A knowledge supplement method and device for operation and maintenance knowledge graph

Abstract

The embodiment of the present application provides a method and device for supplementing knowledge of an operation and maintenance knowledge graph, which relates to the field of information processing technology. The method for supplementing knowledge of the operation and maintenance knowledge graph includes: creating an operation and maintenance knowledge graph; obtaining newly added node information; generating a newly added node knowledge graph based on the newly added node information and the operation and maintenance knowledge graph; processing the newly added node knowledge graph based on the preset DeepWalk algorithm, the preset TransE model and the preset GCN model, respectively obtaining node vector representation data and relationship vector representation data, and then performing regression processing to generate fused node vector representation data and fused relationship vector representation data; completing the knowledge supplement of the operation and maintenance knowledge graph based on the fused node vector representation data and the fused relationship vector representation data. The method for supplementing knowledge of the operation and maintenance knowledge graph can achieve the technical effect of improving the efficiency and accuracy of knowledge supplementation of the knowledge graph.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and specifically to a knowledge supplement method, device, electronic device and computer-readable storage medium for an operation and maintenance knowledge graph. Background Art

[0002] Currently, with economic growth, businesses are becoming more diverse, and the overall architecture required to accommodate these diverse businesses is becoming increasingly complex. Data centers, such as those in the financial and manufacturing sectors, are becoming an effective asset management tool for a growing number of enterprises. Data center operations and maintenance (O&M) have naturally become essential for business survival, and how to achieve better O&M for the environment, IT equipment, and business operations has become a hot research topic. Data centers generate terabytes, or even petabytes, of data daily. This massive amount of data represents both a valuable asset and a significant pain point, posing a significant challenge for businesses in terms of effective data mining and analysis. To ensure a positive user experience and highly reliable business continuity, O&M tasks have become incredibly challenging, and traditional O&M methods are no longer suitable for the big data environment. Intelligent O&M is one effective means of addressing these challenges.

[0003] Knowledge graphs have found significant application in numerous fields, such as finance and traditional Chinese medicine. They store entities, attributes, and relationships and present them in a graph format, intuitively conveying the entities and relationships involved in a given field. Furthermore, knowledge graphs can be used for reasoning and to uncover a wealth of valuable information. Knowledge graphs are still relatively new in the field of operations and maintenance. While data in this field is largely structured, which undoubtedly offers advantages for knowledge graph creation, effectively and rationally defining the framework of entities, attributes, and relationships within the operations and maintenance knowledge graph is crucial for building such a knowledge graph. Furthermore, as businesses expand and scale, dynamic knowledge updates are inevitable. For enterprises, launching new services, adding new servers, and deploying new databases are commonplace, but each step represents a significant update to the operations and maintenance knowledge graph. Adding a new node requires supplementing the relevant knowledge, specifically the relationships associated with that node. Traditional methods for updating knowledge graphs are cumbersome and complex, requiring significant manual effort to manually supplement and verify the knowledge of newly added nodes. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide a knowledge supplement method, device, electronic device and computer-readable storage medium for an operation and maintenance knowledge graph, which can achieve the technical effect of improving the efficiency and accuracy of knowledge supplementation in the knowledge graph.

[0005] This embodiment of the present application provides a method for supplementing knowledge of an operation and maintenance knowledge graph, including:

[0006] Create an operations and maintenance knowledge graph;

[0007] Get new node information;

[0008] Generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph;

[0009] Generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data;

[0010] The newly added node knowledge graph is used as data input to train the preset TransE model to obtain the second node vector representation data and the second relationship vector representation data;

[0011] The newly added node knowledge graph is used as data input to train the preset GCN model to obtain the third node vector representation data;

[0012] performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data;

[0013] The knowledge supplement of the operation and maintenance knowledge graph is completed according to the fusion node vector representation data and the fusion relationship vector representation data.

