Space-time-frequency multi-domain electronic intelligence knowledge graph construction method

Abstract

The invention provides a knowledge graph construction method containing time-space-frequency multi-domain electronic intelligence, which adopts a mode of combining a distributed data storage system and a non-traditional NOSQL non-relational database to construct an electronic intelligence representation ontology model containing multi-dimensional information such as a time domain, a frequency domain, a space domain, a parameter domain and the like. Electronic intelligence original data recorded by electronic reconnaissance equipment and processed and refined electronic intelligence are respectively filled into a distributed data storage system and a non-traditional NOSQL non-relational database, and through automatic import and knowledge reasoning and complementing based on a rule model and a semantic model. According to the method, the electronic intelligence knowledge graph which can be used for carrying out big data analysis, research and judgment on electronic intelligence and situations can be constructed.

Description

Technical Field

[0001] The present invention relates to knowledge graph technology, and in particular to a method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph. Background Art

[0002] Knowledge graphs describe concepts, entities and their relationships in the objective world in a structured form, and express information from books, documents, the Internet, etc. in a form closer to the human cognitive world, providing a way to better organize, manage and understand the existing massive information and knowledge. Knowledge graph technology refers to the technology for establishing and applying knowledge graphs, which is an interdisciplinary study that integrates cognitive computing, knowledge representation and reasoning, information retrieval and extraction, natural language processing and semantic Web, data mining and machine learning, and other methods.

[0003] The current knowledge graph technology mainly studies the expression of knowledge and the automatic construction technology of knowledge graphs. The knowledge graphs it constructs basically contain factual and relational knowledge, but lack specialized knowledge networks for the sub-field of electronic countermeasures. Therefore, it cannot support the needs of electronic countermeasures big data analysis.

[0004] In the prior art, there are generally two ways to construct a knowledge graph: one is based on a basic knowledge base, using NLP (natural language processing) technology to extract triples to construct it, which is generally used for Internet searches. The other requires users to pre-define the knowledge graph, which is generally used in some specific fields (such as libraries). In terms of data storage, the widely used relational database has many advantages, such as standardized design format, simple reading and writing, and strong usability. However, facing the massive heterogeneous data contained in the electromagnetic field knowledge, the performance of traditional relational databases is slightly insufficient.

[0005] Electronic reconnaissance data has a large amount of data, multiple data sources, and complex processing procedures. It takes a lot of time and experience to mine and extract valuable electronic intelligence from electronic reconnaissance data intercepted by a variety of electronic reconnaissance equipment, which is very difficult. It is very difficult to discover the relationship between multi-source electronic intelligence and summarize valuable intelligence. Summary of the invention

[0006] The purpose of the present invention is to propose a method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph.

[0007] The technical solution to achieve the purpose of the present invention is: a method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph, comprising the following steps:

[0008] (1) Knowledge graph ontology model construction

[0009] In the data layer of the knowledge graph, a distributed data storage system is used to store the original data files recorded by the electronic reconnaissance equipment, and a unique access URI of the distributed storage system is generated. The URI will be stored as an entity in the non-relational database; the ontology layer is managed by a labeled ontology library, and the ontology library's support for axioms, rules and constraints is used to standardize entities, relationships, entity types and attributes; the entity layer stores multi-dimensional electronic intelligence entity information including time domain, frequency domain, space domain and parameter domain and their mutual relationships;

[0010] (2) Construction of electronic intelligence entities

[0011] By using real-time stream processing and batch post-processing based on recorded data, the real-time reconnaissance results of electronic reconnaissance equipment or recorded reconnaissance data are imported into the distributed data storage system, and after pre-processing, data cleaning, automatic labeling and compilation, the electronic intelligence entity is constructed;

[0012] (3) Knowledge reasoning completion based on rule model and semantic model

[0013] Extract knowledge from text data, structured data, and expert knowledge data in the field of radar countermeasures according to predetermined rules, and form a corpus of subject-predicate-object structure after segmentation and merging. Construct a synonym index of the vocabulary in the corpus, including words and phrases, and construct two types of knowledge triples: (subject-predicate-object) and (vocabulary-synonym-synonym). After data cleaning and ontology matching and association according to the knowledge graph ontology model, import it into a non-relational database to form a semantic association model.

