Knowledge graph construction method for extreme space weather event influence analysis

Abstract

The invention provides a knowledge graph construction method for extreme space weather event influence analysis, and relates to the technical field of knowledge graph construction. Comprising the steps of crawling key parameters of a space weather event, and performing cleaning treatment on crawled data to obtain a to-be-analyzed data set; performing data mining and correlation analysis on the to-be-analyzed data set based on a physical transmission mechanism of space disastrous weather; extracting event entities, parameter entities and attribute entities on the basis of the obtained data relevance, relevance between events and association rules, constructing a graph triple, and storing the graph triple into a Neo4j graph database; by calling graph triple data in a neo4j graph database, entities are standardized as nodes, relationships among the entities are standardized as edges, and then rendering is performed on a page in a vis.js height customization mode, so that professional knowledge graph visualization for space weather event influence analysis is realized.

Description

Technical Field

[0001] This invention relates to the field of knowledge graph construction technology, and in particular to a method for constructing a knowledge graph for analyzing the impact of extreme space weather events. Background Technology

[0002] Space weather broadly refers to the state and changes in solar / interplanetary space and Earth's space (such as the magnetosphere, ionosphere, thermosphere, and middle and upper atmosphere) caused by solar activity. It can directly affect space-based / ground-based technology systems, impacting their service reliability and performance. Extreme space weather events, specifically solar geomagnetic storms or flares, ionospheric disturbances or scintillation anomalies, are among the natural disasters facing humanity in the high-tech era. They pose a significant potential threat to the service stability and accuracy of Global Navigation Satellite Systems (GNSS) and have become an important aspect of GNSS operational status monitoring and early warning analysis.

[0003] Traditional GNSS service capability analysis for extreme space weather events typically focuses on disastrous weather events, employing a checklist-style correlation analysis of data anomalies. This involves linking relatively isolated anomalous events, such as solar activity anomalies, ionospheric anomalies, magnetospheric anomalies, and GNSS system service anomalies, to achieve post-event analysis and fault tracing of GNSS system anomalies. However, traditional analytical methods are limited by the sporadic nature of space weather events and the complexity of their mechanisms. They cannot construct a complete physical chain of events—from their occurrence and driving forces to their transmission—based on loosely connected events. This makes it difficult to effectively utilize the mechanisms of accumulation, release, transmission, and impact of space weather events to achieve near real-time analysis of GNSS service capability anomalies, rapid fault diagnosis, and prediction and early warning of potential weather events. Summary of the Invention

[0004] This invention addresses the problem that traditional impact analysis methods are limited by the sporadic nature, low coupling, and complexity of the transmission of effects of space weather events, making it difficult to trace causes and effects and make inferences and predictions based on the overall picture of events. It provides a knowledge graph construction method for impact analysis of extreme space weather events.

[0005] The first aspect of this invention discloses a method for constructing a knowledge graph for analyzing the impact of extreme space weather events, comprising:

[0006] S1, crawl key parameters of space weather events, and clean and process the crawled data to obtain the dataset to be analyzed; key parameters of space weather events include solar activity data, satellite-to-ground link environment data, and GNSS service monitoring data;

[0007] S2, based on the physical transmission mechanism of space-related hazardous weather, performs data mining on the dataset to be analyzed to obtain data correlations, and based on the data correlations, mines, analyzes and evaluates the conditions and trends of parameter influence, thereby obtaining the correlations between events and obtaining the corresponding correlation rules;

[0008] S3. Based on the data correlation, event correlation, and correlation rules obtained in step S2, extract event entities, parameter entities, and attribute entities, construct graph triples, and store them in the Neo4j graph database; the graph triples include: event entity triples, event entity and parameter entity triples, and parameter entity attribute triples.

[0009] S4 uses neovis.js in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connecting to the Neo4j graph database. It calls the graph triple data in the Neo4j graph database by executing custom Cypher query statements, normalizes entities as nodes and relationships between entities as edges, and then renders it on the page in a highly customized way using vis.js, realizing professional knowledge graph visualization for spatial weather event impact analysis.

