Intelligent search method and system based on regulation of blood relationship of cloud knowledge graph

Abstract

The application belongs to the technical field of smart grid, and discloses an intelligent search method and system based on blood relationship of a regulation and control cloud knowledge graph, which comprises the following steps: obtaining power grid operation data; converting the obtained power grid operation data into a knowledge graph and storing the knowledge graph in a regulation and control cloud knowledge graph; obtaining a user query statement; performing search intention analysis on the user query statement to obtain a processed query statement; and querying data in the regulation and control cloud knowledge graph according to the processed query statement and returning the data to the user. The application uses data blood analysis to remove repeated and redundant data from multi-source data in a power grid, judges the implicit connection between data in different graphs, and can search the data of multiple knowledge graphs at the same time, so that the search content is more comprehensive.

Description

Technical Field

[0001] This invention belongs to the field of smart grid technology, and specifically relates to an intelligent search method and system based on the lineage relationship of a control cloud knowledge graph. Background Technology

[0002] Smart grids and power big data are crucial for addressing resource constraints and environmental pressures, and achieving green and sustainable development in the construction of a new generation of power systems. Power regulation data permeates all aspects of power production, transmission, dispatch, and management. With the continuous expansion of the power grid, a vast amount of power data has accumulated, containing valuable experience in grid operation and safe production. However, due to the sheer volume and complexity of grid data, despite unified standards for data formats within the power industry, data storage still varies across different platforms and sectors within the grid. When faced with massive amounts of grid data, how to uniformly collect and manage this data, discover implicit relationships between data points, and quickly and accurately retrieve the data needed by users is crucial for utilizing power grid big data. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent search method and system based on regulating the lineage of cloud knowledge graphs, so as to solve the technical problem that existing technologies cannot quickly and accurately retrieve the data needed by users.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] In a first aspect, the present invention provides an intelligent search method based on regulating the kinship relationship of a cloud knowledge graph, comprising:

[0006] Acquire power grid operation data; transform the acquired power grid operation data into a knowledge graph and store it in the control cloud knowledge graph;

[0007] Obtain the user's query statement; perform search intent analysis on the user's query statement to obtain the processed query statement; query data in the control cloud knowledge graph based on the processed query statement and return it to the user.

[0008] A further improvement of this invention is that the step of analyzing the user's query statement to obtain the processed query statement specifically includes:

[0009] The user query statement is segmented into words, and the part-of-speech is classified using a trained classifier. Entities and relational descriptive phrases are extracted from the obtained user query statement to obtain the processed query statement.

[0010] A further improvement of this invention is that the step of querying data in the control cloud knowledge graph based on the processed query statement and returning it to the user specifically includes:

[0011] The processed query statement is used to generate a corresponding knowledge graph query statement; the knowledge graph query statement is used to query the corresponding data from the control cloud knowledge graph to obtain preliminary query results; then, based on the results of data lineage analysis, data with implicit relationships to the preliminary query results are filtered to obtain comprehensive query results.

[0012] A further improvement of this invention is that the step of generating a corresponding knowledge graph query statement from the processed query statement specifically includes:

[0013] Identify entities in the processed query statement and filter out candidate templates that may match from predefined templates;

[0014] The identified entities are combined with the selected candidate templates to determine whether they can match. If the entities and templates match successfully, they can form a subgraph in the knowledge graph; if they cannot form a subgraph in the knowledge graph, they are considered not to match.

[0015] If a template is successfully matched, the matching templates are sorted by matching degree; the template with the highest matching degree is selected and converted into the corresponding knowledge graph query statement.

[0016] If no template match is found, obtain the syntax dependency tree of the processed query statement; find the relation phrases in the syntax dependency tree, identify the entities in the search statement based on the relation phrases and use syntax rules to generate <entity, relation, entity> triples; convert the generated <entity, relation, entity> triples into the corresponding knowledge graph query statement.

[0017] A further improvement of this invention is that it also includes a step of organizing the comprehensive query results and then feeding them back to the user, specifically including:

[0018] For comprehensive query results, sort the results before returning them to the user, displaying the results that best match the user's query statement first.