[0014] In the process of reconnaissance data processing, after the electronic intelligence entity is generated, a set of descriptive words is extracted from the newly generated entity and matched with the vocabulary and its synonyms and near-synonyms in the semantic association model to establish a relationship. At the same time, it is used as the input of the semantic association model to query the ontologies and entities associated with it, and the number of query hits is counted. When the number of hits reaches the threshold, the relationship with the corresponding ontology and entity is constructed to realize knowledge reasoning completion based on the rule model.

[0015] Furthermore, (1) the knowledge graph ontology model is constructed. The specific method is as follows:

[0016] Step 1.1: Build the data layer, deploy a distributed data storage system, and ensure that the distributed data storage system has high availability and scalability;

[0017] Step 1.2: Distribute and store the raw data files collected by the electronic reconnaissance equipment, and generate a unique access URI based on UUID for each file to ensure the uniqueness and traceability of the data;

[0018] Step 1.3: Build the ontology layer, define the concepts, attributes, and relationships of the ontology, and ensure the logical consistency and integrity of the ontology layer;

[0019] Step 1.4: Create a labeled ontology library to manage and maintain the ontology layer information of the knowledge graph;

[0020] Step 1.5: Store axioms, rules, and constraints in the ontology repository to regulate entities, relations, entity types, and attributes;

[0021] Step 1.6: Build the entity layer, define entity types and attributes, and the relationships between entities, and ensure that the entity layer can store multi-dimensional information;

[0022] Step 1.7: Using a top-down construction approach, extract ontology and pattern information from the revised formatted electronic intelligence documents and add them to the knowledge base;

[0023] Step 1.8: Construct a radar radiation source knowledge ontology library, with radar model, purpose, technical system, operating band, radar carrying platform, radar manufacturing country, radar country, and radar user unit as the radar model knowledge ontology library of the knowledge graph; with radar reconnaissance indicator system, including EDW and PDW sequences as the radar signal knowledge ontology library, and the two are combined into the radar radiation source knowledge ontology library;

[0024] Step 1.9: Construct an electronic reconnaissance equipment knowledge ontology library, with the electronic reconnaissance equipment model, technical system, adaptive frequency band, instantaneous bandwidth, instantaneous airspace coverage, electronic reconnaissance equipment carrying platform, electronic reconnaissance equipment production country, electronic reconnaissance equipment country of origin, and electronic reconnaissance equipment user as the electronic reconnaissance equipment knowledge ontology library;

[0025] Step 1.10: Construct a multi-dimensional electronic intelligence knowledge ontology library, with the time, location, track and equipment used in the reconnaissance mission, the pulse parameters and radiation source parameters of the reconnaissance target, the target model, the platform to which it belongs, the target direction finding, and the positioning information as the knowledge graph of the multi-dimensional electronic intelligence knowledge ontology library.

[0026] Furthermore, (2) the electronic intelligence entity is constructed, and the specific method is as follows:

[0027] Step 2.1: Receive and process the reconnaissance results of the electronic reconnaissance equipment in real time through real-time stream processing technology;

[0028] Step 2.2: Batch post-processing of the admitted reconnaissance data, including data preprocessing, cleaning and labeling;

[0029] Step 2.3: Import the processed data into the distributed data storage system to ensure persistent storage of the data;

[0030] Step 2.4: According to the parameter system defined by the knowledge graph ontology model, construct the radar radiation source entity class, electronic reconnaissance equipment entity class and multi-dimensional electronic intelligence related entity class to ensure that the entity class corresponds to the parameter system of the ontology model one by one to ensure the accuracy and consistency of the data.

[0031] Furthermore, (3) knowledge reasoning completion based on rule model and semantic model, the specific method is as follows:

[0032] Step 3.1: Formulate inference rules, including conditional and conclusion parts;

[0033] Step 3.2: Encode the inference rules into executable logical expressions and store them in the rule base, which is managed using the Drools rule engine.