[0014] In the above implementation process, after obtaining the new node information and generating the new node knowledge graph, the knowledge supplement method of the operation and maintenance knowledge graph can obtain the knowledge representation of the nodes and relationships of the new node knowledge graph through TransE (only one result), and can obtain the knowledge representation of the nodes of the new node knowledge graph through DeepWalk and GCN (lack of relationship knowledge representation); the knowledge supplement method of the operation and maintenance knowledge graph completes the knowledge supplement of the knowledge graph by integrating the three algorithm models of DeepWalk, TransE and GCN, avoiding the problems of manual updating of the knowledge graph being inefficient and low in accuracy, and the low accuracy of knowledge supplementation by a single graph representation learning method, and can achieve the technical effect of improving the efficiency and accuracy of knowledge supplementation of the knowledge graph.

[0015] Furthermore, the step of generating a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and operating the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data includes:

[0016] All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct the node relationship matrix, where the matrix value of the node relationship matrix is ​​0 or 1, where 0 indicates that there is no connection between the current row unit and the current column unit, and 1 indicates that there is a connection between the current row unit and the current column unit;

[0017] A node random walk result is obtained based on the preset DeepWalk algorithm and the node relationship matrix, and a model is trained according to the node random walk result to obtain first node vector representation data and first relationship vector representation data.

[0018] In the above implementation, the pre-set DeepWalk algorithm expands the existing node relationship matrix, adding the relationship as a node. As a result, the matrix contents are still either 0 or 1, with 0 indicating no connection and 1 indicating a connection. Repeating the traditional DeepWalk operation yields vector representations of all nodes and relationships in the knowledge graph with the newly added node, i.e., obtaining the first node vector representation data and the first relationship vector representation data.

[0019] Furthermore, the step of performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data includes:

[0020] Performing regression processing on the first node vector representation data, the second node vector representation data, and the third node vector representation data according to a first preset formula to obtain fused node vector representation data;

[0021] Regression processing is performed on the first relationship vector representation data and the second relationship vector representation data according to a second preset formula to generate fused relationship vector representation data.

[0022] Furthermore, the first preset formula is expressed as:

[0023] V=a0+a1*Vd+a2*Vt+a3*Vg;

[0024] Where V represents the fusion node vector representing the data, Vd represents the first node vector representing the data, Vt represents the second node vector representing the data, Vg represents the third node vector representing the data, and a0, a1, a2, and a3 represent the undetermined regression analysis parameters;

[0025] The second preset formula is expressed as:

[0026] E=b0+b1*Ed+b2*Et;

[0027] Among them, E represents the fusion relationship vector representing the data, Ed represents the first relationship vector representing the data, Et represents the second relationship vector representing the data, and b0, b1, and b2 represent the undetermined regression analysis parameters.

[0028] Furthermore, the step of completing the knowledge supplement of the operation and maintenance knowledge graph according to the fused node vector representation data and the fused relationship vector representation data includes:

[0029] The distance information between the newly added nodes, the original nodes and the preset relationships is calculated based on the fused node vector representation data and the fused relationship vector representation data, and the nodes are arranged in ascending order based on the distance information to complete the knowledge supplement of the operation and maintenance knowledge graph.

[0030] In the above implementation process, the distance information is sorted in ascending order, that is, knowledge can be updated through sequential recommendation, which can greatly save the time of manual knowledge supplementation and knowledge verification, improve operation and maintenance efficiency, and better meet the needs of more enterprises to reduce costs and increase efficiency.

[0031] Furthermore, the step of generating a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph includes:

[0032] In the operation and maintenance knowledge graph, a node category and at least one relationship are added to the newly added node of the newly added node information to generate the newly added node knowledge graph.

[0033] In a second aspect, an embodiment of the present application provides a knowledge supplement device for an operation and maintenance knowledge graph, including:

[0034] Graph creation module, used to create operation and maintenance knowledge graph;

[0035] Added a new node module to obtain new node information;

[0036] A new graph module is added, used to generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph;

[0037] A first representation module is configured to generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and to operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data;

[0038] A second representation module is configured to use the newly added node knowledge graph as data input to train a preset TransE model to obtain second node vector representation data and second relationship vector representation data;

[0039] A third representation module is configured to use the newly added node knowledge graph as data input to train a preset GCN model to obtain third node vector representation data;

[0040] a fusion module, configured to perform regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data;

[0041] A supplement module is used to complete the knowledge supplement of the operation and maintenance knowledge graph based on the fusion node vector representation data and the fusion relationship vector representation data.