[0019] A further improvement of this invention is that the matching degree is calculated by the following formula:

[0020]

[0021]

[0022]

[0023] f(p) = αf1(p) + βf2(p)

[0024] Where f1(p) is the weighted result calculated based on the matching degree of the hit entity attributes, f2(p) is the result calculated based on the relevance of the hit entities, and F is the set of attributes of the hit entities. i R(i,j) is the matching value calculated for attribute i. R(i,j) is the relevance between entities i and j. If i and j are related, then R(i,j) = 1, otherwise R(i,j) = 0.

[0025] A further improvement of this invention is that the step of filtering data with a hidden relationship to the preliminary query results based on the data lineage analysis results specifically includes:

[0026] The formula for calculating implicit relationships is:

[0027] f(e i e j )=(α·R1(e i ,e j )+β·R2(e i ,e j ))·γ

[0028]

[0029]

[0030] Where R1 is the entity e computed using graph computation algorithms and deep learning algorithms. i and entity e j The degree of relationship between entities, R², is calculated through lineage analysis. i and entity e j The degree of relation is defined as follows: R is the set of all considered blood relations. For a relation r in R, if i and j are related, then r(i,j) = 1; otherwise, r(i,j) = 0. r ∑λ = 1, where α and β are weights, α + β = 1, and γ is a coefficient added after considering the relationship between region and power grid, 0 < γ < 1.

[0031] Secondly, the present invention provides an intelligent search system based on regulating kinship relationships using a cloud knowledge graph, comprising:

[0032] The knowledge graph data lineage analysis module is used to acquire power grid operation data; the acquired power grid operation data is converted into a knowledge graph and stored in the control cloud knowledge graph;

[0033] The intelligent search module is used to obtain user query statements; analyze the search intent of user query statements to obtain processed query statements; and query data in the control cloud knowledge graph based on the processed query statements and return it to the user.

[0034] Thirdly, the present invention provides an electronic device, including a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the intelligent search method based on the kinship of regulating cloud knowledge graphs.

[0035] Fourthly, the present invention provides a computer-readable storage medium storing at least one instruction, which, when executed by a processor, implements the intelligent search method based on the kinship relationship of a cloud knowledge graph.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] This invention provides an intelligent search method and system based on the lineage relationship of a control cloud knowledge graph. The method involves acquiring power grid operation data; converting the acquired power grid operation data into a knowledge graph and storing it in the control cloud knowledge graph; acquiring user query statements; performing search intent analysis on the user query statements to obtain processed query statements; and retrieving data from the control cloud knowledge graph based on the processed query statements and returning the results to the user. This invention utilizes data lineage analysis to remove duplicate and redundant data from multiple sources in the power grid, identifies implicit relationships between data in different knowledge graphs, and allows for simultaneous searching of data from multiple knowledge graphs, resulting in a more comprehensive search.

[0038] This invention designs a semantic understanding method that combines semantic classification and historical search logs. It initially judges the user's search intent by the keywords in the user's query, and further judges the user's potential search intent or content of interest by combining the user's search and browsing history data. This solves the problem that the search question entered by the user may contain too little information, and avoids the trouble of secondary interaction with the user.

[0039] This invention designs a search statement generation method based on predefined templates and supplemented by semantic extraction. For knowledge graphs in a specific domain, commonly used query templates are preset. When searching, the preset template matching method is used first to generate query statements. Only when the template matching fails is the semantic extraction method used to generate query statements. The use of predefined template matching not only saves system overhead, but also improves the professionalism and accuracy of query statements to a certain extent.

[0040] This invention utilizes data lineage analysis to discover implicit relationships between data in different knowledge graphs of the regulatory cloud, enabling collaborative searching across these graphs. In this invention, the statement intent classification module uses historical search and browsing logs as auxiliary information to determine the user's search intent. Furthermore, the intelligent search based on the lineage relationships of the regulatory cloud knowledge graph calculates the search result matching degree by comprehensively considering the matching degree between the hit entities and the search terms, as well as the relevance between the hit entities, in the result ranking.

