; msg = "";

[0134] Step 2: Construct entity ID: According to the entity ID generation rule "[concept identifier]-[business id]" in the knowledge graph, construct the entity ID:

[0135] entityId=resType+"-"+resId

[0136] Step 3: Entity query: Retrieve the corresponding entity object from the knowledge graph based on the obtained entity ID.

[0137] entity = query(KG,entityId)

[0138] If the entity object obtained is empty, it means that the resource cannot be found in the knowledge graph, the method execution fails, msg = "Unable to find the specified entity", and the method ends; otherwise, the method continues.

[0139] Step 4: Get the associated resource set: According to the knowledge graph embedding model, search for the entity set whose cosine distance is within the threshold (the default is 0.5).

[0140] entities=extract(entity,threshold,KGEV)

[0141] Step 5: Distinguish the affected and traceable physical resources: Locate the current resource node based on the resource's impact path.

[0142] influencedRes=influence(entity,entieis)

[0143] tracedRes=trace(entity,entities)

[0144] Step 6: Method output: Set result to "success" and the method outputs influencedRes, tracedRes, result, and msg.

[0145] In summary, the method pseudocode is described as follows:

[0146]

[0147] This embodiment has at least the following beneficial effects:

[0148] (1) It realizes the unified management and efficient integration of power grid and communication network resources, providing basic support for resource sharing and subsequent intelligent analysis.

[0149] (2) An efficient resource anomaly analysis method is implemented, including impact analysis and traceability analysis; at the same time, the method has the ability to identify potential correlation problems between resources (by adjusting the cosine distance threshold).

[0150] (3) It can effectively improve the information sharing and operation and maintenance efficiency of subsequent power grid and communication network resources, and can further provide more intelligent decision-making and analysis support.

[0151] The embodiment of the present application further provides a resource fusion device, the resource fusion device comprising:

[0152] A first acquisition unit is used to obtain power grid business data resource information and communication network business data resource information;

[0153] A first construction unit is configured to construct, based on the power grid service data resource information and the communication network service data resource information, a unified semantic ontology model integrating the power grid service data resources and the communication network service data resources using semantic network technology, wherein the semantic ontology model defines domain knowledge concepts, object relationships between the domain knowledge concepts, and attribute relationships between the domain knowledge concepts themselves;

[0154] A second construction unit is configured to extract business data from the power grid business system and the communication network business system respectively based on the unified semantic ontology model, and store the business data in a triplet manner to construct a business knowledge graph library; and

[0155] The completion unit is used to complete the business knowledge graph library based on the characteristics of the unified semantic ontology model and the domain business characteristics.

[0156] The present application also provides an anomaly analysis device, which is based on the business knowledge graph library constructed as described above, and includes:

[0157] A second acquiring unit is configured to acquire current abnormal resource entity information; and

[0158] The query unit is used to use the constructed knowledge graph embedding learning model to query and obtain the resource entity results corresponding to the current abnormal resource entity from the business knowledge graph library. The knowledge graph embedding learning model is obtained by training using the business knowledge graph library.

[0159] An embodiment of the present application also provides an electronic device, comprising: a processor, and a memory coupled to the processor, wherein the memory is used to store a computer program; and the processor is used to execute the computer program stored in the memory, so that the electronic device executes the method described in any one of the above embodiments.

[0160] The electronic device may be a computing device such as a desktop computer, a notebook computer, a PDA, a cloud server, etc. The electronic device may include, but is not limited to, a processor and a memory.

[0161] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, connecting various parts of the entire device using various interfaces and lines.

[0162] The memory may be used to store the computer program, and the processor implements various functions of the electronic device by running or executing the computer program stored in the memory and calling the data stored in the memory.

[0163] The memory may primarily include a program storage area and a data storage area, wherein the program storage area may store an operating system, at least one application required for a function, and the like; and the data storage area may store data created based on the use of the mobile phone, and the like. Furthermore, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a SmartMedia Card (SMC), a Secure Digital (SD) card, a flash memory card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0164] The embodiment of the present application also provides a storage medium, which is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0165] An embodiment of the present application further provides a computer program product, including: a computer program or instructions, which, when executed on a computer, causes the computer to execute any of the above-mentioned possible implementation methods.

[0166] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications are also considered to be within the scope of protection of the present application.