[0010] Optionally, in step S1, key parameters of space weather events are crawled, including:

[0011] After identifying the anomalous space weather event to be analyzed, the spatiotemporal window of the event's occurrence and impact is locked. Data files or data packets are obtained through web crawling, and key parameters of the space weather event are extracted according to the data protocol.

[0012] Optionally, in step S1, the specific operations of cleaning and treatment include removing damaged data, removing abnormal data, removing duplicate data, filling in logical missing data, and labeling empty data.

[0013] Optionally, in step S2, the Deepwalk algorithm is used to mine the similarity between dataset nodes of space weather events, ionospheric events, magnetospheric events, and GNSS anomaly events.

[0014] Optionally, in step S2, the Deepwalk algorithm is used to mine the similarity between nodes in the datasets of space weather events, ionospheric events, magnetospheric events, and GNSS anomaly events, specifically including:

[0015] Data on space weather events, ionospheric events, magnetospheric events, and GNSS anomalies are organized into corresponding tuples; the tuples are composed of nodes with different parameters, with each parameter considered as a node.

[0016] A loss function F is constructed, and the Deepwalk algorithm model is used to learn the correlation between events.

[0017] Optionally, in step S3, the triples between event entities include: <Solar activity event, drives, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>, <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, conducts to, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>, <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, affects, GNSS service anomaly event> and <Solar activity event, affects, GNSS service anomaly event>;

[0018] The triples between event entities and parameter entities include: <F10.7 index, belongs to, Solar activity event>, <X-ray index, belongs to, Solar activity event>, <TEC index, belongs to, Ionospheric anomaly event>, <ROTI index, belongs to, Ionospheric anomaly event>, <KP index, belongs to, Magnetospheric anomaly event>, <AP index, belongs to, Magnetospheric anomaly event>, <AE index, belongs to, Magnetospheric anomaly event>, <Dst index, belongs to, Magnetospheric anomaly event>, <SYM-H index, belongs to, Magnetospheric anomaly event>, <SNR, belongs to, GNSS service anomaly event>, <Positioning error, belongs to, GNSS service anomaly event>, <Number of visible satellites, belongs to, GNSS service anomaly event>, <Pseudorange residual, belongs to, GNSS service anomaly event>, <Phase residual, belongs to, GNSS service anomaly event>, <PDOP value, belongs to, GNSS service anomaly event> and <Number of cycle slips, belongs to, GNSS service anomaly event>;

[0019] The triples of parameter entity attributes include: <Parameter, intensity, Numerical value>, <Parameter, observation time, Moment value>, <Parameter, duration, Numerical value>, <Parameter, data source, Observation station>, <Parameter, peak value, Numerical value>, <Parameter, spatial position, Latitude and longitude>, <Parameter, positioning accuracy, Numerical value> and <Parameter, signal quality, Numerical value>.

[0020] Optionally, the method further includes:

[0021] Performing quality assessment on the constructed professional knowledge graph; the quality assessment dimensions are divided into three levels: single triple, multiple triples and the overall knowledge base;

[0022] Optimizing data accuracy, graph integrity, graph consistency, graph coherence, graph timeliness and graph display ability according to the quality assessment results.

[0023] A second aspect of the present invention discloses a knowledge graph construction system for analyzing the impact of extreme space weather events, including:

[0024] The first processing module is configured to crawl key parameters of space weather events and clean and process the crawled data to obtain a dataset to be analyzed; the key parameters of space weather events include solar activity data, satellite-to-ground link environment data, and GNSS service monitoring data.

[0025] The second processing module is configured to perform data mining on the dataset to be analyzed based on the physical transmission mechanism of spatial hazardous weather, obtain data correlations, and based on the data correlations, mine, analyze and evaluate the conditions and trends of parameter influence, thereby obtaining the correlations between events and obtaining the corresponding correlation rules.

[0026] The third processing module is configured to extract event entities, parameter entities, and attribute entities based on the obtained data correlation, event correlation, and correlation rules, construct graph triples, and store them in the Neo4j graph database; the graph triples include: event entity triples, event entity and parameter entity triples, and parameter entity attribute triples;

[0027] The fourth processing module is configured to use neovis.js in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connect to the Neo4j graph database, and call the graph triple data in the Neo4j graph database by executing custom Cypher query statements. It normalizes entities as nodes and relationships between entities as edges, and then renders it on the page in a highly customized way using vis.js to realize the visualization of professional knowledge graphs for the analysis of the impact of space weather events.