[0041] This invention studies intelligent search based on the lineage relationships of a control cloud knowledge graph. It performs data lineage analysis on entities and data within the control cloud knowledge graph, enabling the discovery of implicit relationships between data within complex power grid topology. This provides more comprehensive search content for intelligent search, addressing the issues of diverse and fragmented control cloud data. It broadens the scope of information available to users while improving the accuracy of information retrieval, further enhancing the intelligence level of power grid control and operation. Attached Figure Description

[0042] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0043] Figure 1 This invention provides an overall schematic diagram of an intelligent search system based on regulating the bloodline relationship of a cloud knowledge graph;

[0044] Figure 2 This is a schematic diagram of the data lineage analysis module;

[0045] Figure 3 A schematic diagram of the statement intent classification module;

[0046] Figure 4 This is a schematic diagram of the intelligent search and matching module;

[0047] Figure 5 This is a schematic diagram of an intelligent search method based on regulating the bloodline relationship of a cloud knowledge graph according to the present invention;

[0048] Figure 6 This is a structural block diagram of an intelligent search system based on regulating the bloodline relationship of a cloud knowledge graph according to the present invention;

[0049] Figure 7 This is a structural block diagram of an electronic device according to the present invention. Detailed Implementation

[0050] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0051] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this invention is for describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0052] A knowledge graph is a semantic network that typically uses <entity, relation, entity> triples to represent the attributes of things and the relationships between things.

[0053] Data lineage refers to a relationship similar to blood ties in human society that forms between data during the process of data generation, processing, flow, and eventual disappearance.

[0054] Example 1

[0055] Please see Figure 1 As shown, this invention provides an intelligent search system based on the lineage relationship of a regulatory cloud knowledge graph. It mainly comprises two parts: a knowledge graph data lineage analysis module and an intelligent search module. The knowledge graph data lineage analysis module is used to analyze the regulatory cloud knowledge graph and power grid operation data, standardize the data, and determine the implicit relationships between data to support intelligent search. The intelligent search module is used to query data in the knowledge graph based on the content requested by the user and return it to the user.

[0056] 1. Knowledge Graph Data Lineage Analysis Module

[0057] The knowledge graph data lineage analysis module is used to analyze the control cloud knowledge graph and power grid operation data, standardize the data, and determine the implicit relationships between the data to support intelligent search.

[0058] By utilizing the knowledge graph data lineage analysis module, we can discover implicit relationships between entities, remove duplicate and redundant data from different data sources, standardize the data, and retrieve all relevant data from different graphs as much as possible during the search.

[0059] Please see Figure 2 As shown, the knowledge graph data lineage analysis module transforms power grid operation data into a knowledge graph and stores it in the control cloud knowledge graph. For the data in the control cloud knowledge graph, firstly, graph computation algorithms (shortest path algorithm) and deep learning algorithms (Apriori algorithm) are used to calculate the relationships between the data. The shortest path algorithm calculates the distance between entities in the graph, and the reciprocal is taken as the result of the graph computation algorithm. The Apriori algorithm calculates the support relationship between entities in the graph; those with support below a threshold are recorded as 0. The relationship values ​​calculated by the two algorithms are weighted and summed to initially determine whether a relationship exists between the data. Then, a similarity algorithm is used to calculate the similarity of entity definitions and attributes. If the similarity exceeds a threshold, it is determined that they belong to the same entity, complementary content is merged, and redundant content is deduplicated. Finally, based on the characteristics of the data in the flow process, such as data source, data producer, and data processing flow, the lineage relationship of the data is calculated (the calculation process is shown in R2).

[0060] Furthermore, due to the numerous national, provincial, prefectural, and county-level nodes in the control cloud and their hierarchical relationships, as well as the complex power grid structure, implicit relationships between data from different levels of control clouds within the same region and data from the same power grid must also be considered. The specific calculations are as follows:

[0061]

[0062]

[0063]

[0064] Where R1 is the entity e computed using graph computation algorithms and deep learning algorithms. i and entity e j The degree of relationship between entities, R², is calculated through lineage analysis. i and entity e j The degree of relation is defined as follows: R is the set of all considered blood relations. For a relation r in R, if i and j are related, then r(i,j) = 1; otherwise, r(i,j) = 0. r ∑λ = 1 is the weight of relation r, α and β are the weights of the relation calculated using these two methods, α + β = 1, and γ is the coefficient added after considering the relationship between the region and the power grid, 0 < γ < 1.