Claims

1. A resource fusion method, characterized in that: include: Obtaining power grid business data resource information and communication network business data resource information; Based on the power grid business data resource information and the communication network business data resource information, a unified semantic ontology model integrating the power grid business data resources and the communication network business data resources is constructed using semantic network technology, wherein the semantic ontology model defines domain knowledge concepts, object relationships between the domain knowledge concepts, and attribute relationships between the domain knowledge concepts themselves; Based on the unified semantic ontology model, business data is extracted from the power grid business system and the communication network business system respectively, and the business data is stored in a triple form to construct a business knowledge graph library; as well as Based on the characteristics of the unified semantic ontology model and the business characteristics of the domain, the business knowledge graph library is supplemented with knowledge.

2. The resource fusion method according to claim 1, characterized in that: The domain knowledge concept is defined by the following items: concept identifier, concept name and concept description; the object relationship is defined by the following items: relationship identifier, object relationship name, preceding concept, following concept, transitivity, uniqueness, reflexivity and relationship description; the attribute relationship is defined by the following items: attribute identifier, attribute name, belonging concept and attribute value type; the triple is defined by the following items: subject entity, semantic relationship and object entity or object value; the semantic relationship is an object relationship or an attribute relationship.

3. The resource fusion method according to claim 1, characterized in that: Based on the characteristics of the unified semantic ontology model, the business knowledge graph library is supplemented with knowledge, specifically: based on the object relationship and the attribute relationship in the unified semantic ontology model, knowledge reasoning and supplementation are performed, Based on the domain business characteristics, the business knowledge graph library is supplemented with knowledge, specifically: based on the usage relationship and ownership relationship represented by the domain business characteristics, knowledge reasoning and completion are performed.

4. An abnormality analysis method, characterized in that: The anomaly analysis method is based on the business knowledge graph library constructed according to any one of claims 1 to 3, and the anomaly analysis method includes: Obtaining current abnormal resource entity information; and Using the constructed knowledge graph embedding learning model, the resource entity results corresponding to the current abnormal resource entity are queried from the business knowledge graph library. The knowledge graph embedding learning model is obtained by training using the business knowledge graph library.

5. The abnormality analysis method according to claim 4, wherein: The knowledge graph embedding learning model determines the association between two nodes by calculating the cosine distance between the two nodes.

6. The abnormality analysis method according to claim 4, wherein: The resource entity result includes: a first resource entity set affected by the current abnormal resource entity, and a second resource entity set traced back from the current abnormal resource entity.

7. A resource fusion device, characterized in that: The resource fusion device includes: A first acquisition unit is used to obtain power grid business data resource information and communication network business data resource information; A first construction unit is configured to construct, based on the power grid service data resource information and the communication network service data resource information, a unified semantic ontology model integrating the power grid service data resources and the communication network service data resources using semantic network technology, wherein the semantic ontology model defines domain knowledge concepts, object relationships between the domain knowledge concepts, and attribute relationships between the domain knowledge concepts themselves; A second construction unit is configured to extract business data from the power grid business system and the communication network business system respectively based on the unified semantic ontology model, and store the business data in a triplet manner to construct a business knowledge graph library; and The completion unit is used to complete the business knowledge graph library based on the characteristics of the unified semantic ontology model and the domain business characteristics.

8. An abnormality analysis device, characterized in that: The anomaly analysis device is based on the business knowledge graph library constructed according to any one of claims 1 to 3, and the anomaly analysis device includes: A second acquiring unit is configured to acquire current abnormal resource entity information; and The query unit is used to use the constructed knowledge graph embedding learning model to query and obtain the resource entity results corresponding to the current abnormal resource entity from the business knowledge graph library. The knowledge graph embedding learning model is obtained by training using the business knowledge graph library.

9. An electronic device, characterized in that: The electronic device includes: a processor, and a memory coupled to the processor, The memory is used to store computer programs; and The processor is used to execute the computer program stored in the memory, so that the electronic device executes the resource fusion method according to any one of claims 1 to 3, or executes the anomaly analysis method according to any one of claims 4 to 6.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a computer program or instructions, which, when executed on a computer, enables the computer to execute the resource fusion method according to any one of claims 1 to 3, or the anomaly analysis method according to any one of claims 4 to 6.

Images