[0042] Furthermore, the first representation module is specifically configured to:

[0043] All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct the node relationship matrix, where the matrix value of the node relationship matrix is ​​0 or 1, where 0 indicates that there is no connection between the current row unit and the current column unit, and 1 indicates that there is a connection between the current row unit and the current column unit;

[0044] A node random walk result is obtained based on the preset DeepWalk algorithm and the node relationship matrix, and a model is trained according to the node random walk result to obtain first node vector representation data and first relationship vector representation data.

[0045] Furthermore, the fusion module is specifically used to:

[0046] Performing regression processing on the first node vector representation data, the second node vector representation data, and the third node vector representation data according to a first preset formula to obtain fused node vector representation data;

[0047] Regression processing is performed on the first relationship vector representation data and the second relationship vector representation data according to a second preset formula to generate fused relationship vector representation data.

[0048] Furthermore, the knowledge supplement module is specifically used to:

[0049] The distance information between the newly added nodes, the original nodes and the preset relationships is calculated based on the fused node vector representation data and the fused relationship vector representation data, and the nodes are arranged in ascending order based on the distance information to complete the knowledge supplement of the operation and maintenance knowledge graph.

[0050] Furthermore, the newly added atlas module is specifically used to:

[0051] In the operation and maintenance knowledge graph, a node category and at least one relationship are added to the newly added node of the newly added node information to generate the newly added node knowledge graph.

[0052] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in any one of the first aspects when executing the computer program.

[0053] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which instructions are stored. When the instructions are executed on a computer, the computer executes the method as described in any one of the first aspects.

[0054] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when running on a computer, enables the computer to execute the method as described in any one of the first aspects.

[0055] Other features and advantages disclosed in the present application will be described in the following description, or some features and advantages can be inferred or determined without doubt from the description, or can be learned by implementing the above-mentioned technology disclosed in the present application.

[0056] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0058] Figure 1 A flowchart of a method for supplementing knowledge of an operation and maintenance knowledge graph provided in an embodiment of the present application;

[0059] Figure 2 A flowchart of another method for supplementing knowledge of an operation and maintenance knowledge graph provided in an embodiment of the present application;

[0060] Figure 3 A structural block diagram of a knowledge supplement device for the operation and maintenance knowledge graph provided in an embodiment of the present application;

[0061] Figure 4 This is a structural block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0063] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0064] The embodiments of the present application provide a method, device, electronic device and computer-readable storage medium for knowledge supplementation of an operation and maintenance knowledge graph, which can be applied to data center operation and maintenance; after obtaining the information of the newly added node and generating the knowledge graph of the newly added node, the knowledge supplementation method of the operation and maintenance knowledge graph can obtain the knowledge representation of the nodes and relationships of the newly added node knowledge graph through TransE (only one result), and obtain the knowledge representation of the nodes of the newly added node knowledge graph through DeepWalk and GCN (lack of relationship knowledge representation); the knowledge supplementation method of the operation and maintenance knowledge graph completes the knowledge supplement of the knowledge graph by integrating the three algorithm models of DeepWalk, TransE and GCN, avoiding the problems of manual updating of the knowledge graph being inefficient and inaccurate, and the low accuracy of knowledge supplementation by a single graph representation learning method, and can achieve the technical effect of improving the efficiency and accuracy of knowledge supplementation of the knowledge graph.

[0065] See Figure 1 , Figure 1 This is a flow chart of a method for supplementing knowledge of an operation and maintenance knowledge graph provided in an embodiment of the present application. The method for supplementing knowledge of an operation and maintenance knowledge graph includes the following steps:

[0066] S100: Create an operation and maintenance knowledge graph.

[0067] In some implementations, examples of creating an operation and maintenance knowledge graph are as follows:

[0068] Acquire historical operation and maintenance data, the historical operation and maintenance data including topology data, alarm template data and solution data; determine a first type of entity based on the topology data; determine a second type of entity based on the alarm template data; determine a third type of entity based on the solution data; determine the entity relationship between each entity element of the first type of entity, the second type of entity and the third type of entity in the operation and maintenance knowledge graph based on the topology data, the alarm template data and the solution data; determine the entity attributes of each entity element in the operation and maintenance knowledge graph based on the topology data, the alarm template data and the solution data.