[0065] 2. Intelligent Search Module

[0066] The intelligent search module is used to retrieve data from the knowledge graph based on the content the user wants to query and return it to the user.

[0067] Intelligent search based on the lineage of a cloud knowledge graph mainly comprises three modules: a statement intent classification module, a knowledge search matching module, and a result organization module. The statement intent classification module performs semantic understanding on the acquired user queries and extracts key information. The knowledge search matching module uses the extracted key information to generate corresponding query statements and performs searches within the knowledge graph. The result organization module organizes and sorts the information retrieved by the knowledge search matching module and displays the results to the user.

[0068] 2.1 Statement Intent Classification Module

[0069] Please see Figure 3As shown, the statement intent classification module analyzes the user's desired query content separately through classifier classification, or it combines classifier classification with search logs to analyze the user's desired query content. On one hand, by preprocessing the user's query statement, the module performs word segmentation and classifies parts of speech using a trained classifier, extracting entity and relational descriptive phrases from the obtained user query statement to obtain a preprocessed query statement. On the other hand, the module queries the search logs and, based on the user's interaction with the search engine, finds the content the user focused on when querying related entities from the user's historical query records and browsing history, inferring the information the user might want to query, and correcting the preprocessed query statement. For example, when a user searches for "Longkou Hydropower Plant," the logs record that the user often clicks on the equipment information of Longkou Hydropower Plant. In this case, in addition to querying the relevant information of the entity "Longkou Hydropower Plant," it may also be necessary to query "What equipment does Longkou Hydropower Plant have," i.e., <plant, belongs to, equipment>, to avoid the limited amount of information obtained due to the user's input being too concise, making it difficult to match the search results the user wants.

[0070] 2.2 Knowledge Search and Matching Module

[0071] Please see Figure 4 As shown, the core of the knowledge search matching module is the generation of query statements. Since the data is stored in the knowledge graph in the form of <entity, relation, entity> triples, the key to generating query statements is to use the processed query statements described by natural language to generate the corresponding knowledge graph query statements. The generation of query statements is achieved by using predefined templates as the main method and semantic extraction as the auxiliary method. Based on entities, relations and attributes, the templates can be roughly divided into three types: (1) "entity" template, such as "What is a circuit breaker?", used to view relevant information about the equipment; (2) "entity + relation" template, such as "What equipment does Longkou Hydropower Plant have?", the relation triple is <plant, belongs to, equipment>, "Records of step-down transformer 001?", the relation triple is <equipment, related records, records>; (3) "entity + attribute" template, "Voltage level of step-down transformer 001?", can be used to view the specific attribute information of the entity.

[0072] The specific process for generating query statements is as follows:

[0073] 1) Based on the processed query statement obtained by the statement intent classification module, identify the entities in the processed query statement and filter out the candidate templates that may match from the predefined templates.

[0074] 2) Combine the identified entities with the selected candidate templates and determine whether they can match. Successfully matched entities and templates should be able to form a subgraph in the knowledge graph; if they cannot form a subgraph in the knowledge graph, they are considered not to match.

[0075] 3) If a template is successfully matched, the successfully matched templates are sorted according to their matching degree, which is calculated by the relevance between the template and the search statement;

[0076]

[0077] Where I is the set of entities and relations identified in the query statement, and R... i This indicates whether element i in I is contained in the successfully matched template; if it is, then R... i =1, otherwise R i =0,λ i λ represents the weight of element i; if i is an entity, then λ represents the weight of element i. i =1, otherwise λ i =0.5.

[0078] 4) Select the template with the highest matching degree, convert it into the corresponding knowledge graph query statement, and query the results in the knowledge graph.