[0028] A third aspect of this invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps in the knowledge graph construction method for analyzing the impact of extreme space weather events described in the first aspect of this invention.

[0029] A fourth aspect of this invention discloses a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the knowledge graph construction method for analyzing the impact of extreme space weather events described in the first aspect of this invention.

[0030] The beneficial effects of the technical solution described in this invention include:

[0031] This invention focuses on the impact analysis of extreme space weather events and constructs a navigation time-frequency professional knowledge map, which has the following advantages.

[0032] (1) A complete analysis chain from source to end was constructed.

[0033] This invention, based on the mining and analysis of massive space weather events, ionospheric disturbance events, and GNSS anomaly events, summarizes the potential patterns of interaction between various events from loose, random, and sporadic events, and constructs a complete analysis link from source to end. With solar activity as the driving source, it links the physical transmission mechanisms such as ionospheric and magnetosphere disturbances, and applies them to the GNSS operating status and service capabilities. It systematically integrates the mechanism framework of the occurrence, development, and transmission of space hazardous weather, and completes the effective transformation of "data-information-knowledge", providing professional analysis tools for GNSS monitoring and early warning, fault diagnosis, and protection optimization.

[0034] (2) It provides standardized operating tools that run through data collection, mining and analysis, knowledge construction, knowledge integration and knowledge display.

[0035] In summary, this invention, based on a specialized theoretical analysis framework, employs standardized operational tools to integrate data acquisition, mining analysis, knowledge construction, knowledge fusion, and knowledge visualization, achieving efficient connection and transformation from raw data to domain-specific knowledge. This tool embeds big data mining and knowledge construction technologies, providing a precise and scalable analysis framework. Its hierarchical and causal reasoning structure drives the depth and breadth of data processing, ensuring efficient response and accurate decision support in complex environments. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the map construction method for impact analysis of extreme space weather events according to an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of the association mining path based on the Deepwalk algorithm according to an embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of the event association rules agreed upon in accordance with the physical transmission mechanism of space weather events according to an embodiment of the present invention;

[0039] Figure 4 This is a schematic diagram of the knowledge graph quality assessment hierarchy according to an embodiment of the present invention;

[0040] Figure 5 This is a schematic diagram of a specialized knowledge graph for analyzing the impact of extreme space weather events, according to an embodiment of the present invention. Detailed Implementation

[0041] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0042] It should be noted that the following specific embodiments illustrate the implementation of this application, and those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0043] It should be further noted that any specific structure and / or function described herein are illustrative only. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and that two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or practice the method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0044] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. The illustrations only show the components related to this application and are not drawn according to the actual number, shape and size of the components in the actual implementation. In fact, in the actual implementation, the shape, quantity and proportion of various components can be arbitrarily changed, and the layout of the components may also be more complex.

[0045] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0046] This invention integrates and mines a large amount of solar and geomagnetic activity indices, satellite-to-ground link environmental parameters, and monitoring data from IGS, IGMAS, and industry CORS stations during the period of 1999-2022. Focusing on the impact analysis of extreme space weather events, it constructs a top-down, source-to-end professional knowledge map of navigation time and frequency, achieving effective transformation from "data to information to knowledge to intelligence." This forms a theoretical framework and knowledge network for the analysis and forecasting of the impact of severe space weather events, providing a professional analytical tool for BeiDou / GNSS anomaly monitoring and fault diagnosis under space weather conditions. It constructs a complete physical action chain of the occurrence, development, transmission, and impact of severe space weather events, forming a knowledge-based impact analysis framework for severe space weather events, providing support for the systematic and intelligent analysis and evaluation of space weather events.

[0047] The workflow of the method of the present invention is as follows: Figure 1 As shown, its technical solution is as follows:

[0048] (1) Crawling and cleaning of key parameters for space weather events

[0049] Key parameters of space weather events refer to quantitative data reflecting the state and changes of events, from the driving source to the manifestation, including solar activity data, satellite-to-ground link environmental data, and GNSS service monitoring data. All of these key parameters can be crawled from authoritative / official websites (see Table 1) using the Scrapy framework written in Python to obtain highly reliable and trustworthy data.