[0034] Step 3.3: Combined with the Spark Streaming framework, build a real-time reasoning processing system based on data streams to support dynamic intelligence analysis;

[0035] Step 3.4: Execute the inference rules to generate new knowledge triples through logical reasoning, expanding the depth and breadth of the knowledge graph;

[0036] Step 3.5: Extract knowledge from radar countermeasure field text data, compiled structured data, and organized expert knowledge data, build a corpus, and use NLP technology to perform word segmentation, part-of-speech tagging, and dependency analysis on text data to form a structured knowledge expression;

[0037] Step 3.6: Build a semantic index, including a word vector model and a synonym dictionary, to support the calculation of semantic similarity;

[0038] Step 3.7: Identify potential relationships between entities by matching entity descriptions with semantic indices, and use these relationships as input for reasoning;

[0039] Step 3.8: Set a threshold for the number of query hits and dynamically adjust the threshold based on the importance of the inference results to balance the accuracy and coverage of the inference;

[0040] Step 3.9: Use graph neural network to predict links in knowledge graph and realize context-based entity relationship reasoning;

[0041] Step 3.10: When the inference result reaches the threshold preset in step 3.8, the knowledge graph is automatically updated and new entity relationships are added to achieve continuous proliferation of the knowledge graph.

[0042] A space-time and frequency multi-domain electronic intelligence knowledge graph construction system implements the space-time and frequency multi-domain electronic intelligence knowledge graph construction method to achieve space-time and frequency multi-domain electronic intelligence knowledge graph construction.

[0043] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for constructing a spatiotemporal-frequency multi-domain electronic intelligence knowledge graph is implemented to realize the construction of a spatiotemporal-frequency multi-domain electronic intelligence knowledge graph.

[0044] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the method for constructing a spatiotemporal, frequency-multi-domain electronic intelligence knowledge graph to achieve the construction of a spatiotemporal, frequency-multi-domain electronic intelligence knowledge graph.

[0045] Compared with the prior art, the present invention has the following significant advantages: it constructs a multidimensional electronic intelligence knowledge graph based on radar reconnaissance data, can import and process the reconnaissance results and recorded data in the daily work of the reconnaissance equipment, and generate a multidimensional electronic intelligence knowledge graph; it also uses a rule-based reasoning model and a semantic-based association model to perform knowledge reasoning and completion on the constructed knowledge graph, and mines the hidden connection between new intelligence and the original intelligence in the knowledge graph; it comprehensively utilizes the acquired information to establish a more complete electronic intelligence knowledge graph for electronic intelligence big data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is the knowledge graph ontology model construction process.

[0047] Figure 2 It is the process of building electronic reconnaissance intelligence entity.

[0048] Figure 3 It is a knowledge reasoning and completion process based on rule models and semantic models. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] The present invention adopts a combination of a distributed data storage system and a non-traditional NOSQL non-relational database. First, an electronic intelligence representation ontology model containing multi-dimensional information such as time domain, frequency domain, spatial domain, and parameter domain is constructed. The original electronic intelligence data recorded by the electronic reconnaissance equipment and the processed and refined electronic intelligence are respectively filled into the distributed data storage system and the non-traditional NOSQL non-relational database. Through automatic import and knowledge reasoning and completion based on rule models and semantic models, an electronic intelligence knowledge graph that can be used for big data analysis and judgment of electronic intelligence and situation is constructed.

[0051] The present invention is divided into three parts: knowledge graph ontology model construction, electronic reconnaissance entity construction, and knowledge reasoning completion based on rules and semantic models:

[0052] (1) Knowledge graph ontology model construction

[0053] The electronic intelligence knowledge graph can be logically divided into the data layer, the ontology layer, and the entity layer. In the data layer of the knowledge graph, a distributed data storage system is used to store the original data files recorded by the electronic reconnaissance equipment, and a unique access URI of the distributed storage system is generated. The URI will be stored as an entity in the non-relational database. The ontology layer is the core of the knowledge graph. The designed knowledge framework ontology is stored in the ontology layer. Usually, a labeled ontology library is used to manage the ontology layer of the knowledge graph. The ontology library's support for axioms, rules, and constraints is used to standardize the connection between entities, relationships, entity types, and attributes. The entity layer is the main body of the knowledge graph, which is used to construct and store electronic intelligence entity information and their mutual relationships containing multi-dimensional information such as time domain, frequency domain, spatial domain, and parameter domain within the scope of the knowledge framework ontology.