[0079] If no matching template can be found for a search query, a semantic extraction-based approach is used to generate the query.

[0080] 1) Use the statement intent classification module to obtain the syntax dependency tree of the processed query statement;

[0081] 2) Find the relation phrases in the grammatical dependency tree, and use grammatical rules to identify the entities in the search statement based on the relation phrases, generating <entity, relation, entity> triples;

[0082] 3) Convert the generated <entity, relation, entity> triples into corresponding knowledge graph query statements and perform queries in the knowledge graph.

[0083] After converting the user's search query into a corresponding structured query, the required data is retrieved from the control cloud knowledge graph to obtain preliminary query results. Then, based on the results of data lineage analysis, data with implicit relationships to the preliminary query results are filtered out to search for the results of the user's query content as comprehensively as possible.

[0084] 2.3 Results Organization Module

[0085] For accurate results obtained from the knowledge graph, the results are sorted before being returned to the user, with the results that match the search query most highly displayed first. When calculating the matching score, in addition to considering the match between query terms and entities, the relationships between the matched entities also need to be taken into account. Therefore, the score for ranking search results mainly consists of two parts: the match value between the entity and the search term, and the relationship between the entities. The matching score is calculated using the following formula:

[0086]

[0087]

[0088]

[0089] f(p) = αf1(p) + βf2(p)

[0090] Where f1(p) is the weighted result calculated based on the matching degree of the hit entity attributes, f2(p) is the result calculated based on the relevance of the hit entities, and F is the set of attributes of the hit entities. i R(i,j) is the matching value calculated for attribute i, where R(i,j) represents the relevance between entities i and j. If i and j are related, then R(i,j) = 1; otherwise, R(i,j) = 0. Based on the above formula, the specific steps for calculating the matching degree are as follows:

[0091] 1) Calculate the matching value F between each attribute of each matched entity and the search term. i This is obtained through similarity calculation;

[0092] 2) Perform a weighted summation of the matching values ​​of each hit entity attribute to obtain f1(p);

[0093] 3) Calculate the relationship value between each hit entity. If any two entities are related, the relationship value is incremented by 1. That is, calculate how many pairs of related entities there are among all hit entities.

[0094] 4) The matching value and the relationship value are weighted and summed to obtain the final matching score f(p) of the search results.

[0095] Example 2

[0096] Please see Figure 5 As shown, this invention provides an intelligent search method based on regulating the lineage relationships of a cloud knowledge graph, comprising:

[0097] Acquire power grid operation data; transform the acquired power grid operation data into a knowledge graph and store it in the control cloud knowledge graph;

[0098] Obtain the user's query statement; perform search intent analysis on the user's query statement to obtain the processed query statement; query data in the control cloud knowledge graph based on the processed query statement and return it to the user.

[0099] In one specific implementation, the step of performing search intent analysis on the user's query statement to obtain the processed query statement specifically includes:

[0100] The user query statement is segmented into words, and the part-of-speech is classified using a trained classifier. Entities and relational descriptive phrases are extracted from the obtained user query statement to obtain the processed query statement.

[0101] In one specific implementation, the step of querying data in the control cloud knowledge graph based on the processed query statement and returning it to the user specifically includes:

[0102] The processed query statement is used to generate a corresponding knowledge graph query statement; the knowledge graph query statement is used to query the corresponding data from the control cloud knowledge graph to obtain preliminary query results; then, based on the results of data lineage analysis, data with implicit relationships to the preliminary query results are filtered to obtain comprehensive query results.

[0103] In one specific implementation, the step of generating a corresponding knowledge graph query statement from the processed query statement specifically includes:

[0104] Identify entities in the processed query statement and filter out candidate templates that may match from predefined templates;

[0105] The identified entities are combined with the selected candidate templates to determine whether they can match. If the entities and templates match successfully, they can form a subgraph in the knowledge graph; if they cannot form a subgraph in the knowledge graph, they are considered not to match.

[0106] If a template is successfully matched, the matching templates are sorted by matching degree; the template with the highest matching degree is selected and converted into the corresponding knowledge graph query statement.