[0050] Table 1. List of websites for obtaining key parameters of space weather events

[0051]

[0052] Knowledge is carried by events. Data crawling is driven by abnormal space weather events. After anchoring the event to be analyzed, the spatiotemporal window of the event occurrence and impact is locked. Data files or data packets are obtained through network crawling. Key parameters of space weather events are extracted according to data protocols. The specific key parameters and their meanings are shown in Table 2.

[0053] Table 2 Key Parameters for Space Weather Event Analysis

[0054]

[0055]

[0056] It is important to note that after the key parameter extraction process is completed, the key parameters need to be standardized and cleaned according to data attributes, range, and format protocols. Specific operations include, but are not limited to: removing damaged data, removing abnormal data, removing duplicate data, filling in logical missing data, and labeling empty data, to ensure data availability, integrity, and validity, create a high-quality dataset (dataset to be analyzed), and thus promote the accuracy of subsequent data correlation analysis.

[0057] (2) Big data mining and correlation analysis based on space weather events

[0058] Following the physical transmission mechanism of space-borne hazardous weather, the Deepwalk algorithm is used to integrate and mine key parameters such as solar activity, satellite-to-ground link environment, and GNSS service monitoring during the period from 1999 to 2022, especially the data correlation of anomalous parameters. Based on the data correlation of key parameters, the conditions and trends of parameter influence are analyzed and evaluated, thereby obtaining the correlation between events and sorting out the driving hierarchy and impact transmission between events.

[0059] Event correlation analysis employs the Deepwalk algorithm to mine similarities between nodes in datasets including space weather events, ionospheric events, magnetospheric events, and GNSS anomalies. The mining path is shown below. Figure 2 As shown, the analysis reveals the transitive relationships and potential patterns between events.

[0060] The specific analysis method is as follows: The datasets to be associated are constructed into corresponding tuples, with each tuple consisting of nodes with different parameters. Each parameter is considered a node. The theoretical probability of node a generating node b is:

[0061]

[0062] In the formula, u b v a Let represent the spatial embedding vectors of node b and node a, respectively; k represent all nodes adjacent to node a; X represent the set of all nodes; u k This represents the spatial embedding vector of node k, and the superscript T indicates matrix transpose.

[0063] The actual calculated probability is:

[0064]

[0065] In the formula, ε ab ε is the edge weight between node a and node b; ak Let Y be the edge weights of nodes a and k; Y represents the set of nodes adjacent to a in the actual sample, derived from the actual statistical tuple library. The learning process aims to make the theoretical probability as close as possible to the actual probability. The loss function G is defined as:

[0066]

[0067] In the formula, KuL represents the KL divergence, expressed as: Where n represents the possible values ​​of random variables p and q. By using negative sampling to reduce computation, the loss function becomes F:

[0068]

[0069] In the formula, σ represents the sigmoid function, and L represents the number of negative samples used.

[0070] Based on the correlation analysis of the DeepWalk algorithm and combined with the physical transmission mechanism of space weather events, the correlation rules are clarified as follows: Figure 3 As shown, space weather events and their associated transmission events are constructed as a graph structure. The nodes of the graph represent events, and the edges represent the direct or indirect influence relationships between events. The frequency of event occurrence and the temporal relationship between events are used as weights or edge attributes.

[0071] (3) Knowledge extraction, construction and storage based on the transfer mechanism of space weather

[0072] According to the data / event correlation and association rules mined and analyzed in step (2), entities such as events, parameters, and attributes are extracted. Based on event association rules, parameter subordination relationships, parameter attribute characteristics, etc., graph triple is constructed, and the main forms include: <entity, relationship, entity>, <entity, attribute, value>, etc. Construct a knowledge structured representation among events, between events and parameters, and parameter attributes, and store it in the Neo4j graph database according to the above structure to achieve an orderly and efficient storage of basic attribute knowledge, association knowledge, event knowledge, data resource knowledge, etc.