[0069] It should be noted that the above-mentioned operation and maintenance knowledge graph creation process is only an example and not a limitation. The creation of the operation and maintenance knowledge graph can also refer to other existing methods, which will not be repeated here.

[0070] S200: Acquire new node information.

[0071] S300: Generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph.

[0072] Exemplarily, after obtaining the newly added node information, the newly added node in the newly added node information is added to the operation and maintenance knowledge graph, that is, a node category and at least one relationship are added to the newly added node to generate a new node knowledge graph.

[0073] S400: Generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data;

[0074] S500: Using the newly added node knowledge graph as data input to train the preset TransE model, and obtaining the second node vector representation data and the second relationship vector representation data;

[0075] S600: Using the newly added node knowledge graph as data input to train the preset GCN model to obtain the third node vector representation data;

[0076] S700: Perform regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data;

[0077] S800: Complete the knowledge supplement of the operation and maintenance knowledge graph based on the fused node vector representation data and the fused relationship vector representation data.

[0078] Exemplarily, after obtaining the new node information and generating the new node knowledge graph, the knowledge supplement method of the operation and maintenance knowledge graph can obtain the knowledge representation of the nodes and relationships of the new node knowledge graph through TransE (only one result), and obtain the knowledge representation of the nodes of the new node knowledge graph through DeepWalk and GCN (lack of relationship knowledge representation); the knowledge supplement method of the operation and maintenance knowledge graph completes the knowledge supplement of the knowledge graph by integrating the three algorithm models of DeepWalk, TransE and GCN, avoiding the problems of manual updating of the knowledge graph being inefficient and inaccurate, and the low accuracy of knowledge supplementation by a single graph representation learning method, and can achieve the technical effect of improving the efficiency and accuracy of knowledge supplementation of the knowledge graph.

[0079] See Figure 2 , Figure 2 A flowchart of another method for supplementing knowledge of an operation and maintenance knowledge graph provided in an embodiment of the present application.

[0080] Exemplarily, S400: generating a node relationship matrix based on nodes and relationships in the newly added node knowledge graph as matrix parameters, and operating the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data, includes:

[0081] S410: All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct a node relationship matrix. The matrix value of the node relationship matrix is ​​0 or 1, where 0 indicates that there is no connection between the current row unit and the current column unit, and 1 indicates that there is a connection between the current row unit and the current column unit;

[0082] S420: Obtain node random walk results based on a preset DeepWalk algorithm and a node relationship matrix, perform model training according to the node random walk results, and obtain first node vector representation data and first relationship vector representation data.

[0083] For example, the traditional DeepWalk is based on the random walk and skip-gram modes. It builds a node relationship matrix (each row and each column represents a node in the knowledge graph, and the matrix value is 0 or 1, 0 means there is no relationship between the current row node and the current column node, and 1 means there is a relationship), and then starts from a node and randomly walks through several nodes (the basis of walking is that there is a relationship between the nodes), and records the results of each random walk. By repeating the previous action, we can get a large number of random walk results. These data can be regarded as text in natural language processing. According to the skip-gram in word2vec, we can use the results of these random walks to train the model, and we can output the vector representation of all nodes, that is, the knowledge representation of the nodes; but in knowledge supplementation, the vector representation of relationships is particularly important;

[0084] In the knowledge supplementation method for the operation and maintenance knowledge graph provided in the embodiments of the present application, a preset DeepWalk algorithm expands the original node relationship matrix and places the relationship as a node in the node relationship matrix; thus, the value of the matrix content is also 0 or 1, 0 indicating that the two are not connected, and 1 indicating that the two are connected. Repeating the traditional DeepWalk operation can obtain the vector representation of all nodes and relationships in the knowledge graph with the newly added node, that is, obtaining the first node vector representation data and the first relationship vector representation data.