[0054] The knowledge graph ontology model construction methods include top-down and bottom-up construction methods. The present invention uses a top-down construction method to extract ontology and pattern information from high-quality information and add it to the knowledge base.

[0055] The knowledge ontology related to the electronic intelligence knowledge graph mainly includes three categories:

[0056] Radar emitter knowledge ontology

[0057] Electronic reconnaissance equipment knowledge ontology

[0058] ·Multi-dimensional electronic intelligence knowledge ontology in time domain, frequency domain, space domain and parameter domain

[0059] The radar model knowledge ontology library uses radar model, purpose, technical system, operating band, radar carrying platform, radar manufacturing country, radar country, and radar user unit as the knowledge graph; the radar reconnaissance indicator system, including EDW (radiator descriptor) and typical PDW (pulse descriptor) sequences, is used as the radar signal knowledge ontology library; the two are combined into the radar radiation source knowledge ontology library.

[0060] The electronic reconnaissance equipment knowledge ontology database is composed of electronic reconnaissance equipment models, technical systems, applicable frequency bands, instantaneous bandwidth, instantaneous airspace coverage, electronic reconnaissance equipment carrying platforms, electronic reconnaissance equipment producing countries, electronic reconnaissance equipment country of origin, and electronic reconnaissance equipment using units.

[0061] A multi-dimensional electronic intelligence knowledge ontology library with the reconnaissance mission time, location, track and equipment used, reconnaissance target pulse parameters and radiation source parameters (including EDW and all intercepted PDW), target model, platform, target direction finding, positioning information, etc. as the knowledge graph.

[0062] (2) Construction of electronic reconnaissance intelligence entity

[0063] By using real-time stream processing and batch post-processing based on recorded data, the real-time reconnaissance results of electronic reconnaissance equipment or the recorded reconnaissance data are imported into the distributed data storage system, and after preprocessing, data cleaning, automatic labeling and compilation, the electronic reconnaissance intelligence entity is constructed.

[0064] Corresponding to the knowledge graph ontology model, the constructed electronic reconnaissance intelligence entities include the following categories:

[0065] Radar radiation source entity class

[0066] Electronic reconnaissance equipment entity

[0067] ·Multi-dimensional electronic intelligence related entity classes in time domain, frequency domain, space domain, and parameter domain

[0068] The parameter system of each entity class corresponds one-to-one to the parameter system of the corresponding ontology model.

[0069] (3) Knowledge reasoning completion based on rule model and semantic model

[0070] Knowledge graph reasoning is to use logical thinking ability based on existing knowledge to draw some implicit conclusions. In the knowledge graph, the existing knowledge (triples) in the graph are used to infer some new relationships between entities or attributes of entities (triples).

[0071] The present invention adopts a rule-based reasoning model and a semantic-based association model to perform knowledge reasoning and completion on the constructed knowledge graph.

[0072] A set of entity and relationship reasoning rule models are predefined according to the ontology model and imported into the non-relational database. Using big data stream processing technology, an automatic reasoning process for reconnaissance data streams is constructed to achieve knowledge reasoning completion based on rule models.

[0073] Knowledge is extracted from text data, structured data, and expert knowledge data in the field of radar confrontation according to predetermined rules. After segmentation and merging, a corpus of subject-predicate-object structure is formed, and synonyms and synonyms indexes of vocabulary (including words and phrases) in the corpus are constructed to construct two types of knowledge triples: (subject-predicate-object) and (vocabulary-synonyms-synonyms). After data cleaning and ontology matching association according to the knowledge graph ontology model, it is imported into a non-relational database to form a semantic association model. In the process of reconnaissance data processing, after the electronic reconnaissance intelligence entity is generated, a set of descriptive words is extracted from the newly generated entity, and matched with the vocabulary and its synonyms and synonyms in the semantic association model to establish a relationship. At the same time, it is used as the input of the semantic association model to query the associated ontology and entity, and the query hit count is counted. When the hit count reaches the threshold, the relationship with the corresponding ontology and entity is automatically constructed. The threshold should be dynamically adjusted according to the size of the semantic association model.

[0074] Example

[0075] A method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph includes the following steps:

[0076] (1) Knowledge graph ontology model construction

[0077] Step 1.1: Deploy a distributed data storage system to ensure high availability and scalability.