[0107] If no template match is found, obtain the syntax dependency tree of the processed query statement; find the relation phrases in the syntax dependency tree, identify the entities in the search statement based on the relation phrases and use syntax rules to generate <entity, relation, entity> triples; convert the generated <entity, relation, entity> triples into the corresponding knowledge graph query statement.

[0108] In one specific implementation, the method further includes a step of organizing the comprehensive query results and then feeding them back to the user, specifically including:

[0109] For comprehensive query results, sort the results before returning them to the user, displaying the results that best match the user's query statement first.

[0110] In one specific implementation, the matching degree is calculated using the following formula:

[0111]

[0112]

[0113]

[0114] f(p) = αf1(p) + βf2(p)

[0115] Where f1(p) is the weighted result calculated based on the matching degree of the hit entity attributes, f2(p) is the result calculated based on the relevance of the hit entities, and F is the set of attributes of the hit entities. i R(i,j) is the matching value calculated for attribute i. R(i,j) is the relevance between entities i and j. If i and j are related, then R(i,j) = 1, otherwise R(i,j) = 0.

[0116] In one specific implementation, the step of further filtering data with implicit relationships to the preliminary query results based on the data lineage analysis results specifically includes:

[0117] The formula for calculating implicit relationships is:

[0118] f(e i e j )=(α·R1(e i e j )+β·R2(e i e j ))·γ

[0119]

[0120]

[0121] Where R1 is the entity e computed using graph computation algorithms and deep learning algorithms. i and entity e j The degree of relationship between entities, R², is calculated through lineage analysis. i and entity e j The degree of relation is defined as follows: R is the set of all considered blood relations. For a relation r in R, if i and j are related, then r(i,j) = 1; otherwise, r(i,j) = 0. r λ is the weight of relation r, ∑λ=1, α and β are weights, α+β=1, r is the coefficient added after considering the relationship between region and power grid, 0<γ<1.

[0122] Example 3

[0123] Please see Figure 6 As shown, this invention provides an intelligent search system based on regulating the kinship relationships of a cloud knowledge graph, comprising:

[0124] The knowledge graph data lineage analysis module is used to acquire power grid operation data; the acquired power grid operation data is converted into a knowledge graph and stored in the control cloud knowledge graph;

[0125] The intelligent search module is used to obtain user query statements; analyze the search intent of user query statements to obtain processed query statements; and query data in the control cloud knowledge graph based on the processed query statements and return it to the user.

[0126] In one specific implementation, the step of performing search intent analysis on the user's query statement to obtain the processed query statement specifically includes:

[0127] The user query statement is segmented into words, and the part-of-speech is classified using a trained classifier. Entities and relational descriptive phrases are extracted from the obtained user query statement to obtain the processed query statement.

[0128] In one specific implementation, the step of querying data in the control cloud knowledge graph based on the processed query statement and returning it to the user specifically includes:

[0129] The processed query statement is used to generate a corresponding knowledge graph query statement; the knowledge graph query statement is used to query the corresponding data from the control cloud knowledge graph to obtain preliminary query results; then, based on the results of data lineage analysis, data with implicit relationships to the preliminary query results are filtered to obtain comprehensive query results.

[0130] In one specific implementation, the step of generating a corresponding knowledge graph query statement from the processed query statement specifically includes:

[0131] Identify entities in the processed query statement and filter out candidate templates that may match from predefined templates;

[0132] The identified entities are combined with the selected candidate templates to determine whether they can match. If the entities and templates match successfully, they can form a subgraph in the knowledge graph; if they cannot form a subgraph in the knowledge graph, they are considered not to match.

[0133] If a template is successfully matched, the matching templates are sorted by matching degree; the template with the highest matching degree is selected and converted into the corresponding knowledge graph query statement.

[0134] If no template match is found, obtain the syntax dependency tree of the processed query statement; find the relation phrases in the syntax dependency tree, identify the entities in the search statement based on the relation phrases and use syntax rules to generate <entity, relation, entity> triples; convert the generated <entity, relation, entity> triples into the corresponding knowledge graph query statement.