[0085] Exemplarily, S700: performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data includes:

[0086] S710: Perform regression processing on the first node vector representation data, the second node vector representation data, and the third node vector representation data according to a first preset formula to obtain fused node vector representation data;

[0087] S720: Perform regression processing on the first relationship vector representation data and the second relationship vector representation data according to a second preset formula to generate fused relationship vector representation data.

[0088] Exemplarily, the first preset formula is expressed as:

[0089] V=a0+a1*Vd+a2*Vt+a3*Vg;

[0090] Where V represents the fusion node vector representing the data, Vd represents the first node vector representing the data, Vt represents the second node vector representing the data, Vg represents the third node vector representing the data, and a0, a1, a2, and a3 represent the undetermined regression analysis parameters;

[0091] The second preset formula is expressed as:

[0092] E=b0+b1*Ed+b2*Et;

[0093] Among them, E represents the fusion relationship vector representing the data, Ed represents the first relationship vector representing the data, Et represents the second relationship vector representing the data, and b0, b1, and b2 represent the undetermined regression analysis parameters.

[0094] Exemplarily, the step of completing the knowledge supplement of the operation and maintenance knowledge graph according to the fused node vector representation data and the fused relationship vector representation data in S800 includes:

[0095] S810: Calculate the distance information between the newly added nodes, the original nodes, and the preset relationships based on the fused node vector representation data and the fused relationship vector representation data, and arrange them in ascending order based on the distance information to complete the knowledge supplement of the operation and maintenance knowledge graph.

[0096] For example, by arranging in ascending order according to distance information, knowledge can be updated through sequential recommendation, which can greatly save the time of manual knowledge supplementation and knowledge verification, improve operation and maintenance efficiency, and better meet the needs of more enterprises to reduce costs and increase efficiency.

[0097] Exemplarily, S300: the step of generating a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph includes:

[0098] S310: Add a node category and at least one relationship to the newly added node of the newly added node information in the operation and maintenance knowledge graph to generate a newly added node knowledge graph.

[0099] In some embodiments, combined Figures 1 to 2 The knowledge supplement method of the operation and maintenance knowledge graph provided in the embodiment of the present application has the following specific steps and processes:

[0100] Step 1: Create an operation and maintenance knowledge graph;

[0101] Step 2: Generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph, keeping the knowledge graph complete;

[0102] A complete graph requires relationship support, meaning that each node in the knowledge graph must have at least one relationship. There cannot be unrelated nodes or fragmented graphs. This requires that when a new node is added, at least one relationship and the node's category can be manually added to ensure the integrity of the knowledge graph.

[0103] Step 3: DeepWalk method improvement;

[0104] The preset DeepWalk algorithm is an improvement based on the traditional DeepWalk method; the traditional DeepWalk is based on the random walk and skip-gram modes, which is achieved by constructing a node relationship matrix (each row and each column represents a node in the knowledge graph, and the matrix value is 0 or 1, 0 means there is no relationship between the current row node and the current column node, and 1 means there is a relationship), and then starting from a node, randomly walk through several nodes (the basis of walking is that there is a relationship between the nodes), and record the results of each random walk. By repeating the previous action, we can get a large number of random walk results. These data can be regarded as text in natural language processing. According to the skip-gram in word2vec, we can use the results of these random walks to train the model, and output the vector representation of all nodes, that is, the knowledge representation of the node;

[0105] However, in knowledge supplementation, the vector representation of relationships is particularly important. Therefore, we improve DeepWalk (preset DeepWalk algorithm) here, expand the original node relationship matrix, and put the relationship as a node into the node relationship matrix. The value of the matrix content is also 0 or 1, 0 means that the two are not connected, and 1 means that the two are connected. Repeating the traditional DeepWalk operation can obtain the vector representation of all nodes and relationships in the knowledge graph with the newly added node, denoted as Vd and Ed;

[0106] Step 4: Knowledge representation of TransE and GCN;

[0107] Here, we use the knowledge graph with newly added nodes as data input, and train the TransE model and GCN model separately to obtain the vector representations of nodes and relationships: Vt (TransE node vector representation), Et (TransE relationship vector representation), and Vg (GCN node vector representation). GCN is mainly used to classify nodes in the knowledge graph, so its output does not have a vector representation of relationships.