[0078] Step 1.2: The original data files recorded by the electronic reconnaissance equipment are distributed and a unique access URI based on UUID is generated for each file to ensure the uniqueness and traceability of the data.

[0079] Step 1.3: Design the knowledge framework ontology, including defining the concepts, attributes, relationships, etc. of the ontology layer to ensure the logical consistency and integrity of the ontology layer.

[0080] Step 1.4: Create a labeled ontology library to manage and maintain the ontology layer information of the knowledge graph.

[0081] Step 1.5: Store axioms, rules, and constraints in the ontology repository to regulate the connections between objects such as entities, relations, entity types, and attributes.

[0082] Step 1.6: Build the entity layer, define entity types and attributes, and the relationships between entities to ensure that the entity layer can store multi-dimensional information.

[0083] Step 1.7: Use a top-down construction approach to extract ontology and pattern information from high-quality information sources and add it to the knowledge base.

[0084] Step 1.8: Build a knowledge ontology library of radar radiation sources, including radar model, purpose, technical system and other attributes.

[0085] Step 1.9: Construct the electronic reconnaissance equipment knowledge ontology library, including equipment model, technical system, applicable frequency band and other attributes.

[0086] Step 1.10: Construct a multi-dimensional electronic intelligence knowledge ontology library, including attributes such as time, location, and track of reconnaissance missions.

[0087] Furthermore, step 1.8: construct a radar radiation source knowledge ontology library, and the specific method is: use radar model, purpose, technical system, operating band, radar carrying platform, radar manufacturing country, radar country, radar user unit as the radar model knowledge ontology library of the knowledge graph; use radar reconnaissance indicator system, including EDW (radiator descriptor) and typical PDW (pulse descriptor) sequence as the radar signal knowledge ontology library; the two are combined into the radar radiation source knowledge ontology library.

[0088] Furthermore, step 1.9: construct an electronic reconnaissance equipment knowledge ontology library, and the specific method is: the electronic reconnaissance equipment model, technical system, adaptive frequency band, instantaneous bandwidth, instantaneous airspace coverage, electronic reconnaissance equipment carrying platform, electronic reconnaissance equipment producing country, electronic reconnaissance equipment country of origin, and electronic reconnaissance equipment user unit are the electronic reconnaissance equipment knowledge ontology library.

[0089] Furthermore, step 1.10: construct a multidimensional electronic intelligence knowledge ontology library, and the specific method is: use the reconnaissance mission time, location, track and equipment used, reconnaissance target pulse parameters and radiation source parameters (including EDW and all intercepted PDW), target model, platform, target direction finding, positioning information, etc. as the knowledge graph of the multidimensional electronic intelligence knowledge ontology library.

[0090] (2) Construction of electronic reconnaissance intelligence entity

[0091] Step 2.1: Receive and process the reconnaissance results of the electronic reconnaissance equipment in real time through real-time stream processing technology.

[0092] Step 2.2: Batch post-processing of the admitted reconnaissance data, including data preprocessing, cleaning and labeling.

[0093] Step 2.3: Import the processed data into the distributed data storage system to ensure persistent storage of the data.

[0094] Step 2.4: According to the parameter system defined by the ontology model, automatically construct the radar radiation source entity class, electronic reconnaissance equipment entity class and multi-dimensional electronic intelligence related entity class.

[0095] Step 2.5: Ensure that the entity class corresponds to the parameter system of the ontology model one by one to ensure the accuracy and consistency of the data.

[0096] (3) Knowledge reasoning completion based on rule model and semantic model

[0097] Step 3.1: Formulate inference rules, including the conditional part (antecedent) and the conclusion part (consequent), such as "If entity A is a radar and its operating frequency is in the X-band, then entity A may be a radar used for air surveillance."

[0098] Step 3.2: Encode the inference rules into executable logical expressions and store them in a rule base, such as using the Drools rule engine for management.

[0099] Step 3.3: Combine with big data processing frameworks such as Spark Streaming to implement real-time reasoning processing of data streams to support dynamic intelligence analysis.

[0100] Step 3.4: Execute inference rules to generate new knowledge triples through logical reasoning to expand the depth and breadth of the knowledge graph.