[0073] It involves the construction of triples, specifically including the following three aspects:

[0074] Construction of triples between event entities

[0075] For the knowledge graph constructed based on the事理 logic of the evolution law and pattern change between events, with typical space weather events as the center, first construct event evolution transfer rules for event entities. Specific rules for events are as follows:

[0076] a. <Solar activity event, drives, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>;

[0077] b. <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, conducts to, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>;

[0078] c. <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, affects, GNSS service anomaly event>;

[0079] d. <Solar activity event, affects, GNSS service anomaly event>.

[0080] Construction of triples between event entities and parameter entities

[0081] Taking various typical events as the center, construct the triple form between event entities and key parameter entities. The specific form is:

[0082] a. According to the key parameters of the mined solar activity events, construct a set of parameter entities for solar activity parameters, including but not limited to (extensible): F10.7 index, X-ray index. The constructed triple form is:

[0083] <F10.7 index, belongs to, Solar activity event>

[0084] <X-ray index, affiliation, solar activity event>

[0085] b. Based on the key parameters of the ionospheric anomaly events mined, construct an entity set of ionospheric anomaly class parameters. The parameters include but are not limited to (extensible): TEC index, ROTI index. The constructed triple form is:

[0086] <TEC index, affiliation, ionospheric anomaly event>

[0087] <ROTI index, affiliation, ionospheric anomaly event>

[0088] c. Based on the key parameters of the magnetospheric anomaly events mined, construct an entity set of magnetospheric anomaly class parameters. The parameters mainly include but are not limited to (extensible): KP index, AP index, AE index, Dst index, SYM-H index. The constructed triple form is:

[0089] <KP index, affiliation, magnetospheric anomaly event>

[0090] <AP index, affiliation, magnetospheric anomaly event>

[0091] <AE index, affiliation, magnetospheric anomaly event>

[0092] <Dst index, affiliation, magnetospheric anomaly event>

[0093] <SYM-H index, affiliation, magnetospheric anomaly event>

[0094] d. Based on the key parameters of the GNSS service anomaly events mined, construct an entity set of GNSS service anomaly class parameters. The parameters mainly include but are not limited to (extensible): SNR, positioning error, number of visible satellites, pseudorange residual, phase residual, PDOP value, cycle slip count. The constructed triple form is:

[0095] <SNR, affiliation, GNSS service anomaly event>

[0096] <Positioning error, affiliation, GNSS service anomaly event>

[0097] <Number of visible satellites, affiliation, GNSS service anomaly event>

[0098] <Pseudorange residual, affiliation, GNSS service anomaly event>

[0099] <Phase residual, affiliation, GNSS service anomaly event>

[0100] <PDOP value, affiliation, GNSS service anomaly event>

[0101] <Cycle slip count, affiliation, GNSS service anomaly event>

[0102] Parameter entity attribute triple construction

[0103] It provides a complete description of the parameter entity in terms of time, space, trend, and changes, and supports event reasoning, prediction, and early warning based on knowledge graphs. The attribute description of the specific parameter entity is represented as follows:

[0104] <Parameters, Intensity, Value>

[0105] <Parameters, observation time, time value>

[0106] <Parameter, duration, value>

[0107] <Parameters, data source, observation station>

[0108] <Parameter, peak value>

[0109] <Parameters, spatial location, latitude and longitude>

[0110] <Parameters, positioning accuracy, numerical value>

[0111] <Parameters, signal quality, numerical value>

[0112] (4) Visualization of professional knowledge graphs for space weather event impact analysis

[0113] For the structured graph data constructed and stored in step (3), neovis.js is used in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connected to the Neo4j graph database. By executing a custom cypher query statement, the graph data in the Neo4j graph database is called, entities are standardized as nodes, and the relationships between entities are standardized as edges. Then, the data is rendered on the page in a highly customized way using vis.js to form an intuitive knowledge network, so as to professionally represent the transmission characteristics and impact mechanisms of space weather events.

[0114] In terms of detailed display, different shapes, colors, and sizes are used to distinguish nodes according to their type and importance, thereby enhancing the intuitiveness of information expression;

[0115] In terms of visual noise reduction, the visibility of the map is further improved by reducing redundant information and optimizing the layout (such as using multi-sided hashing for opposite edges to reduce edge overlap).