[0108] Step 5: Knowledge representation fusion;

[0109] The above three methods have obtained the knowledge representation of different nodes and relationships. We need to fuse them by regression method, that is, V=a0+a1*Vd+a2*Vt+a3*Vg, E=b0+b1*Ed+b2*Et. We believe that the relationships in the original knowledge graph are correct. For example, node 1 (V1) and node 2 (V2) and the relationship between them (E12) all existed before the new node was added (the direction of E12 is ). The three should have D=||V1+E12-V2|| in the graph representation, and D=0, that is, if there is a relationship between two nodes, then the distance between the two nodes and the relationship should be 0, where ||·|| represents the norm. Then according to this relationship, we can use the original nodes and relationships to perform regression analysis on a0, a1, a2, a3, b0, b1, b2 to obtain the vector representation of the nodes and relationships after fusion;

[0110] Step 6: Knowledge supplement;

[0111] This application uses the method of knowledge representation to achieve knowledge supplementation, and the supplementation of the relationship between nodes. Through the above operations, we have been able to obtain the vector representation of all nodes and relationships, and also perform knowledge supplementation according to the principle of regression analysis above, that is, if there is a certain relationship between two points, then ||V1+E12-V2||=0 is true. Here we choose Et (the output result of TransE) for calculation, mainly considering that we are supplementing knowledge in a recommendation manner, so we only need to calculate the distance between the newly added nodes, the original nodes and a certain type of relationship and arrange them in ascending order according to the distance, and recommend according to the sorting results. The specific process is that we select the vector representation Vtar of the newly added node, the vector representation Es of a certain type of relationship and another type of node set Vgroup of the current relationship = {V0, V1, V2,......, Vn}. It is necessary to clarify the direction of Es here. Here it is assumed that Vtar points to Vgroup. After determining the node and relationship data, the distances between Vtar, Es, and Vi are calculated according to the formula D = || Vtar + Es - Vi ||, i = 0, 1, 2, ...., n, and the results are sorted in ascending order according to the distance and recommended.

[0112] See Figure 3 , Figure 3 This is a structural block diagram of a knowledge supplement device for an operation and maintenance knowledge graph provided in an embodiment of the present application. The knowledge supplement device for an operation and maintenance knowledge graph includes:

[0113] A graph creation module 100 is used to create an operation and maintenance knowledge graph;

[0114] A new node module 200 is used to obtain new node information;

[0115] A new graph module 300 is used to generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph;

[0116] A first representation module 400 is configured to generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and to operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data;

[0117] The second representation module 500 is used to use the newly added node knowledge graph as data input to train the preset TransE model to obtain second node vector representation data and second relationship vector representation data;

[0118] The third representation module 600 is used to input the newly added node knowledge graph as data to train the preset GCN model to obtain third node vector representation data;

[0119] a fusion module 700 for performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data;

[0120] The supplement module 800 is used to complete the knowledge supplement of the operation and maintenance knowledge graph based on the fused node vector representation data and the fused relationship vector representation data.

[0121] Exemplarily, the first representation module 400 is specifically configured to:

[0122] All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct a node relationship matrix. The matrix value of the node relationship matrix is ​​0 or 1. 0 means there is no connection between the current row unit and the current column unit, and 1 means there is a connection between the current row unit and the current column unit.

[0123] Based on the preset DeepWalk algorithm and the node relationship matrix, the node random walk results are obtained, and the model is trained according to the node random walk results to obtain the first node vector representation data and the first relationship vector representation data.

[0124] Exemplarily, the fusion module 700 is specifically configured to:

[0125] Performing regression processing on the first node vector representation data, the second node vector representation data, and the third node vector representation data according to a first preset formula to obtain fused node vector representation data;

[0126] Regression processing is performed on the first relationship vector representation data and the second relationship vector representation data according to a second preset formula to generate fused relationship vector representation data.

[0127] Exemplarily, the knowledge supplement module 800 is specifically configured to:

[0128] Based on the fused node vector representation data and the fused relationship vector representation data, the distance information between the newly added nodes, the original nodes and the preset relationships is calculated, and they are arranged in ascending order according to the distance information to complete the knowledge supplement of the operation and maintenance knowledge graph.