[0135] In one specific implementation, the method further includes a step of organizing the comprehensive query results and then feeding them back to the user, specifically including:

[0136] For comprehensive query results, sort the results before returning them to the user, displaying the results that best match the user's query statement first.

[0137] In one specific implementation, the matching degree is calculated using the following formula:

[0138]

[0139]

[0140]

[0141] f(p) = αf1(p) + βf2(p)

[0142] Where f1(p) is the weighted result calculated based on the matching degree of the hit entity attributes, f2(p) is the result calculated based on the relevance of the hit entities, and F is the set of attributes of the hit entities. i R(i,j) is the matching value calculated for attribute i. R(i,j) is the relevance between entities i and j. If i and j are related, then R(i,j) = 1, otherwise R(i,j) = 0.

[0143] In one specific implementation, the step of further filtering data with implicit relationships to the preliminary query results based on the data lineage analysis results specifically includes:

[0144] The formula for calculating implicit relationships is:

[0145] f(e i e j )=(α·R1(e i e j )+β·R2(e i e j ))·γ

[0146]

[0147]

[0148] Where R1 is the entity e computed using graph computation algorithms and deep learning algorithms. i and entity e j The degree of relationship between entities, R², is calculated through lineage analysis. i and entity e j The degree of relation is defined as follows: R is the set of all considered blood relations. For a relation r in R, if i and j are related, then r(i,h) = 1; otherwise, r(i,j) = 0. r∑λ = 1, where α and β are weights, α + β = 1, and γ is a coefficient added after considering the relationship between region and power grid, 0 < γ < 1.

[0149] Example 4

[0150] Please see Figure 7 As shown, the present invention also provides an electronic device 100 for implementing an intelligent search method based on the kinship of regulating cloud knowledge graphs; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0151] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the intelligent search method based on the kinship of the cloud knowledge graph described in Embodiment 1 or 2 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0152] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0153] The memory 101 in the electronic device 100 stores multiple instructions to implement an intelligent search method based on regulating cloud knowledge graph kinship relationships, and the processor 102 can execute the multiple instructions to achieve the following:

[0154] Acquire power grid operation data; transform the acquired power grid operation data into a knowledge graph and store it in the control cloud knowledge graph;

[0155] Obtain the user's query statement; perform search intent analysis on the user's query statement to obtain the processed query statement; query data in the control cloud knowledge graph based on the processed query statement and return it to the user.