[0116] In terms of interaction design, it provides users with interactive functions with the knowledge graph, such as zooming in and out, dragging nodes, searching, filtering, and clicking to view details, allowing users to explore the knowledge graph structure according to their needs for in-depth analysis.

[0117] (5) Quality assessment of professional knowledge graphs

[0118] The quality of the constructed knowledge graph was evaluated, with evaluation dimensions divided into three levels: single triples, multiple triples, and the overall knowledge base. Figure 4 As shown, professional evaluation guides the optimization and completion of the knowledge graph, thereby improving the quality of the professional knowledge graph and its ability to be applied in various scenarios.

[0119] At the single triplet level, the focus is on evaluation indicators such as accuracy, completeness, consistency, connectivity, redundancy, timeliness, and availability.

[0120] a. Accuracy: Precision and recall are used to measure the accuracy of data in the graph.

[0121] Accuracy:

[0122]

[0123] Recall rate:

[0124]

[0125] Wherein, TP (True Positive) represents the number of correctly labeled entities or relations, FP (False Positive) represents the number of incorrectly labeled entities or relations, and FN (False Negative) represents the number of unlabeled entities or relations that should exist.

[0126] b. Completeness: The completeness of information in the map is measured using the coverage metric, ensuring that the core content of the target domain is covered. The corresponding formula is:

[0127]

[0128] Among them, |E KG | represents the number of entities or relations contained in the knowledge graph, |E total | indicates the number of all entities or relationships that should be included in the target domain.

[0129] c. Consistency: The consistency score is used to quantify the proportion of consistency in the map. The higher the proportion, the better the consistency of the map. The corresponding formula is:

[0130]

[0131] Among them, |C conflict | indicates the number of conflicting relationships or entities, |C total | indicates the total number of relations or entities.

[0132] d. Connectivity: Average Path Length (APL) is used to evaluate the connectivity between entities in the knowledge graph. Shorter path lengths generally indicate a more compact graph structure. The corresponding formula is:

[0133]

[0134] Where N represents the number of entities in the knowledge graph, and d(i,j) represents the shortest path length between entity i and entity j.

[0135] e. Redundancy: The redundancy rate is used to quantify the proportion of duplicate data in the knowledge graph. The corresponding formula is:

[0136]

[0137] Among them, |E dup | represents the number of repeated entities or relations, |E total | indicates the total number of entities or relationships.

[0138] f. Timeliness: The timeliness of the map is quantified using update latency. The smaller the update latency, the higher the timeliness of the map. The corresponding formula is:

[0139]

[0140] Among them, t KG，i t represents the update time of the i-th entity or relation in the knowledge graph. source，i This indicates the update time of the corresponding data source.

[0141] (6) Optimization of knowledge graph for space weather event impact analysis

[0142] Based on the quality assessment results of the knowledge graph, feedback is provided to the graph construction system to guide the optimization / supplementation of specialized knowledge graphs for comprehensive space weather event impact analysis, thereby improving the overall quality and application effectiveness of the graphs.

[0143] Based on the map quality assessment indicators, the map optimization and completion work mainly includes the following aspects:

[0144] a. Data accuracy optimization: By identifying inaccuracies or incompleteness in knowledge graph entities or relationships, prioritize optimizing data source quality, introduce high-precision data sources, or perform deep cleaning and governance such as data verification and filtering to ensure the credibility of data assets.

[0145] b. Knowledge Graph Completeness Optimization: To address the issue of insufficient coverage of the knowledge graph identified in the assessment, the graph entities and relationship sets will be enriched by expanding data sources. For example, more typical space weather events and impact events will be introduced to ensure that the knowledge graph fully covers space weather events and their consequences, thereby enhancing the professional analysis capabilities for complex event chains.

[0146] c. Graph Consistency Optimization: Establish strict constraints and rules to ensure that the relationships and attributes in the graph are logically consistent. By using constraint satisfaction problem (CSP) or logical reasoning techniques, detect and repair conflicts or inconsistencies in the graph to improve the internal consistency of the graph.