[0101] Step 3.5: Extract knowledge from radar countermeasure field text data, compiled structured data, and organized expert knowledge data to build a corpus. Use NLP technology to perform word segmentation, part-of-speech tagging, and dependency analysis on text data to form a structured knowledge expression.

[0102] Step 3.6: Build a semantic index, including a word vector model and a synonym dictionary, to support the calculation of semantic similarity.

[0103] Step 3.7: Identify potential relationships between entities by matching entity descriptions with semantic indices, and use these relationships as input for reasoning.

[0104] Step 3.8: Develop a semantic reasoning module to implement context-based entity relationship reasoning, such as using graph neural networks (GNNs) to perform link prediction on knowledge graphs.

[0105] Step 3.9: Set a threshold for the query hit count and dynamically adjust the threshold based on the importance of the inference results to balance the accuracy and coverage of the inference.

[0106] Step 3.10: When the inference result reaches the preset threshold, the knowledge graph is automatically updated and new entity relationships are added to achieve continuous proliferation of the knowledge graph.

[0107] In summary, the present invention adopts a combination of a distributed data storage system and a non-traditional NOSQL non-relational database to construct an electronic intelligence knowledge graph that can be used for big data analysis and assessment of electronic intelligence and situations.

[0108] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0109] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph, characterized in that: The following steps are involved: (1) Knowledge graph ontology model construction In the data layer of the knowledge graph, a distributed data storage system is used to store the original data files recorded by the electronic reconnaissance equipment, and a unique access URI of the distributed storage system is generated. The URI will be stored as an entity in the non-relational database; The ontology layer is managed by a labeled ontology library, which uses the ontology library's support for axioms, rules and constraints to standardize entities, relationships, entity types and attributes; the entity layer stores multi-dimensional electronic intelligence entity information including time domain, frequency domain, space domain and parameter domain and their mutual relationships; (2) Construction of electronic intelligence entities By using real-time stream processing and batch post-processing based on recorded data, the real-time reconnaissance results of electronic reconnaissance equipment or recorded reconnaissance data are imported into the distributed data storage system, and after pre-processing, data cleaning, automatic labeling and compilation, the electronic intelligence entity is constructed; (3) Knowledge reasoning completion based on rule model and semantic model Extract knowledge from text data, structured data, and expert knowledge data in the field of radar countermeasures according to predetermined rules, and form a corpus of subject-predicate-object structure after segmentation and merging. Construct a synonym index of the vocabulary in the corpus, including words and phrases, and construct two types of knowledge triples: (subject-predicate-object) and (vocabulary-synonym-synonym). After data cleaning and ontology matching and association according to the knowledge graph ontology model, import it into a non-relational database to form a semantic association model. In the process of reconnaissance data processing, after the electronic intelligence entity is generated, a set of descriptive words is extracted from the newly generated entity and matched with the vocabulary and its synonyms and near-synonyms in the semantic association model to establish a relationship. At the same time, it is used as the input of the semantic association model to query the ontologies and entities associated with it, and the number of query hits is counted. When the number of hits reaches the threshold, the relationship with the corresponding ontology and entity is constructed to realize knowledge reasoning completion based on the rule model.

2. The method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph according to claim 1 is characterized in that: (1) Construction of knowledge graph ontology model. The specific method is as follows: Step 1.1: Build the data layer, deploy a distributed data storage system, and ensure that the distributed data storage system has high availability and scalability; Step 1.2: Distribute and store the raw data files collected by the electronic reconnaissance equipment, and generate a unique access URI based on UUID for each file to ensure the uniqueness and traceability of the data; Step 1.3: Build the ontology layer, define the concepts, attributes, and relationships of the ontology, and ensure the logical consistency and integrity of the ontology layer; Step 1.4: Create a labeled ontology library to manage and maintain the ontology layer information of the knowledge graph; Step 1.5: Store axioms, rules, and constraints in the ontology repository to regulate entities, relations, entity types, and attributes; Step 1.6: Build the entity layer, define entity types and attributes, and the relationships between entities, and ensure that the entity layer can store multi-dimensional information; Step 1.7: Using a top-down construction approach, extract ontology and pattern information from the revised formatted electronic intelligence documents and add them to the knowledge base; Step 1.8: Construct a radar radiation source knowledge ontology library, with radar model, purpose, technical system, operating band, radar carrying platform, radar manufacturing country, radar country, and radar user unit as the radar model knowledge ontology library of the knowledge graph; with radar reconnaissance indicator system, including EDW and PDW sequences as the radar signal knowledge ontology library, and the two are combined into the radar radiation source knowledge ontology library; Step 1.9: Construct an electronic reconnaissance equipment knowledge ontology library, with the electronic reconnaissance equipment model, technical system, adaptive frequency band, instantaneous bandwidth, instantaneous airspace coverage, electronic reconnaissance equipment carrying platform, electronic reconnaissance equipment production country, electronic reconnaissance equipment country of origin, and electronic reconnaissance equipment user as the electronic reconnaissance equipment knowledge ontology library; Step 1.10: Construct a multi-dimensional electronic intelligence knowledge ontology library, with the time, location, track and equipment used in the reconnaissance mission, the pulse parameters and radiation source parameters of the reconnaissance target, the target model, the platform to which it belongs, the target direction finding, and the positioning information as the knowledge graph of the multi-dimensional electronic intelligence knowledge ontology library.