[0156] Example 5

[0157] If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0158] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0159] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0160] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An intelligent search method based on regulating blood relationship of cloud knowledge graph, characterized in that, The method comprises the following steps: acquiring power grid operation data; converting the acquired power grid operation data into a knowledge graph and storing the knowledge graph into a regulation and control cloud knowledge graph; acquiring a user query statement; performing search intent analysis on the user query statement to obtain a processed query statement; querying data in the regulation and control cloud knowledge graph according to the processed query statement and returning the data to the user; the step of performing search intent analysis on the user query statement to obtain the processed query statement specifically comprises: performing word segmentation on the user query statement, classifying the word forms through a trained classifier, extracting entities and relationship description phrases from the acquired user query statement, and obtaining the processed query statement; the step of querying data in the regulation and control cloud knowledge graph according to the processed query statement and returning the data to the user specifically comprises: generating a corresponding knowledge graph query statement from the processed query statement; querying corresponding data from the regulation and control cloud knowledge graph by using the knowledge graph query statement to obtain a preliminary query result; and filtering data having an implicit relationship with the preliminary query result according to a result of data blood relationship analysis to obtain a comprehensive query result; the step of filtering data having an implicit relationship with the preliminary query result according to the result of data blood relationship analysis specifically comprises: the calculation formula of the implicit relationship is: wherein, is the relationship degree between entity and entity calculated according to a graph computing algorithm and a deep learning algorithm, is the relationship degree between entity and entity calculated by blood relation analysis, R is a set of all considered blood relations, for a relation r in R, if i and j have a relation then , otherwise , is the weight of relation r, , α and β are weights, α + β = 1, is a coefficient added after considering the regional and power grid relations, . 2.The intelligent search method based on the blood relationship of the regulation cloud knowledge graph according to claim 1, characterized in that, the step of generating the corresponding knowledge graph query statement from the processed query statement specifically comprises: identifying entities in the processed query statement and screening possible candidate templates from predefined templates; combining the identified entities and the screened candidate templates to determine whether the entities and the templates can be matched, wherein the entities and the templates that are successfully matched can constitute a subgraph in the knowledge graph, and the entities and the templates that cannot constitute the subgraph are considered to be unable to be matched; if the templates are successfully matched, the successfully matched templates are sorted according to matching degrees; the template having the highest matching degree is selected and converted into a corresponding knowledge graph query statement; if the templates are not successfully matched, a syntax dependency tree of the processed query statement is acquired; relationship phrases in the syntax dependency tree are found out; entities in the search statement are identified according to the relationship phrases by using syntax rules to generate <entity, relationship, entity> triples; and the generated <entity, relationship, entity> triples are converted into a corresponding knowledge graph query statement. 3.The intelligent search method based on the blood relationship of the regulation cloud knowledge graph according to claim 1, characterized in that, The method further comprises a step of feeding back the comprehensive query result to the user after the comprehensive query result is organized, and the step specifically comprises: before the comprehensive query result is returned to the user, the result is sorted, and the result having the highest matching degree with the user query statement is displayed in the front. 4.The intelligent search method based on the blood relationship of the regulation cloud knowledge graph according to claim 3, characterized in that, the matching degree is calculated by the following formula: wherein, is a result of a weighting computation according to a matching degree of a hit entity attribute, is a result of a correlation computation of a hit entity, F is a set of attributes of the hit entity, is a matching value computed for attribute i, R(i,j) is a correlation of entities i and j, R(i,j)=1 if i and j are correlated, otherwise R(i,j)=0.

5. The intelligent search system based on the regulation of blood relationship of cloud knowledge graph, characterized in that, The method comprises the following steps: a knowledge graph data blood relationship analysis module is configured to acquire power grid operation data; and convert the acquired power grid operation data into a knowledge graph and store the knowledge graph into a regulation and control cloud knowledge graph; an intelligent search module is configured to acquire a user query statement; perform search intent analysis on the user query statement to obtain a processed query statement; and query data in the regulation and control cloud knowledge graph according to the processed query statement and return the data to the user; the step of performing search intent analysis on the user query statement to obtain the processed query statement specifically comprises: The user query statement is segmented, the parts of speech are classified by the trained classifier, entity and relation description phrases are extracted from the obtained user query statement, and a processed query statement is obtained; The step of querying data in the regulation cloud knowledge graph according to the processed query statement and returning the data to the user specifically comprises: The processed query statement is used to generate a corresponding knowledge graph query statement, the corresponding data in the regulation cloud knowledge graph is queried by using the knowledge graph query statement, a preliminary query result is obtained, and then data having an implicit relationship with the preliminary query result is screened according to the result of data blood relationship analysis, and a comprehensive query result is obtained; The step of screening data having an implicit relationship with the preliminary query result according to the result of data blood relationship analysis specifically comprises: The calculation formula of the implicit relationship is: wherein, is the relationship degree between entities and entities calculated according to graph computing algorithm and deep learning algorithm, is the relationship degree between entities and entities calculated by blood analysis, R is the set of all considered blood relationships, for a relationship r in R, if i and j have a relationship then , otherwise , is the weight of relationship r, , a and b are weights, a + b = 1, is the coefficient added after considering the regional and power grid relationship, .

6. An electronic device, comprising: The computer readable storage medium stores at least one instruction, and the at least one instruction is executed by the processor to implement the intelligent search method based on the blood relationship of the regulation cloud knowledge graph as claimed in any one of claims 1 to 4.

7. A computer readable storage medium characterized by, The computer readable storage medium stores at least one instruction, and the at least one instruction is executed by the processor to implement the intelligent search method based on the blood relationship of the regulation cloud knowledge graph as claimed in any one of claims 1 to 4.

Images