[0147] d. Graph coherence optimization: To address the issues of isolated nodes or poor connectivity in the graph, key bridging entities or relationships are added to ensure that an effective connection network is formed between various entities, and that each event can form a complete causal chain in the graph, thereby improving the overall connectivity and information flow of the graph.

[0148] e. Map Timeliness Optimization: In response to the dynamic changes in space weather events, an automated update mechanism is introduced to periodically refresh map content based on real-time data sources, ensuring the timeliness of map applications and event response capabilities, and providing support for GNSS monitoring, early warning, and emergency optimization.

[0149] f. Optimization of Knowledge Graph Display Capabilities: Addressing issues such as ambiguity, lack of intuitiveness, and ineffability in knowledge graph display, a graphical interface and dynamic interactive display tools are used to enable users to intuitively understand the impact paths and extent of space weather events, strengthening the role of knowledge graphs in decision support. The optimized knowledge graph is as follows: Figure 5 As shown.

[0150] This invention also provides a knowledge graph construction system for analyzing the impact of extreme space weather events, comprising:

[0151] The first processing module is configured to crawl key parameters of space weather events and clean and process the crawled data to obtain a dataset to be analyzed; the key parameters of space weather events include solar activity data, satellite-to-ground link environment data, and GNSS service monitoring data.

[0152] The second processing module is configured to perform data mining on the dataset to be analyzed based on the physical transmission mechanism of spatial hazardous weather, obtain data correlations, and based on the data correlations, mine, analyze and evaluate the conditions and trends of parameter influence, thereby obtaining the correlations between events and obtaining the corresponding correlation rules.

[0153] The third processing module is configured to extract event entities, parameter entities, and attribute entities based on the obtained data correlation, event correlation, and correlation rules, construct graph triples, and store them in the Neo4j graph database; the graph triples include: event entity triples, event entity and parameter entity triples, and parameter entity attribute triples;

[0154] The fourth processing module is configured to use neovis.js in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connect to the Neo4j graph database, and call the graph triple data in the Neo4j graph database by executing custom Cypher query statements. It normalizes entities as nodes and relationships between entities as edges, and then renders it on the page in a highly customized way using vis.js to realize the visualization of professional knowledge graphs for the analysis of the impact of space weather events.

[0155] The present invention also provides an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps in the above-described method. The electronic device includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. The processor of this electronic device provides computing and control capabilities.

[0156] The above examples are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for constructing a knowledge graph for the impact analysis of extreme space weather events, characterized in that, include: S1, crawl key parameters of space weather events, and clean and process the crawled data to obtain the dataset to be analyzed; key parameters of space weather events include solar activity data, satellite-to-ground link environment data, and GNSS service monitoring data; S2, based on the physical transmission mechanism of space-related hazardous weather, performs data mining on the dataset to be analyzed to obtain data correlations, and based on the data correlations, mines, analyzes and evaluates the conditions and trends of parameter influence, thereby obtaining the correlations between events and obtaining the corresponding correlation rules; S3. Based on the data correlation, event correlation, and correlation rules obtained in step S2, extract event entities, parameter entities, and attribute entities, construct graph triples, and store them in the Neo4j graph database; the graph triples include: event entity triples, event entity and parameter entity triples, and parameter entity attribute triples; S4 uses neovis.js in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connecting to the Neo4j graph database. It calls the graph triple data in the Neo4j graph database by executing custom Cypher query statements, normalizes entities as nodes and relationships between entities as edges, and then renders it on the page in a highly customized way using vis.js, realizing professional knowledge graph visualization for spatial weather event impact analysis.

2. The method as described in claim 1, characterized in that, In step S1, key parameters of space weather events are crawled, including: After identifying the anomalous space weather event to be analyzed, the spatiotemporal window of the event's occurrence and impact is locked. Data files or data packets are obtained through web crawling, and key parameters of the space weather event are extracted according to the data protocol.

3. The method as described in claim 1, characterized in that, In step S1, the specific operations of cleaning and treatment include removing damaged data, removing abnormal data, removing duplicate data, filling in logical missing data, and labeling empty data.

4. The method as described in claim 1, characterized in that, In step S2, the Deepwalk algorithm is used to mine the similarity between dataset nodes of space weather events, ionospheric events, magnetospheric events, and GNSS anomaly events to obtain the correlation between events.