3. The method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph according to claim 2 is characterized in that: (2) Construction of electronic intelligence entity: The specific method is as follows: Step 2.1: Receive and process the reconnaissance results of the electronic reconnaissance equipment in real time through real-time stream processing technology; Step 2.2: Batch post-processing of the admitted reconnaissance data, including data preprocessing, cleaning and labeling; Step 2.3: Import the processed data into the distributed data storage system to ensure persistent storage of the data; Step 2.4: According to the parameter system defined by the knowledge graph ontology model, construct the radar radiation source entity class, electronic reconnaissance equipment entity class and multi-dimensional electronic intelligence related entity class to ensure that the entity class corresponds to the parameter system of the ontology model one by one to ensure the accuracy and consistency of the data.

4. The method for constructing a time-space-frequency multi-domain electronic intelligence knowledge graph according to claim 1 is characterized in that: (3) Knowledge reasoning completion based on rule model and semantic model. The specific method is as follows: Step 3.1: Formulate inference rules, including conditional and conclusion parts; Step 3.2: Encode the inference rules into executable logical expressions and store them in the rule base, which is managed using the Drools rule engine. Step 3.3: Combined with the Spark Streaming framework, build a real-time reasoning processing system based on data streams to support dynamic intelligence analysis; Step 3.4: Execute the inference rules to generate new knowledge triples through logical reasoning, expanding the depth and breadth of the knowledge graph; Step 3.5: Extract knowledge from radar countermeasure field text data, compiled structured data, and organized expert knowledge data, build a corpus, and use NLP technology to perform word segmentation, part-of-speech tagging, and dependency analysis on text data to form a structured knowledge expression; Step 3.6: Build a semantic index, including a word vector model and a synonym dictionary, to support the calculation of semantic similarity; Step 3.7: Identify potential relationships between entities by matching entity descriptions with semantic indices, and use these relationships as input for reasoning; Step 3.8: Set a threshold for the number of query hits and dynamically adjust the threshold based on the importance of the inference results to balance the accuracy and coverage of the inference; Step 3.9: Use graph neural network to predict links in knowledge graph and realize context-based entity relationship reasoning; Step 3.10: When the inference result reaches the threshold preset in step 3.8, the knowledge graph is automatically updated and new entity relationships are added to achieve continuous proliferation of the knowledge graph.

5. A time-space-frequency multi-domain electronic intelligence knowledge graph construction system, characterized in that: Implement the method for constructing a spatiotemporal, frequency-multi-domain electronic intelligence knowledge graph as described in any one of claims 1-4 to realize the construction of a spatiotemporal, frequency-multi-domain electronic intelligence knowledge graph.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for constructing a spatiotemporal and frequency-multi-domain electronic intelligence knowledge graph as described in any one of claims 1 to 4 is implemented to realize the construction of a spatiotemporal and frequency-multi-domain electronic intelligence knowledge graph.

7. A computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for constructing a spatiotemporal, frequency-domain electronic intelligence knowledge graph as described in any one of claims 1 to 4 is implemented to realize the construction of a spatiotemporal, frequency-domain electronic intelligence knowledge graph.

Images