[0129] Exemplarily, the newly added atlas module 300 is specifically used to:

[0130] In the operation and maintenance knowledge graph, a node category and at least one relationship are added to the newly added node of the newly added node information to generate a new node knowledge graph.

[0131] It should be noted that the knowledge supplement device for the operation and maintenance knowledge graph provided in the embodiment of the present application is the same as Figures 1 to 2 The method embodiments described above correspond to each other and will not be described again here to avoid repetition.

[0132] This application also provides an electronic device, see Figure 4 , Figure 4 This is a block diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include a processor 510, a communication interface 520, a memory 530, and at least one communication bus 540. The communication bus 540 is used to enable direct communication between these components. The communication interface 520 of the electronic device in this embodiment of the present application is used to communicate signaling or data with other node devices. The processor 510 may be an integrated circuit chip with signal processing capabilities.

[0133] The processor 510 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor, or the processor 510 can also be any conventional processor.

[0134] The memory 530 may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The memory 530 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 510, the electronic device can perform the above-mentioned operations. Figures 1 to 2 The various steps involved in the method embodiment.

[0135] Optionally, the electronic device may further include a storage controller and an input / output unit.

[0136] The memory 530, storage controller, processor 510, peripheral interface, and input / output units are electrically connected to each other directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses 540. The processor 510 is used to execute executable modules stored in the memory 530, such as software function modules or computer programs included in the electronic device.

[0137] The input and output unit is used to provide users with the ability to create tasks and to create optional time periods or preset execution times for the tasks to enable interaction between the user and the server. The input and output unit can be, but is not limited to, a mouse and a keyboard.

[0138] I understand. Figure 4 The structure shown is only for illustration, and the electronic device may also include Figure 4 More or fewer components than shown, or with Figure 4 Different configurations shown. Figure 4 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0139] An embodiment of the present application further provides a storage medium having instructions stored thereon. When the instructions are run on a computer, the computer program is executed by a processor to implement the method described in the method embodiment. To avoid repetition, details are not given here.

[0140] The present application also provides a computer program product, which, when running on a computer, enables the computer to execute the method described in the method embodiment.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0142] In addition, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0143] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0144] The foregoing is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined or explained in subsequent figures.

[0145] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0146] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

Claims

1. A knowledge supplement method for operation and maintenance knowledge graph, characterized in that: include: Create an operations and maintenance knowledge graph; Get new node information; Generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph; Generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data; The newly added node knowledge graph is used as data input to train the preset TransE model to obtain the second node vector representation data and the second relationship vector representation data; The newly added node knowledge graph is used as data input to train the preset GCN model to obtain the third node vector representation data; performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data; Completing the knowledge supplement of the operation and maintenance knowledge graph according to the fused node vector representation data and the fused relationship vector representation data; The steps to create an operation and maintenance knowledge graph include: Acquire historical operation and maintenance data, including topology data, alarm template data, and solution data; Determine the first type of entity based on the topological structure data; determine the second type of entity based on the alarm template data; determine the third type of entity based on the solution data; determine the entity relationship between each entity element of the first type of entity, the second type of entity, and the third type of entity in the operation and maintenance knowledge graph based on the topological structure data, the alarm template data and the solution data; determine the entity attributes of each entity element in the operation and maintenance knowledge graph based on the topological structure data, the alarm template data and the solution data.

2. The knowledge supplement method of the operation and maintenance knowledge graph according to claim 1 is characterized in that: The step of generating a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and operating the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data includes: All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct the node relationship matrix, where the matrix value of the node relationship matrix is ​​0 or 1, where 0 indicates that there is no connection between the current row unit and the current column unit, and 1 indicates that there is a connection between the current row unit and the current column unit; A node random walk result is obtained based on the preset DeepWalk algorithm and the node relationship matrix, and a model is trained according to the node random walk result to obtain first node vector representation data and first relationship vector representation data.

3. The knowledge supplement method of the operation and maintenance knowledge graph according to claim 1 or 2, characterized in that: The step of performing regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data includes: Performing regression processing on the first node vector representation data, the second node vector representation data, and the third node vector representation data according to a first preset formula to obtain fused node vector representation data; Regression processing is performed on the first relationship vector representation data and the second relationship vector representation data according to a second preset formula to generate fused relationship vector representation data.