5. The method as described in claim 4, characterized in that, In step S2, the Deepwalk algorithm is used to mine the similarity between nodes in the datasets of space weather events, ionospheric events, magnetospheric events, and GNSS anomaly events to obtain the correlation between events, specifically including: Data on space weather events, ionospheric events, magnetospheric events, and GNSS anomalies are organized into corresponding tuples; the tuples are composed of nodes with different parameters, with each parameter considered as a node. A loss function F is constructed, and the Deepwalk algorithm model is used to learn the correlation between events.

6. The method as described in claim 1, characterized in that, In step S3, the triples between event entities include: <Solar activity event, drives, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>, <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, conducts to, Ionospheric anomaly event>, <Ionospheric anomaly event, affects, GNSS service anomaly event>, <Solar activity event, drives, Magnetospheric anomaly event>, <Magnetospheric anomaly event, affects, GNSS service anomaly event>, and <Solar activity event, affects, GNSS service anomaly event>; The triples between event entities and parameter entities include: <F10.7 index, belongs to, Solar activity event>, <X-ray index, belongs to, Solar activity event>, <TEC index, belongs to, Ionospheric anomaly event>, <ROTI index, belongs to, Ionospheric anomaly event>, <KP index, belongs to, Magnetospheric anomaly event>, <AP index, belongs to, Magnetospheric anomaly event>, <AE index, belongs to, Magnetospheric anomaly event>, <Dst index, belongs to, Magnetospheric anomaly event>, <SYM-H index, belongs to, Magnetospheric anomaly event>, <SNR, belongs to, GNSS service anomaly event>, <Positioning error, belongs to, GNSS service anomaly event>, <Number of visible satellites, belongs to, GNSS service anomaly event>, <Pseudorange residual, belongs to, GNSS service anomaly event>, <Phase residual, belongs to, GNSS service anomaly event>, <PDOP value, belongs to, GNSS service anomaly event>, and <Number of cycle slips, belongs to, GNSS service anomaly event>; The triples of parameter entity attributes include: <Parameter, intensity, numerical value>, <Parameter, observation time, time value>, <Parameter, duration, numerical value>, <Parameter, data source, observation station>, <Parameter, peak value, numerical value>, <Parameter, spatial position, longitude and latitude>, <Parameter, positioning accuracy, numerical value>, and <Parameter, signal quality, numerical value>.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: S4, performing quality assessment on the constructed professional knowledge graph; the quality assessment dimensions are divided into three levels: single triple, multiple triples, and the overall knowledge base; S5, optimizing data accuracy, graph integrity, graph consistency, graph coherence, graph timeliness, and graph display ability according to the quality assessment results.

8. A knowledge graph construction system for analyzing the impact of extreme space weather events, characterized in that, It includes: The first processing module is configured to crawl the key parameters of space weather events and clean and manage the crawled data to obtain a dataset to be analyzed; the key parameters of space weather events include solar activity data, space-ground link environment data, and GNSS service monitoring data; The second processing module is configured to perform data mining on the dataset to be analyzed based on the physical transmission mechanism of space catastrophic weather, obtain data correlations, and based on the data correlations, mine, analyze, and evaluate the conditions and trends affected by parameters, and then obtain the correlations between events and obtain corresponding association rules; The third processing module is configured to extract event entities, parameter entities, and attribute entities based on the obtained data correlations, event correlations, and correlation rules, construct graph triples, and store them in the Neo4j graph database; the graph triples include: event entity triples, event entity and parameter entity triples, and parameter entity attribute triples. The fourth processing module is configured to use neovis.js in conjunction with JavaScript to drive the visualization libraries Neo4j and vis.js, directly connect to the Neo4j graph database, and call the graph triple data in the Neo4j graph database by executing custom Cypher query statements. It normalizes entities as nodes and relationships between entities as edges, and then renders it on the page in a highly customized way using vis.js to realize the visualization of professional knowledge graphs for the analysis of the impact of space weather events.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps in the knowledge graph construction method for impact analysis of extreme space weather events according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the knowledge graph construction method for impact analysis of extreme space weather events according to any one of claims 1 to 7.

Images