4. The knowledge supplement method of the operation and maintenance knowledge graph according to claim 3 is characterized in that: The first preset formula is expressed as: V = a0 + a1*Vd + a2*Vt + a3*Vg; Where V represents the fusion node vector representing the data, Vd represents the first node vector representing the data, Vt represents the second node vector representing the data, Vg represents the third node vector representing the data, and a0, a1, a2, and a3 represent the undetermined regression analysis parameters; The second preset formula is expressed as: E = b0 + b1*Ed + b2*Et; Among them, E represents the fusion relationship vector representing the data, Ed represents the first relationship vector representing the data, Et represents the second relationship vector representing the data, and b0, b1, and b2 represent the undetermined regression analysis parameters.

5. The knowledge supplement method of the operation and maintenance knowledge graph according to claim 1 is characterized in that: The step of completing knowledge supplement of the operation and maintenance knowledge graph according to the fused node vector representation data and the fused relationship vector representation data includes: The distance information between the newly added nodes, the original nodes and the preset relationships is calculated based on the fused node vector representation data and the fused relationship vector representation data, and the nodes are arranged in ascending order based on the distance information to complete the knowledge supplement of the operation and maintenance knowledge graph.

6. The knowledge supplement method of the operation and maintenance knowledge graph according to claim 1 is characterized in that: The step of generating a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph includes: In the operation and maintenance knowledge graph, a node category and at least one relationship are added to the newly added node of the newly added node information to generate the newly added node knowledge graph.

7. A knowledge supplement device for operation and maintenance knowledge graph, characterized in that: include: Graph creation module, used to create operation and maintenance knowledge graph; Added a new node module to obtain new node information; A new graph module is added, which is used to generate a new node knowledge graph based on the new node information and the operation and maintenance knowledge graph; A first representation module is configured to generate a node relationship matrix based on the nodes and relationships in the newly added node knowledge graph as matrix parameters, and to operate the node relationship matrix based on a preset DeepWalk algorithm to obtain first node vector representation data and first relationship vector representation data; A second representation module is configured to use the newly added node knowledge graph as data input to train a preset TransE model to obtain second node vector representation data and second relationship vector representation data; A third representation module is configured to use the newly added node knowledge graph as data input to train a preset GCN model to obtain third node vector representation data; a fusion module, configured to perform regression processing on the first node vector representation data, the first relationship vector representation data, the second node vector representation data, the second relationship vector representation data, and the third node vector representation data to generate fused node vector representation data and fused relationship vector representation data; A supplementing module, configured to complete knowledge supplement of the operation and maintenance knowledge graph according to the fused node vector representation data and the fused relationship vector representation data; The graph creation module is also used to: obtain historical operation and maintenance data, which includes topology data, alarm template data and solution data; determine the first type of entity based on the topology data; determine the second type of entity based on the alarm template data; determine the third type of entity based on the solution data; determine the entity relationship between each entity element of the first type of entity, the second type of entity and the third type of entity in the operation and maintenance knowledge graph based on the topology data, the alarm template data and the solution data; determine the entity attributes of each entity element in the operation and maintenance knowledge graph based on the topology data, the alarm template data and the solution data.

8. The knowledge supplement device for the operation and maintenance knowledge graph according to claim 7 is characterized in that: The first representation module is specifically used for: All nodes and relationships in the newly added node knowledge graph are used as row and column units to construct the node relationship matrix, where the matrix value of the node relationship matrix is ​​0 or 1, where 0 indicates that there is no connection between the current row unit and the current column unit, and 1 indicates that there is a connection between the current row unit and the current column unit; A node random walk result is obtained based on the preset DeepWalk algorithm and the node relationship matrix, and a model is trained according to the node random walk result to obtain first node vector representation data and first relationship vector representation data.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for supplementing the knowledge of the operation and maintenance knowledge graph as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the computer executes the knowledge supplement method for the operation and maintenance knowledge graph according to any one of claims 1 to 6.

Images